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
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
40 JH/02 Support "G" suffix to numbers in ${if comparisons.
42 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
44 NM/01 Bugzilla 1197 - Spec typo
45 Bugzilla 1196 - Spec examples corrections
47 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
49 PP/09 Add gnutls_enable_pkcs11 option.
51 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
52 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
54 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
55 "acl {{name}{arg}...}", and optional args on acl condition
58 JH/05 Permit multiple router/transport headers_add/remove lines.
63 PP/01 Handle short writes when writing local log-files.
64 In practice, only affects FreeBSD (8 onwards).
65 Bugzilla 1053, with thanks to Dmitry Isaikin.
67 NM/01 Bugzilla 949 - Documentation tweak
69 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
72 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
74 PP/02 Implemented gsasl authenticator.
76 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
78 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
79 `pkg-config foo` for cflags/libs.
81 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
82 with rest of GSASL and with heimdal_gssapi.
84 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
85 `pkg-config foo` for cflags/libs for the TLS implementation.
87 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
88 properties get this fed in as external SSF. A number of robustness
89 and debugging improvements to the cyrus_sasl authenticator.
91 PP/08 cyrus_sasl server now expands the server_realm option.
93 PP/09 Bugzilla 1214 - Log authentication information in reject log.
94 Patch by Jeremy Harris.
96 PP/10 Added dbmjz lookup type.
98 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
100 PP/12 MAIL args handles TAB as well as SP, for better interop with
101 non-compliant senders.
102 Analysis and variant patch by Todd Lyons.
104 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
105 Bug report from Lars Müller <lars@samba.org> (via SUSE),
106 Patch from Dirk Mueller <dmueller@suse.com>
108 PP/13 tls_peerdn now print-escaped for spool files.
109 Observed some $tls_peerdn in wild which contained \n, which resulted
110 in spool file corruption.
112 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
113 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
114 or write after TLS renegotiation, which otherwise led to messages
117 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
118 as a tracking header (ie: a signed header comes before the signature).
119 Patch from Wolfgang Breyha.
121 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
122 comma-sep list; embedded commas doubled.
124 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
126 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
128 Report and patch from Dmitry Banschikov.
130 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
131 Removed SSL_clear() after SSL_new() which led to protocol negotiation
132 failures. We appear to now support TLS1.1+ with Exim.
134 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
135 lets Exim select keys and certificates based upon TLS SNI from client.
136 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
137 before an outbound SMTP session. New log_selector, +tls_sni.
139 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
140 NULL dereference. Report and patch from Alun Jones.
142 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
143 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
144 Not seeing resolver debug output on NetBSD, but suspect this is a
145 resolver implementation change.
147 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
148 Left warnings. Added "eximon gdb" invocation mode.
150 PP/21 Defaulting "accept_8bitmime" to true, not false.
152 PP/22 Added -bw for inetd wait mode support.
154 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
155 locate the relevant includes and libraries. Made this the default.
157 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
158 Bugzilla 1246, report and most of solution from Tomasz Kusy.
160 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
161 This may cause build issues on older platforms.
163 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
164 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
165 gnutls_require_mac & gnutls_require_protocols (no longer supported).
166 Added SNI support via GnuTLS too.
167 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
169 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
171 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
172 Applied second patch from Janne, implementing suggestion to default
173 multiple-strings-in-record handling to match SPF spec.
175 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
177 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
178 read-only, out of scope).
179 Patch from Wolfgang Breyha, report from Stuart Northfield.
181 PP/29 Fix three issues highlighted by clang analyser static analysis.
182 Only crash-plausible issue would require the Cambridge-specific
183 iplookup router and a misconfiguration.
184 Report from Marcin Mirosław.
186 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
188 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
189 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
190 As part of this, removing so much warning spew let me fix some minor
191 real issues in debug logging.
193 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
194 assignment on my part. Fixed.
196 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
197 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
198 Janne Snabb (who went above and beyond: thank you).
200 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
201 string otherwise requires a connection and a bunch more work and it's
202 relatively easy to get wrong. Should also expose TLS library linkage
205 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
206 64-bit ${eval} (JH/03).
208 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
209 GNU libc to support some of the 64-bit stuff, should not lead to
210 conflicts. Defined before os.h is pulled in, so if a given platform
211 needs to override this, it can.
213 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
214 protection layer was required, which is not implemented.
215 Bugzilla 1254, patch from Wolfgang Breyha.
217 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
218 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
219 tls_dhparam take prime identifiers. Also unbreak combination of
220 OpenSSL+DH_params+TLSSNI.
222 PP/39 Disable SSLv2 by default in OpenSSL support.
228 PP/01 Solaris build fix for Oracle's LDAP libraries.
229 Bugzilla 1109, patch from Stephen Usher.
231 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
233 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
236 TF/02 Fix a couple more cases where we did not log the error message
237 when unlink() failed. See also change 4.74-TF/03.
239 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
240 lock up or crash if it happened to be inside a call to libc when it
241 got a SIGUSR1 from exiwhat.
243 The SIGUSR1 handler appends the current process status to the process
244 log which is later printed by exiwhat. It used to use the general
245 purpose logging code to do this, but several functions it calls are
246 not safe for signals.
248 The new output code in the SIGUSR1 handler is specific to the process
249 log, and simple enough that it's easy to inspect for signal safety.
250 Removing some special cases also simplifies the general logging code.
251 Removing the spurious timestamps from the process log simplifies
254 TF/04 Improved ratelimit ACL condition.
256 The /noupdate option has been deprecated in favour of /readonly which
257 has clearer semantics. The /leaky, /strict, and /readonly update modes
258 are mutually exclusive. The update mode is no longer included in the
259 database key; it just determines when the database is updated. (This
260 means that when you upgrde Exim will forget old rate measurements.)
262 Exim now checks that the per_* options are used with an update mode that
263 makes sense for the current ACL. For example, when Exim is processing a
264 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
265 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
266 must specify per_mail/readonly. If you omit the update mode it defaults to
267 /leaky where that makes sense (as before) or /readonly where required.
269 The /noupdate option is now undocumented but still supported for
270 backwards compatibility. It is equivalent to /readonly except that in
271 ACLs where /readonly is required you may specify /leaky/noupdate or
272 /strict/noupdate which are treated the same as /readonly.
274 A useful new feature is the /count= option. This is a generalization
275 of the per_byte option, so that you can measure the throughput of other
276 aggregate values. For example, the per_byte option is now equivalent
277 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
279 The per_rcpt option has been generalized using the /count= mechanism
280 (though it's more complicated than the per_byte equivalence). When it is
281 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
282 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
283 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
284 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
285 Note that using per_rcpt with a non-readonly update mode in more than
286 one ACL will cause the recipients to be double-counted. (The per_mail
287 and per_byte options don't have this problem.)
289 The handling of very low rates has changed slightly. If the computed rate
290 is less than the event's count (usually one) then this event is the first
291 after a long gap. In this case the rate is set to the same as this event's
292 count, so that the first message of a spam run is counted properly.
294 The major new feature is a mechanism for counting the rate of unique
295 events. The new per_addr option counts the number of different
296 recipients that someone has sent messages to in the last time period. It
297 behaves like per_rcpt if all the recipient addresses are different, but
298 duplicate recipient addresses do not increase the measured rate. Like
299 the /count= option this is a general mechanism, so the per_addr option
300 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
301 example, measure the rate that a client uses different sender addresses
302 with the options per_mail/unique=$sender_address. There are further
303 details in the main documentation.
305 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
307 TF/06 Removed a few PCRE remnants.
309 TF/07 Automatically extract Exim's version number from tags in the git
310 repository when doing development or release builds.
312 PP/02 Raise smtp_cmd_buffer_size to 16kB.
313 Bugzilla 879. Patch from Paul Fisher.
315 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
316 Heavily based on revision 40f9a89a from Simon Arlott's tree.
319 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
321 PP/05 Variable $av_failed, true if the AV scanner deferred.
322 Bugzilla 1078. Patch from John Horne.
324 PP/06 Stop make process more reliably on build failure.
325 Bugzilla 1087. Patch from Heiko Schlittermann.
327 PP/07 Make maildir_use_size_file an _expandable_ boolean.
328 Bugzilla 1089. Patch from Heiko Schlittermann.
330 PP/08 Handle ${run} returning more data than OS pipe buffer size.
331 Bugzilla 1131. Patch from Holger Weiß.
333 PP/09 Handle IPv6 addresses with SPF.
334 Bugzilla 860. Patch from Wolfgang Breyha.
336 PP/10 GnuTLS: support TLS 1.2 & 1.1.
338 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
341 PP/11 match_* no longer expand right-hand-side by default.
342 New compile-time build option, EXPAND_LISTMATCH_RHS.
343 New expansion conditions, "inlist", "inlisti".
345 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
347 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
349 PP/14 fix log_write() format string regression from TF/03.
350 Bugzilla 1152. Patch from Dmitry Isaikin.
356 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
358 PP/02 Harmonised TLS library version reporting; only show if debugging.
359 Layout now matches that introduced for other libraries in 4.74 PP/03.
361 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
363 PP/04 New "dns_use_edns0" global option.
365 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
368 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
369 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
371 TK/01 Updated PolarSSL code to 0.14.2.
372 Bugzilla 1097. Patch from Andreas Metzler.
374 PP/07 Catch divide-by-zero in ${eval:...}.
377 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
380 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
381 format-string attack -- SECURITY: remote arbitrary code execution.
383 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
384 time unintentionally subject to list matching rules, letting the header
385 cause arbitrary Exim lookups (of items which can occur in lists, *not*
386 arbitrary string expansion). This allowed for information disclosure.
388 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
389 INT_MIN/-1 -- value coerced to INT_MAX.
395 NM/01 Workround for PCRE version dependency in version reporting
398 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
399 This fixes portability to compilers other than gcc, notably
400 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
402 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
403 makefiles for portability to HP-UX and POSIX correctness.
405 PP/01 Permit LOOKUP_foo enabling on the make command-line.
406 Also via indented variable definition in the Makefile.
407 (Debugging by Oliver Heesakkers).
409 PP/02 Restore caching of spamd results with expanded spamd_address.
410 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
412 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
413 Improves build reliability. Fix from: Frank Elsner
415 NM/02 Fix wide character breakage in the rfc2047 coding
416 Fixes bug 1064. Patch from Andrey N. Oktyabrski
418 NM/03 Allow underscore in dnslist lookups
419 Fixes bug 1026. Patch from Graeme Fowler
421 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
422 Code patches from Adam Ciarcinski of NetBSD.
424 NM/04 Fixed exiqgrep to cope with mailq missing size issue
427 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
428 is logged, to avoid truncation. Patch from John Horne.
430 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
431 Patch from Jakob Hirsch.
433 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
434 SQL string expansion failure details.
435 Patch from Andrey Oktyabrski.
437 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
438 Patch from Simon Arlott.
440 PP/09 New lookups functionality failed to compile on old gcc which rejects
441 extern declarations in function scope.
442 Patch from Oliver Fleischmann
444 PP/10 Use sig_atomic_t for flags set from signal handlers.
445 Check getgroups() return and improve debugging.
446 Fixed developed for diagnosis in bug 927 (which turned out to be
449 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
450 Patch from Mark Zealey.
452 PP/12 Bugzilla 1056: Improved spamd server selection.
453 Patch from Mark Zealey.
455 PP/13 Bugzilla 1086: Deal with maildir quota file races.
456 Based on patch from Heiko Schlittermann.
458 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
459 Patch from Uwe Doering, sign-off by Michael Haardt.
461 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
462 variable declaration deep within a block. Bug and patch from
465 PP/15 lookups-Makefile IRIX compatibilty coercion.
467 PP/16 Make DISABLE_DKIM build knob functional.
469 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
470 Patch by Simon Arlott
472 TF/03 Fix valgrind.h portability to C89 compilers that do not support
473 variable argument macros. Our copy now differs from upstream.
479 TF/01 Failure to get a lock on a hints database can have serious
480 consequences so log it to the panic log.
482 TF/02 Log LMTP confirmation messages in the same way as SMTP,
483 controlled using the smtp_confirmation log selector.
485 TF/03 Include the error message when we fail to unlink a spool file.
487 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
488 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
489 for maintaining out-of-tree patches for some time.
491 PP/01 Bugzilla 139: Documentation and portability issues.
492 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
493 Handle per-OS dynamic-module compilation flags.
495 PP/02 Let /dev/null have normal permissions.
496 The 4.73 fixes were a little too stringent and complained about the
497 permissions on /dev/null. Exempt it from some checks.
498 Reported by Andreas M. Kirchwitz.
500 PP/03 Report version information for many libraries, including
501 Exim version information for dynamically loaded libraries. Created
502 version.h, now support a version extension string for distributors
503 who patch heavily. Dynamic module ABI change.
505 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
506 privilege escalation vulnerability whereby the Exim run-time user
507 can cause root to append content of the attacker's choosing to
510 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
513 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
514 If dropping privileges for untrusted macros, we disabled normal logging
515 on the basis that it would fail; for the Exim run-time user, this is not
516 the case, and it resulted in successful deliveries going unlogged.
517 Fixed. Reported by Andreas Metzler.
523 PP/01 Date: & Message-Id: revert to normally being appended to a message,
524 only prepend for the Resent-* case. Fixes regression introduced in
525 Exim 4.70 by NM/22 for Bugzilla 607.
527 PP/02 Include check_rfc2047_length in configure.default because we're seeing
528 increasing numbers of administrators be bitten by this.
530 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
532 PP/03 Bugzilla 994: added openssl_options main configuration option.
534 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
536 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
538 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
540 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
541 without a peer certificate, leading to a segfault because of an
542 assumption that peers always have certificates. Be a little more
543 paranoid. Problem reported by Martin Tscholak.
545 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
546 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
547 NB: ClamAV planning to remove STREAM in "middle of 2010".
548 CL also introduces -bmalware, various -d+acl logging additions and
549 more caution in buffer sizes.
551 PP/09 Implemented reverse_ip expansion operator.
553 PP/10 Bugzilla 937: provide a "debug" ACL control.
555 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
557 PP/12 Bugzilla 973: Implement --version.
559 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
561 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
563 PP/15 Bugzilla 816: support multiple condition rules on Routers.
565 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
566 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
567 ignore trailing whitespace.
569 JJ/02 prevent non-panic DKIM error from being sent to paniclog
571 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
574 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
575 Notification from Dr Andrew Aitchison.
577 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
578 ExtendedDetectionInfo response format.
579 Notification from John Horne.
581 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
584 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
585 XSL and documented dependency on system catalogs, with examples of how
588 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
591 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
592 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
593 configuration file which is writeable by the Exim user or group.
595 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
596 of configuration files to cover files specified with the -C option if
597 they are going to be used with root privileges, not just the default
600 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
601 option (effectively making it always true).
603 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
604 files to be used while preserving root privileges.
606 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
607 that rogue child processes cannot use them.
609 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
610 run-time user, instead of root.
612 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
613 Exim run-time user without dropping privileges.
615 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
616 result string, instead of calling string_vformat() twice with the same
619 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
620 for other users. Others should always drop root privileges if they use
621 -C on the command line, even for a whitelisted configure file.
623 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
625 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
631 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
632 $data_path, and $header_path variables; fixed documentation bugs and
635 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
636 exipick to access non-standard spools, including the "frozen" queue
639 NM/01 Bugzilla 965: Support mysql stored procedures.
640 Patch from Alain Williams
642 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
644 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
645 Patch from Andreas Metzler
647 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
648 Patch from Kirill Miazine
650 NM/05 Bugzilla 671: Added umask to procmail example.
652 JJ/03 installed exipick 20100323.0, fixing doc bug
654 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
655 directory. Notification and patch from Dan Rosenberg.
657 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
659 TK/02 Improve log output when DKIM signing operation fails.
661 MH/01 Treat the transport option dkim_domain as a colon separated
662 list, not as a single string, and sign the message with each element,
663 omitting multiple occurences of the same signer.
665 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
666 Bugzilla 985, 986. Patch by Simon Arlott
668 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
669 Patch by Simon Arlott
671 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
672 MBX locking. Notification from Dan Rosenberg.
678 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
680 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
682 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
684 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
686 NM/04 Bugzilla 915: Flag broken perl installation during build.
692 TK/01 Added patch by Johannes Berg that expands the main option
693 "spamd_address" if it starts with a dollar sign.
695 TK/02 Write list of recipients to X-Envelope-Sender header when building
696 the mbox-format spool file for content scanning (suggested by Jakob
699 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
700 (http://www.dcc-servers.net/) support via dccifd. Activated by
701 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
703 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
704 by Mark Daniel Reidel <mr@df.eu>.
706 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
707 When building exim an external PCRE library is now needed -
708 PCRE is a system library on the majority of modern systems.
709 See entry on PCRE_LIBS in EDITME file.
711 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
712 conversation. Added nologin parameter to request.
713 Patch contributed by Kirill Miazine.
715 TF/01 Do not log submission mode rewrites if they do not change the address.
717 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
719 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
720 log files in place. Contributed by Roberto Lima.
722 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
724 TF/03 Bugzilla 615: When checking the local_parts router precondition
725 after a local_part_suffix or local_part_prefix option, Exim now
726 does not use the address's named list lookup cache, since this
727 contains cached lookups for the whole local part.
729 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
730 Robert Millan. Documentation is in experimental-spec.txt.
732 TF/04 Bugzilla 668: Fix parallel build (make -j).
734 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
736 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
737 Patch provided by Jan Srzednicki.
739 TF/05 Leading white space used to be stripped from $spam_report which
740 wrecked the formatting. Now it is preserved.
742 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
743 that they are available at delivery time.
745 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
747 TF/08 TLS error reporting now respects the incoming_interface and
748 incoming_port log selectors.
750 TF/09 Produce a more useful error message if an SMTP transport's hosts
751 setting expands to an empty string.
753 NM/06 Bugzilla 744: EXPN did not work under TLS.
754 Patch provided by Phil Pennock.
756 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
757 Patch provided by Richard Godbee.
759 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
760 acl_smtp_notquit, added index entry.
762 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
763 Patch provided by Eugene Bujak.
765 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
766 accept(). Patch provided by Maxim Dounin.
768 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
769 Patch provided by Phil Pennock.
771 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
773 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
774 Patch provided by Brad "anomie" Jorsch.
776 NM/14 Bugzilla 622: Added timeout setting to transport filter.
777 Patch provided by Dean Brooks.
779 TK/05 Add native DKIM support (does not depend on external libraries).
781 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
782 Patch provided by Graeme Fowler.
784 NM/16 Bugzilla 851: Documentation example syntax fix.
786 NM/17 Changed NOTICE file to remove references to embedded PCRE.
788 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
791 NM/19 Bugzilla 745: TLS version reporting.
792 Patch provided by Phil Pennock.
794 NM/20 Bugzilla 167: bool: condition support.
795 Patch provided by Phil Pennock.
797 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
798 clients. Patch provided by Phil Pennock.
800 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
801 Patch provided by Brad "anomie" Jorsch.
803 NM/23 Bugzilla 687: Fix misparses in eximstats.
804 Patch provided by Heiko Schlittermann.
806 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
807 Patch provided by Heiko Schlittermann.
809 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
810 plus update to original patch.
812 NM/26 Bugzilla 799: Documentation correction for ratelimit.
814 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
815 Patch provided by David Brownlee.
817 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
819 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
821 NM/30 Bugzilla 888: TLS documentation bugfixes.
823 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
825 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
826 Unlike the original bugzilla I have changed all shell scripts in src tree.
828 NM/33 Bugzilla 898: Transport filter timeout fix.
829 Patch by Todd Rinaldo.
831 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
832 Patch by Serge Demonchaux.
834 NM/35 Bugzilla 39: Base64 decode bug fixes.
835 Patch by Jakob Hirsch.
837 NM/36 Bugzilla 909: Correct connect() call in dcc code.
839 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
841 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
843 NM/39 Bugzilla 911: Fixed MakeLinks build script.
849 TK/01 Add preliminary DKIM support. Currently requires a forked version of
850 ALT-N's libdkim that I have put here:
851 http://duncanthrax.net/exim-experimental/
853 Note to Michael Haardt: I had to rename some vars in sieve.c. They
854 were called 'true' and it seems that C99 defines that as a reserved
855 keyword to be used with 'bool' variable types. That means you could
856 not include C99-style headers which use bools without triggering
857 build errors in sieve.c.
859 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
860 as mailq or other aliases. Changed the --help handling significantly
861 to do whats expected. exim_usage() emits usage/help information.
863 SC/01 Added the -bylocaldomain option to eximstats.
865 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
867 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
869 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
875 PH/01 Another patch from the Sieve maintainer.
877 PH/02 When an IPv6 address is converted to a string for single-key lookup
878 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
879 dots are used instead of colons so that keys in lsearch files need not
880 contain colons. This was done some time before quoting was made available
881 in lsearch files. However, iplsearch files do require colons in IPv6 keys
882 (notated using the quote facility) so as to distinguish them from IPv4
883 keys. This meant that lookups for IP addresses in host lists did not work
884 for iplsearch lookups.
886 This has been fixed by arranging for IPv6 addresses to be expressed with
887 colons if the lookup type is iplsearch. This is not incompatible, because
888 previously such lookups could never work.
890 The situation is now rather anomolous, since one *can* have colons in
891 ordinary lsearch keys. However, making the change in all cases is
892 incompatible and would probably break a number of configurations.
894 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
897 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
898 conversion specification without a maximum field width, thereby enabling
899 a rogue spamd server to cause a buffer overflow. While nobody in their
900 right mind would setup Exim to query an untrusted spamd server, an
901 attacker that gains access to a server running spamd could potentially
902 exploit this vulnerability to run arbitrary code as the Exim user.
904 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
905 $primary_hostname instead of what libspf2 thinks the hosts name is.
907 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
908 a directory entry by the name of the lookup key. Previously, if a
909 symlink pointed to a non-existing file or a file in a directory that
910 Exim lacked permissions to read, a lookup for a key matching that
911 symlink would fail. Now it is enough that a matching directory entry
912 exists, symlink or not. (Bugzilla 503.)
914 PH/03 The body_linecount and body_zerocount variables are now exported in the
917 PH/04 Added the $dnslist_matched variable.
919 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
920 This means they are set thereafter only if the connection becomes
923 PH/06 Added the client_condition to authenticators so that some can be skipped
924 by clients under certain conditions.
926 PH/07 The error message for a badly-placed control=no_multiline_responses left
927 "_responses" off the end of the name.
929 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
931 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
932 (without spaces) instead of just copying the configuration text.
934 PH/10 Added the /noupdate option to the ratelimit ACL condition.
936 PH/11 Added $max_received_linelength.
938 PH/12 Added +ignore_defer and +include_defer to host lists.
940 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
941 way in which PCRE > 7.0 is built.
943 PH/14 Implemented queue_only_load_latch.
945 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
946 MAIL command. The effect was to mangle the value on 64-bit systems.
948 PH/16 Another patch from the Sieve maintainer.
950 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
952 PH/18 If a system quota error occurred while trying to create the file for
953 a maildir delivery, the message "Mailbox is full" was not appended to the
954 bounce if the delivery eventually timed out. Change 4.67/27 below applied
955 only to a quota excession during the actual writing of the file.
957 PH/19 It seems that peer DN values may contain newlines (and other non-printing
958 characters?) which causes problems in log lines. The DN values are now
959 passed through string_printing() before being added to log lines.
961 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
962 and InterBase are left for another time.)
964 PH/21 Added message_body_newlines option.
966 PH/22 Guard against possible overflow in moan_check_errorcopy().
968 PH/23 POSIX allows open() to be a macro; guard against that.
970 PH/24 If the recipient of an error message contained an @ in the local part
971 (suitably quoted, of course), incorrect values were put in $domain and
972 $local_part during the evaluation of errors_copy.
978 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
979 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
982 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
983 issue a MAIL command.
985 PH/02 In an ACL statement such as
987 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
989 if a client was not listed at all, or was listed with a value other than
990 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
991 the condition was not true (as it should be), so access was not denied.
992 The bug was that the ! inversion was incorrectly passed on to the second
993 item. This has been fixed.
995 PH/03 Added additional dnslists conditions == and =& which are different from
996 = and & when the dns lookup returns more than one IP address.
998 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
999 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1001 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1002 FSYNC, which compiles an option called disable_fsync that allows for
1003 bypassing fsync(). The documentation is heavily laced with warnings.
1005 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1007 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1008 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1009 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1010 including adding "make clean"; (3) Added -fPIC when compiling the test
1011 dynamically loaded module, to get rid of a warning.
1013 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1014 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1015 = 0s. The bug is otherwise harmless.
1017 PH/07 There was a bug in the dovecot authenticator such that the value of
1018 $auth1 could be overwritten, and so not correctly preserved, after a
1019 successful authentication. This usually meant that the value preserved by
1020 the server_setid option was incorrect.
1022 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1024 PH/09 Installed PCRE release 7.0.
1026 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1027 run for batched SMTP input. It is now run at the start of every message
1028 in the batch. While fixing this I discovered that the process information
1029 (output by running exiwhat) was not always getting set for -bs and -bS
1030 input. This is fixed, and it now also says "batched" for BSMTP.
1032 PH/11 Added control=no_pipelining.
1034 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1035 patch, slightly modified), and move the expansion of helo_data till after
1036 the connection is made in the smtp transport (so it can use these
1039 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1041 PH/14 Added log_selector = +pid.
1043 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1045 PH/16 Add ${if forany and ${if forall.
1047 PH/17 Added dsn_from option to vary the From: line in DSNs.
1049 PH/18 Flush SMTP output before performing a callout, unless control =
1050 no_callout_flush is set.
1052 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1053 was true (the default) a successful delivery failed to delete the retry
1054 item, thus causing premature timeout of the address. The bug is now
1057 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1059 PH/21 Long custom messages for fakedefer and fakereject are now split up
1060 into multiline reponses in the same way that messages for "deny" and
1061 other ACL rejections are.
1063 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1064 with slight modification.
1066 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1067 draft, changing the syntax and factoring some duplicate code".
1069 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1070 for deliveries of the second and subsequent messages over the same SMTP
1073 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1074 ${reduce, with only minor "tidies".
1076 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1078 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1079 expansion side effects.
1081 PH/27 When a message times out after an over-quota error from an Exim-imposed
1082 quota, the bounce message says "mailbox is full". This message was not
1083 being given when it was a system quota that was exceeded. It now should
1086 MH/03 Made $recipients available in local_scan(). local_scan() already has
1087 better access to the recipient list through recipients_list[], but
1088 $recipients can be useful in postmaster-provided expansion strings.
1090 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1091 in the case of a MAIL command with additional options following the
1092 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1093 were accidentally chopped off.
1095 PH/29 SMTP synchronization checks are implemented when a command is read -
1096 there is a check that no more input is waiting when there shouldn't be
1097 any. However, for some commands, a delay in an ACL can mean that it is
1098 some time before the response is written. In this time, more input might
1099 arrive, invalidly. So now there are extra checks after an ACL has run for
1100 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1101 pipelining has not been advertised.
1103 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1105 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1106 returned in the SMTP response when smtp_return_error_details was false.
1107 This has been fixed.
1109 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1110 instead of the C I/O that was originally supplied, because problems were
1111 reported on Solaris.
1113 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1114 Exim which did not show up earlier: it was assuming that a call to
1115 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1116 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1117 was a macro that became an assignment, so it seemed to work. This has
1118 changed to a proper function call with a void return, hence the compile
1119 error. Exim's code has been fixed.
1121 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1124 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1126 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1128 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1129 because it turns out to be too big or there is a timeout) there is no
1130 "Completed" line in the log. When some messages of this type were
1131 selected by exigrep, they were listed as "not completed". Others were
1132 picked up by some special patterns. I have improved the selection
1133 criteria to be more general.
1135 PH/38 The host_find_failed option in the manualroute router can now be set
1136 to "ignore", to completely ignore a host whose IP address cannot be
1137 found. If all hosts are ignored, the behaviour is controlled by the new
1138 host_all_ignored option.
1140 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1141 homing or because of multiple MX records with /mx) generated more than
1142 one IP address, and the following item turned out to be the local host,
1143 all the secondary addresses of the first item were incorrectly removed
1144 from the list, along with the local host and any following hosts (which
1145 is what is supposed to happen).
1147 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1148 whoever called Exim into the -H file. In the case of the daemon it was
1149 behaving confusingly. When first started, it used values for whoever
1150 started the daemon, but after a SIGHUP it used the Exim user (because it
1151 calls itself on a restart). I have changed the code so that it now always
1154 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1155 message are rejected with the same error (e.g. no authentication or bad
1156 sender address), and a DATA command is nevertheless sent (as can happen
1157 with PIPELINING or a stupid MUA), the error message that was given to the
1158 RCPT commands is included in the rejection of the DATA command. This is
1159 intended to be helpful for MUAs that show only the final error to their
1162 PH/42 Another patch from the Sieve maintainer.
1164 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1165 Eximstats - Fixed some broken HTML links and added missing column headers
1167 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1168 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1170 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1172 PH/43 Yet another patch from the Sieve maintainer.
1174 PH/44 I found a way to check for a TCP/IP connection going away before sending
1175 the response to the final '.' that terminates a message, but only in the
1176 case where the client has not sent further data following the '.'
1177 (unfortunately, this is allowed). However, in many cases there won't be
1178 any further data because there won't be any more messages to send. A call
1179 to select() can be used: if it shows that the input is "ready", there is
1180 either input waiting, or the socket has been closed. An attempt to read
1181 the next input character can distinguish the two cases. Previously, Exim
1182 would have sent an OK response which the client would never have see.
1183 This could lead to message repetition. This fix should cure that, at
1184 least in a lot of common cases.
1186 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1187 advertised in response to EHLO.
1193 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1194 fixed by 4.65/MH/01 (is this a record?) are fixed:
1196 (i) An empty string was always treated as zero by the numeric comparison
1197 operators. This behaviour has been restored.
1199 (ii) It is documented that the numeric comparison operators always treat
1200 their arguments as decimal numbers. This was broken in that numbers
1201 starting with 0 were being interpreted as octal.
1203 While fixing these problems I realized that there was another issue that
1204 hadn't been noticed. Values of message_size_limit (both the global option
1205 and the transport option) were treated as octal if they started with 0.
1206 The documentation was vague. These values are now always treated as
1207 decimal, and I will make that clear in the documentation.
1213 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1214 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1217 MH/01 Don't check that the operands of numeric comparison operators are
1218 integers when their expansion is in "skipping" mode (fixes bug
1219 introduced by 4.64-PH/07).
1221 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1222 child addresses, Exim now panics and dies. Previously, because the count
1223 is held in a short int, deliveries were likely to be lost. As such a
1224 large number of recipients for a single message is ridiculous
1225 (performance will be very, very poor), I have chosen to impose a limit
1226 rather than extend the field.
1232 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1233 leftover -K file (the existence of which was triggered by #402).
1234 While we were at it, introduced process PID as part of the -K
1235 filename. This should rule out race conditions when creating
1238 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1239 processing considerably. Previous code took too long for large mails,
1240 triggering a timeout which in turn triggers #401.
1242 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1243 in the DK code in transports.c. sendfile() is not really portable,
1244 hence the _LINUX specificness.
1246 TF/01 In the add_headers option to the mail command in an Exim filter,
1247 there was a bug that Exim would claim a syntax error in any
1248 header after the first one which had an odd number of characters
1251 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1252 callout verification, Exim cached a "reject" for the entire domain. This
1253 is correct for most verifications, but it is not correct for a recipient
1254 verification with use_sender or use_postmaster set, because in that case
1255 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1256 case of MAIL FROM:<> rejection from other early rejections (e.g.
1257 rejection of HELO). When verifying a recipient using a non-null MAIL
1258 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1259 Whatever the result of the callout, the value of the domain cache is
1260 left unchanged (for any other kind of callout, getting as far as trying
1261 RCPT means that the domain itself is ok).
1263 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1266 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1267 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1268 session, the daemon ignores ECONNECT errors and logs others; it now
1269 ignores EPIPE as well.
1271 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1272 (quoted-printable decoding).
1274 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1275 later the small subsequent patch to fix an introduced bug.
1277 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1279 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1281 PH/08 An error is now given if message_size_limit is specified negative.
1283 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1284 to be given (somewhat) arbitrary names.
1286 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1289 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1290 miscellaneous code fixes
1292 PH/10 Added the log_reject_target ACL modifier to specify where to log
1295 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1296 hostname. This is wrong, because it relates to the incoming message (and
1297 probably the interface on which it is arriving) and not to the outgoing
1298 callout (which could be using a different interface). This has been
1299 changed to use the value of the helo_data option from the smtp transport
1300 instead - this is what is used when a message is actually being sent. If
1301 there is no remote transport (possible with a router that sets up host
1302 addresses), $smtp_active_hostname is used.
1304 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1305 tweaks were necessary in order to get it to work (see also 21 below):
1306 (a) The code assumed that strncpy() returns a negative number on buffer
1307 overflow, which isn't the case. Replaced with Exim's string_format()
1309 (b) There were several signed/unsigned issues. I just did the minimum
1310 hacking in of casts. There is scope for a larger refactoring.
1311 (c) The code used strcasecmp() which is not a standard C function.
1312 Replaced with Exim's strcmpic() function.
1313 (d) The code set only $1; it now sets $auth1 as well.
1314 (e) A simple test gave the error "authentication client didn't specify
1315 service in request". It would seem that Dovecot has changed its
1316 interface. Fortunately there's a specification; I followed it and
1317 changed what the client sends and it appears to be working now.
1319 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1322 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1323 address A is aliased to B and C, where B exists and C does not. Without
1324 -v the output is "A verified" because verification stops after a
1325 successful redirection if more than one address is generated. However,
1326 with -v the child addresses are also verified. Exim was outputting "A
1327 failed to verify" and then showing the successful verification for C,
1328 with its parentage. It now outputs "B failed to verify", showing B's
1329 parentage before showing the successful verification of C.
1331 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1332 look up a TXT record in a specific list after matching in a combined
1335 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1336 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1337 they consult the DNS. I had assumed they would set it the way they
1338 wanted; and indeed my experiments on Linux seem to show that in some
1339 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1340 To be on the safe side, however, I have now made the interface to
1341 host_find_byname() similar to host_find_bydns(), with an argument
1342 containing the DNS resolver options. The host_find_byname() function now
1343 sets these options at its start, just as host_find_bydns() does. The smtp
1344 transport options dns_qualify_single and dns_search_parents are passed to
1345 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1346 of host_find_byname() use the default settings of RES_DEFNAMES
1347 (qualify_single) but not RES_DNSRCH (search_parents).
1349 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1350 spool_read_header() do less string testing, by means of a preliminary
1351 switch on the second character of optional "-foo" lines. (This is
1352 overdue, caused by the large number of possibilities that now exist.
1353 Originally there were few.) While I was there, I also converted the
1354 str(n)cmp tests so they don't re-test the leading "-" and the first
1355 character, in the hope this might squeeze out yet more improvement.
1357 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1358 flag allowing group syntax was set by the header_syntax check but not
1359 turned off, possible causing trouble later; (2) The flag was not being
1360 set at all for the header_verify test, causing "group"-style headers to
1361 be rejected. I have now set it in this case, and also caused header_
1362 verify to ignore an empty address taken from a group. While doing this, I
1363 came across some other cases where the code for allowing group syntax
1364 while scanning a header line wasn't quite right (mostly, not resetting
1365 the flag correctly in the right place). These bugs could have caused
1366 trouble for malformed header lines. I hope it is now all correct.
1368 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1369 with the "reply" argument non-NULL. The code, however (which originally
1370 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1371 but it didn't always do it. This confused somebody who was copying the
1372 code for some other use. I have removed all the tests.
1374 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1375 feature that was used to support insecure browsers during the U.S. crypto
1376 embargo. It requires special client support, and Exim is probably the
1377 only MTA that supported it -- and would never use it because real RSA is
1378 always available. This code has been removed, because it had the bad
1379 effect of slowing Exim down by computing (never used) parameters for the
1380 RSA_EXPORT functionality.
1382 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1383 authenticator to fail if there's a tab character in the incoming data
1384 (there should never be unless someone is messing about, as it's supposed
1385 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1386 if the connection is using TLS or if the remote IP is the same as the
1387 local IP, and the "valid-client-cert option" if a client certificate has
1390 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1391 authenticators. This can be used for authorization after authentication
1392 succeeds. (In the case of plaintext, it servers for both authentication
1395 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1396 if any retry times were supplied.
1398 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1399 connection, where there is no client IP address to check. In this
1400 situation, the verify now always succeeds.
1402 PH/25 Applied John Jetmore's -Mset patch.
1404 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1406 PH/27 In a string expansion for a processed (not raw) header when multiple
1407 headers of the same name were present, leading whitespace was being
1408 removed from all of them, but trailing whitespace was being removed only
1409 from the last one. Now trailing whitespace is removed from each header
1410 before concatenation. Completely empty headers in a concatenation (as
1411 before) are ignored.
1413 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1414 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1416 PH/29 [Removed. This was a change that I later backed out, and forgot to
1417 correct the ChangeLog entry (that I had efficiently created) before
1418 committing the later change.]
1420 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1421 address errors (4xx response to RCPT) over the same connection as other
1422 messages routed to the same hosts. Such deliveries are always "forced",
1423 so retry times are not inspected. This resulted in far too many retries
1424 for the affected addresses. The effect occurred only when there were more
1425 hosts than the hosts_max_try setting in the smtp transport when it had
1426 the 4xx errors. Those hosts that it had tried were not added to the list
1427 of hosts for which the message was waiting, so if all were tried, there
1428 was no problem. Two fixes have been applied:
1430 (i) If there are any address or message errors in an SMTP delivery, none
1431 of the hosts (tried or untried) are now added to the list of hosts
1432 for which the message is waiting, so the message should not be a
1433 candidate for sending over the same connection that was used for a
1434 successful delivery of some other message. This seems entirely
1435 reasonable: after all the message is NOT "waiting for some host".
1436 This is so "obvious" that I'm not sure why it wasn't done
1437 previously. Hope I haven't missed anything, but it can't do any
1438 harm, as the worst effect is to miss an optimization.
1440 (ii) If, despite (i), such a delivery is accidentally attempted, the
1441 routing retry time is respected, so at least it doesn't keep
1442 hammering the server.
1444 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1445 in ${readsocket because some servers need this prod.
1447 PH/32 Added some extra debug output when updating a wait-xxx database.
1449 PH/33 The hint "could be header name not terminated by colon", which has been
1450 given for certain expansion errors for a long time, was not being given
1451 for the ${if def:h_colon_omitted{... case.
1453 PH/34 The spec says: "With one important exception, whenever a domain list is
1454 being scanned, $domain contains the subject domain." There was at least
1455 one case where this was not true.
1457 PH/35 The error "getsockname() failed: connection reset by peer" was being
1458 written to the panic log as well as the main log, but it isn't really
1459 panic-worthy as it just means the connection died rather early on. I have
1460 removed the panic log writing for the ECONNRESET error when getsockname()
1463 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1464 runs only) independently of the message's sender address. This meant
1465 that, if the 4xx error was in fact related to the sender, a different
1466 message to the same recipient with a different sender could confuse
1467 things. In particualar, this can happen when sending to a greylisting
1468 server, but other circumstances could also provoke similar problems.
1469 I have changed the default so that the retry time for these errors is now
1470 based a combination of the sender and recipient addresses. This change
1471 can be overridden by setting address_retry_include_sender=false in the
1474 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1475 remote server are returned as part of bounce messages. This was not
1476 happening for LMTP over a pipe (the lmtp transport), but now it is the
1477 same for both kinds of LMTP.
1479 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1480 in header lines that were in fact CNAMEs. This is no longer the case.
1482 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1483 and queue runs started by the daemon processed all messages. This has
1484 been fixed so that -R and -S can now usefully be given with -q<time>.
1486 PH/40 Import PCRE release 6.7 (fixes some bugs).
1488 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1490 PH/42 Give an error if -q is specified more than once.
1492 PH/43 Renamed the variables $interface_address and $interface_port as
1493 $received_ip_address and $received_port, to make it clear that these
1494 values apply to message reception, and not to the outgoing interface when
1495 a message is delivered. (The old names remain recognized, of course.)
1497 PH/44 There was no timeout on the connect() call when using a Unix domain
1498 socket in the ${readsocket expansion. There now is.
1500 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1501 be meaningful with "accept".
1503 SC/01 Eximstats V1.43
1504 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1506 SC/02 Eximstats V1.44
1507 Use a glob alias rather than an array ref in the generated
1508 parser. This improves both readability and performance.
1510 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1511 Collect SpamAssassin and rejection statistics.
1512 Don't display local sender or destination tables unless
1513 there is data to show.
1514 Added average volumes into the top table text output.
1516 SC/04 Eximstats V1.46
1517 Collect data on the number of addresses (recipients)
1518 as well as the number of messages.
1520 SC/05 Eximstats V1.47
1521 Added 'Message too big' to the list of mail rejection
1522 reasons (thanks to Marco Gaiarin).
1524 SC/06 Eximstats V1.48
1525 Mainlog lines which have GMT offsets and are too short to
1526 have a flag are now skipped.
1528 SC/07 Eximstats V1.49 (Alain Williams)
1529 Added the -emptyok flag.
1531 SC/08 Eximstats V1.50
1532 Fixes for obtaining the IP address from reject messages.
1534 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1535 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1536 whitesspace changes from 4.64-PH/27
1538 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1541 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1542 are found, allow negative numbers in numeric criteria)
1544 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1546 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1547 to match changes made in 4.64-PH/43
1549 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1551 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1552 "domain" is actually an IP address, and force a failure. This locks out
1553 those revolvers/nameservers that support "A-for-A" lookups, in
1554 contravention of the specifications.
1556 PH/48 When a host name was looked up from an IP address, and the subsequent
1557 forward lookup of the name timed out, the host name was left in
1558 $sender_host_name, contrary to the specification.
1560 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1561 restricted to single-key lookups, Exim was not diagnosing an error if
1562 * or *@ was used with a query-style lookup.
1564 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1566 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1567 long ago, but noone interested enough thought of it. Let's just say that
1568 the "1.1" means that there are some new functions that weren't there at
1569 some point in the past.
1571 PH/51 Error processing for expansion failure of helo_data from an smtp
1572 transport during callout processing was broken.
1574 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1575 tested/used via the -bh/-bhc/-bs options.
1577 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1578 bug, fixed in subsequent PCRE releases).
1580 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1581 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1583 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1589 SC/01 Use a glob alias rather than an array ref in eximstats generated
1590 parser. This improves both readability and performance.
1592 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1593 Don't display local sender or destination tables in eximstats unless
1594 there is data to show.
1595 Added average volumes into the eximstats top table text output.
1597 SC/03 Collect data on the number of addresses (recipients) as well
1598 as the number of messages in eximstats.
1600 TF/01 Correct an error in the documentation for the redirect router. Exim
1601 does (usually) call initgroups() when daemonizing.
1603 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1604 with consistent privilege compared to when running as a daemon.
1606 TF/03 Note in the spec that $authenticated_id is not set for local
1607 submissions from trusted users.
1609 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1610 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1612 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1613 by adding some example configuration directives to the default
1614 configuration file. A little bit of work is required to uncomment the
1615 directives and define how usernames and passwords are checked, but
1616 there is now a framework to start from.
1618 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1619 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1620 without this. I don't know how relevant this is to other LDAP libraries.
1622 PH/02 Add the verb name to the "unknown ACL verb" error.
1624 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1626 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1628 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1629 directory not expanded when it should be if an expanded home directory
1630 was set for the address (which is overridden by the transport).
1632 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1635 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1636 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1637 because it is too late at that time, and has no effect.
1639 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1640 security issue with \' (bugzilla #107). I could not use the
1641 PQescapeStringConn() function, because it needs a PGconn value as one of
1644 PH/08 When testing addresses using -bt, indicate those final addresses that
1645 are duplicates that would not cause an additional delivery. At least one
1646 person was confused, thinking that -bt output corresponded to deliveries.
1647 (Suppressing duplicates isn't a good idea as you lose the information
1648 about possibly different redirections that led to the duplicates.)
1650 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1651 systems where poll() doesn't work, in particular OS X.
1653 PH/10 Added more information to debugging output for retry time not reached.
1655 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1656 operations in malware.c.
1658 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1661 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1662 syslog_duplication set false, log lines that would normally be written
1663 both the the main log and to the reject log were not written to syslog at
1666 PH/14 In the default configuration, change the use of "message" in ACL warn
1667 statements to "add_header".
1669 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1670 not followed by a command (e.g. "seen endif").
1672 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1673 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1676 PH/17 Added extra conditions to the default value of delay_warning_condition
1680 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1681 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1682 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1685 The Auto-Submitted: and various List- headers are standardised, whereas I
1686 don't think Precedence: ever was.
1688 PH/18 Refactored debugging code in route_finduser() to show more information,
1689 in particular, the error code if getpwnam() issues one.
1691 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1692 This is apparently needed in addition to the PH/07 change above to avoid
1693 any possible encoding problems.
1695 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1696 but not after initializing Perl.
1698 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1699 output them only if debugging. By default they are written stderr,
1700 apparently, which is not desirable.
1702 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1705 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1708 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1710 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1711 authenticated or an ident call has been made. Suppress the default
1712 values for $authenticated_id and $authenticated_sender (but permit -oMai
1713 and -oMas) when testing with -bh.
1715 PH/24 Re-jigged the order of the tests in the default configuration so that the
1716 tests for valid domains and recipients precede the DNS black list and CSA
1717 tests, on the grounds that those ones are more expensive.
1719 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1720 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1721 This bug exists in every version of Exim that I still have, right back to
1724 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1725 However, an attempt to turn on case-sensitivity in a regex key by
1726 including (?-i) didn't work because the subject string was already
1727 lowercased, and the effects were non-intuitive. It turns out that a
1728 one-line patch can be used to allow (?-i) to work as expected.
1734 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1735 other effects) broke the use of negated acl sub-conditions.
1737 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1740 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1741 "Deny" causes Exim to reject the incoming connection with a 554 error.
1742 Unfortunately, if there is a major crisis, such as a disk failure,
1743 tcp-wrappers gives "deny", whereas what one would like would be some
1744 kind of temporary error. A kludge has been added to help with this.
1745 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1746 554 error is used if errno is still zero or contains ENOENT (which occurs
1747 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1750 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1752 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1753 errors. Otherwise a message that provokes a temporary error (when other
1754 messages do not) can cause a whole host to time out.
1756 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1757 addresses were routed directly to files or pipes from a redirect router.
1758 File deliveries just didn't batch; pipe deliveries might have suffered
1761 PH/06 A failure to get a lock for a hints database would erroneously always say
1762 "Failed to get write lock", even when it was really a read lock.
1764 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1765 of 0600. This has been changed to use the value of the lockfile_mode
1766 option (which defaults to 0600).
1768 PH/08 Applied small patch from the Sieve maintainer.
1770 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1771 folder from quota calculations, a direct delivery into this folder messed
1772 up the contents of the maildirsize file. This was because the regex was
1773 used only to exclude .Trash (or whatever) when the size of the mailbox
1774 was calculated. There was no check that a delivery was happening into an
1775 excluded directory. This bug has been fixed by ignoring all quota
1776 processing for deliveries into excluded directories.
1778 PH/10 Added the maildirfolder_create_regex option to appendfile.
1784 PH/01 The code for finding all the local interface addresses on a FreeBSD
1785 system running IPv6 was broken. This may well have applied to all BSD
1786 systems, as well as to others that have similar system calls. The broken
1787 code found IPv4 interfaces correctly, but gave incorrect values for the
1788 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1789 that it would not match correctly against @[] and not recognize the IPv6
1792 PH/02 The ipliteral router was not recognizing addresses of the form user@
1793 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1795 PH/03 Added disable_ipv6.
1797 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1798 decoded form, because it is most often used to construct To: headers
1799 lines in autoreplies, and the decoded form may well be syntactically
1800 invalid. However, $reply_address has leading white space removed, and all
1801 newlines turned into spaces so that the autoreply transport does not
1804 PH/05 If group was specified without a user on a router, and no group or user
1805 was specified on a transport, the group from the router was ignored.
1807 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1808 for visible compile-time settings that can be used to change these
1809 numbers, for those that want even more. Backwards compatibility with old
1810 spool files has been maintained. However, going back to a previous Exim
1811 release will lost any variables that are in spool files.
1813 PH/07 Two small changes when running in the test harness: increase delay when
1814 passing a TCP/IP connection to a new process, in case the original
1815 process has to generate a bounce, and remove special handling of
1816 127.0.0.2 (sic), which is no longer necessary.
1818 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1819 be the same on different OS.
1821 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1824 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1825 whether --show-vars was specified or not
1827 JJ/02 exipick: Added support for new ACL variable spool format introduced
1830 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1831 syntactically invalid From: or Reply-to: line, and a filter used this to
1832 generate an autoreply, and therefore failed to obtain an address for the
1833 autoreply, Exim could try to deliver to a non-existent relative file
1834 name, causing unrelated and misleading errors. What now happens is that
1835 it logs this as a hard delivery error, but does not attempt to create a
1838 PH/11 The exinext utility has a -C option for testing purposes, but although
1839 the given file was scanned by exinext itself; it wasn't being passed on
1840 when Exim was called.
1842 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1843 an end-of-file indication when reading a command response.
1845 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1846 compiled. In many other places in Exim, IPv6 addresses are always
1847 recognized, so I have changed this. It also means that IPv4 domain
1848 literals of the form [IPV4:n.n.n.n] are now always recognized.
1850 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1851 used if the router is not running as root, for example, when verifying at
1852 ACL time, or when using -bh. The debugging output from this situation was
1853 non-existent - all you got was a failure to exec. I have made two
1856 (a) Failures to set uid/gid, the current directory, or a process leader
1857 in a subprocess such as that created by queryprogram now generate
1858 suitable debugging ouput when -d is set.
1860 (b) The queryprogram router detects when it is not running as root,
1861 outputs suitable debugging information if -d is set, and then runs
1862 the subprocess without attempting to change uid/gid.
1864 PH/15 Minor change to Makefile for building test_host (undocumented testing
1867 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1868 additional section of a DNS packet that returns MX or SRV records.
1869 Instead, it always explicitly searches for A/AAAA records. This avoids
1870 major problems that occur when a DNS server includes only records of one
1871 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1872 fixed another bug: if SRV records were looked up and the corresponding
1873 address records were *not* found in the additional section, the port
1874 values from the SRV records were lost.
1876 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1877 using the correct key (the original address) when searching the retry
1878 rules in order to find which one to use for generating the retry hint.
1880 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1881 adding the default one. Similarly, if it contains a Reply-To: header, the
1882 errors_reply_to option, if set, is not used.
1884 PH/19 When calculating a retry time, Exim used to measure the "time since
1885 failure" by looking at the "first failed" field in the retry record. Now
1886 it does not use this if it is later than than the arrival time of the
1887 message. Instead it uses the arrival time. This makes for better
1888 behaviour in cases where some deliveries succeed, thus re-setting the
1889 "first failed" field. An example is a quota failure for a huge message
1890 when small messages continue to be delivered. Without this change, the
1891 "time since failure" will always be short, possible causing more frequent
1892 delivery attempts for the huge message than are intended.
1893 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1895 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1896 $1, $2, $3) because the numerical variables can be reset during some
1897 expansion items (e.g. "match"), thereby losing the authentication data.
1899 PH/21 Make -bV show the size of off_t variables so that the test suite can
1900 decide whether to run tests for quotas > 2G.
1902 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1903 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1904 filecount value is greater than 2G or if a quota value is greater than 2G
1905 on a system where the size of off_t is not greater than 4, a panic error
1908 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1909 never match. The debug and -bh output now contains an explicit error
1910 message indicating a malformed IPv4 address or mask.
1912 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1913 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1914 PH/23 above applies.
1916 PH/25 Do not write to syslog when running in the test harness. The only
1917 occasion when this arises is a failure to open the main or panic logs
1918 (for which there is an explicit test).
1920 PH/26 Added the /no_tell option to "control=freeze".
1922 PH/27 If a host name lookup failed very early in a connection, for example, if
1923 the IP address matched host_lookup and the reverse lookup yielded a name
1924 that did not have a forward lookup, an error message of the form "no IP
1925 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1926 could be logged. Now it outputs the IP address instead of "NULL".
1928 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1929 allows the sender and the authenticated sender to be set when
1930 submitting a message from within Exim. Since child_open_exim() is
1931 documented for local_scan(), the new function should be too.
1933 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1934 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1935 results in an empty string is now treated as unset.
1937 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1939 PH/31 Added sender_verify_fail logging option.
1941 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1942 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1943 tidied the source and removed it altogether.
1945 PH/33 When a queue run was abandoned because the load average was too high, a
1946 log line was always written; now it is written only if the queue_run log
1947 selector is set. In addition, the log line for abandonment now contains
1948 information about the queue run such as the pid. This is always present
1949 in "start" and "stop" lines but was omitted from the "abandon" line.
1951 PH/34 Omit spaces between a header name and the colon in the error message that
1952 is given when verify = headers_syntax fails (if there are lots of them,
1953 the message gets confusing).
1955 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1956 names, as there are now some PTR records that contain slashes. This check
1957 is only to protect against broken name servers that fall over on strange
1958 characters, so the fact that it applies to all lookups doesn't matter.
1960 PH/36 Now that the new test suite is complete, we can remove some of the
1961 special code in Exim that was needed for the old test suite. For example,
1962 sorting DNS records because real resolvers return them in an arbitrary
1963 order. The new test suite's fake resolver always returns records in the
1966 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1967 bounces) except when queue_only is set, to avoid logging races between
1968 the different processes.
1970 PH/38 Panic-die if .include specifies a non-absolute path.
1972 PH/39 A tweak to the "H" retry rule from its user.
1974 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1975 a label. They prevented compilation on older perls.
1977 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1978 a warning to be raised on newish perls.
1980 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1981 on queue. Changes to match documented behaviour of showing count of
1982 messages matching specified criteria.
1984 PH/40 Changed the default ident timeout from 30s to 5s.
1986 PH/41 Added support for the use of login_cap features, on those BSD systems
1987 that have them, for controlling the resources used by pipe deliveries.
1989 PH/42 The content-scanning code uses fopen() to create files in which to put
1990 message data. Previously it was not paying any attention to the mode of
1991 the files. Exim runs with umask(0) because the rest of the code creates
1992 files with open(), and sets the required mode explicitly. Thus, these
1993 files were ending up world-writeable. This was not a big issue, because,
1994 being within the spool directory, they were not world-accessible. I have
1995 created a function called modefopen, which takes an additional mode
1996 argument. It sets umask(777), creates the file, chmods it to the required
1997 mode, then resets the umask. All the relevant calls to fopen() in the
1998 content scanning code have been changed to use this function.
2000 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2001 to 24 hours. This avoids potential overflow problems when processing G
2002 and H retry rules. I suspect nobody ever tinkers with this value.
2004 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2006 PH/45 When the plaintext authenticator is running as a client, the server's
2007 challenges are checked to ensure they are valid base64 strings. By
2008 default, the authentication attempt is cancelled if an invalid string is
2009 received. Setting client_ignore_invalid_base64 true ignores these errors.
2010 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2011 they are received. Thus, the responses can be made to depend on the
2012 challenges. If an invalid string is ignored, an empty string is placed in
2015 PH/46 Messages that are created by the autoreply transport now contains a
2016 References: header, in accordance with RFCs 2822 and 3834.
2018 PH/47 Added authenticated_sender_force to the smtp transport.
2020 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2022 PH/49 Installed latest patch from the Sieve maintainer.
2024 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2025 was also set, the appendfile transport was unnecessarily scanning a
2026 directory of message files (e.g. for maildir delivery) to find the count
2027 of files (along with the size), even though it did not need this
2028 information. It now does the scan only if it needs to find either the
2029 size of the count of files.
2031 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2033 PH/52 Two bugs concerned with error handling when the smtp transport is
2036 (i) Exim was not creating retry information for temporary errors given
2037 for individual recipients after the DATA command when the smtp transport
2038 was used in LMTP mode. This meant that they could be retried too
2039 frequently, and not timed out correctly.
2041 (ii) Exim was setting the flag that allows error details to be returned
2042 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2043 recipients that were returned after the DATA command.
2045 PH/53 This is related to PH/52, but is more general: for any failing address,
2046 when detailed error information was permitted to be returned to the
2047 sender, but the error was temporary, then after the final timeout, only
2048 "retry timeout exceeded" was returned. Now it returns the full error as
2049 well as "retry timeout exceeded".
2051 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2052 do this, and (what is worse) MTAs that accept it.
2054 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2055 will now be deprecated.
2057 PH/56 New os.c-cygwin from the Cygwin maintainer.
2059 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2060 formats (previously only available in exim formats via -bpr, -bpru,
2061 and -bpra. Now also available in native and exiqgrep formats)
2063 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2064 with very large, slow to parse queues
2066 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2068 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2070 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2071 responses to authentication challenges, though it was showing the
2072 challenges; (ii) I've removed the CR characters from the debug output for
2075 PH/58 Allow for the insertion of a newline as well as a space when a string
2076 is turned into more than one encoded-word during RFC 2047 encoding. The
2077 Sieve code now uses this.
2079 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2080 data_4xx, lost_connection, tls_required.
2082 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2083 message was being sent as an SMTP response.
2085 PH/61 Add -l and -k options to exicyclog.
2087 PH/62 When verifying, if an address was redirected to one new address, so that
2088 verification continued, and the new address failed or deferred after
2089 having set something in $address_data, the value of $address_data was not
2090 passed back to the ACL. This was different to the case when no
2091 redirection occurred. The value is now passed back in both cases.
2093 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2094 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2095 use login_cap.h, so on its own it isn't the distinguishing feature. The
2096 new name refers directly to the setclassresources() function.
2098 PH/65 Added configuration files for NetBSD3.
2100 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2102 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2103 is preferred over IPv4.
2105 PH/68 The bounce_return_message and bounce_return_body options were not being
2106 honoured for bounces generated during the reception of non-SMTP messages.
2107 In particular, this applied to messages rejected by the ACL. This bug has
2108 been fixed. However, if bounce_return_message is true and bounce_return_
2109 body is false, the headers that are returned for a non-SMTP message
2110 include only those that have been read before the error was detected.
2111 (In the case of an ACL rejection, they have all been read.)
2113 PH/69 The HTML version of the specification is now built in a directory called
2114 spec_html instead of spec.html, because the latter looks like a path with
2115 a MIME-type, and this confuses some software.
2117 PH/70 Catch two compiler warnings in sieve.c.
2119 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2120 function verify_get_ident() calls ip_connect() to connect a socket, but
2121 if the "connect()" function timed out, ip_connect() used to close the
2122 socket. However, verify_get_ident() also closes the socket later, and in
2123 between Exim writes to the log, which may get opened at this point. When
2124 the socket was closed in ip_connect(), the log could get the same file
2125 descriptor number as the socket. This naturally causes chaos. The fix is
2126 not to close the socket in ip_connect(); the socket should be closed by
2127 the function that creates it. There was only one place in the code where
2128 this was missing, in the iplookup router, which I don't think anybody now
2129 uses, but I've fixed it anyway.
2131 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2132 well as to direct DNS lookups. Otherwise the handling of names in host
2133 lists is inconsistent and therefore confusing.
2139 PH/01 Two changes to the default runtime configuration:
2141 (1) Move the checks for relay_from_hosts and authenticated clients from
2142 after to before the (commented out) DNS black list checks.
2144 (2) Add control=submission to the relay_from_hosts and authenticated
2145 clients checks, on the grounds that messages accepted by these
2146 statements are most likely to be submissions.
2148 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2150 (1) Generate an error if the third argument for the ${prvs expansion is
2153 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2156 (3) Reset the variables that are obtained from the first argument of
2157 ${prvscheck and used in the second argument before leaving the code,
2158 because their memory is reclaimed, so using them afterwards may do
2161 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2162 one (it's much easier than Tom thought :-).
2164 (5) Because of (4), we can now allow for the use of $prvscheck_result
2165 inside the third argument.
2167 PH/03 For some reason, the default setting of PATH when running a command from
2168 a pipe transport was just "/usr/bin". I have changed it to
2171 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2172 anything to be listed in the output from -bV.
2174 PH/05 When a filter generated an autoreply, the entire To: header line was
2175 quoted in the delivery log line, like this:
2177 => >A.N.Other <ano@some.domain> <original@ddress> ...
2179 This has been changed so that it extracts the operative address. There
2180 may be more than one such address. If so, they are comma-separated, like
2183 => >ano@some.domain,ona@other.domain <original@ddress> ...
2185 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2186 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2187 not being looked up in the rDNS to get a host name, Exim was showing the
2188 IP address twice in Received: lines, even though the IP addresses were
2189 identical. For example:
2191 Received: from [1.2.3.4] (helo=[1.2.3.4])
2193 However, if the real host name was known, it was omitting the HELO data
2194 if it matched the actual IP address. This has been tidied up so that it
2195 doesn't show the same IP address twice.
2197 PH/07 When both +timestamp and +memory debugging was on, the value given by
2198 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2199 called by the debug printing, thereby overwriting the timestamp buffer.
2200 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2202 PH/08 When the original message was included in an autoreply transport, it
2203 always said "this is a copy of the message, including all the headers",
2204 even if body_only or headers_only was set. It now gives an appropriate
2207 PH/09 Applied a patch from the Sieve maintainer which:
2209 o fixes some comments
2210 o adds the (disabled) notify extension core
2211 o adds some debug output for the result of if/elsif tests
2212 o points to the current vacation draft in the documentation
2213 and documents the missing references header update
2217 o fixes a bug in processing the envelope test (when testing
2218 multiple envelope elements, the last element determinted the
2221 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2222 Electronic Mail") by including:
2224 Auto-submitted: auto-generated
2226 in the messages that it generates (bounce messages and others, such as
2227 warnings). In the case of bounce messages for non-SMTP mesages, there was
2228 also a typo: it was using "Auto_submitted" (underscore instead of
2229 hyphen). Since every message generated by Exim is necessarily in response
2230 to another message, thes have all been changed to:
2232 Auto-Submitted: auto-replied
2234 in accordance with these statements in the RFC:
2236 The auto-replied keyword:
2238 - SHOULD be used on messages sent in direct response to another
2239 message by an automatic process,
2241 - MUST NOT be used on manually-generated messages,
2243 - MAY be used on Delivery Status Notifications (DSNs) and Message
2244 Disposition Notifications (MDNs),
2246 - MUST NOT be used on messages generated by automatic or periodic
2247 processes, except for messages which are automatic responses to
2250 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2251 to the default Received: header definition.
2253 PH/12 Added log selector acl_warn_skipped (default on).
2255 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2256 variables because (a) they are in the wrong storage pool and (b) even if
2257 they were copied, it wouldn't work properly because of the caching.
2259 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2260 checking when decoding. Apparently there are clients that generate
2261 overlong encoded strings. Why am I not surprised?
2263 PH/15 If the first argument of "${if match_address" was not empty, but did not
2264 contain an "@" character, Exim crashed. Now it writes a panic log message
2265 and treats the condition as false.
2267 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2269 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2270 extension "envelope-auth". The code is finished and in agreement with
2271 other implementations, but there is no documentation so far and in fact,
2272 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2273 not changing the active code.
2275 Print executed "if" and "elsif" statements when debugging is used. This
2276 helps a great deal to understand what a filter does.
2278 Document more things not specified clearly in RFC3028. I had all this
2279 sorted out, when out of a sudden new issues came to my mind. Oops."
2281 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2284 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2285 canonical form (as documented). However, after a host name lookup from
2286 the IP address, check_host() was doing a simple string comparison with
2287 addresses acquired from the DNS when checking that the found name did
2288 have the original IP as one of its addresses. Since any found IPv6
2289 addresses are likely to be in abbreviated form, the comparison could
2290 fail. Luckily, there already exists a function for doing the comparison
2291 by converting both addresses to binary, so now that is used instead of
2292 the text comparison.
2294 PH/20 There was another similar case to PH/19, when a complete host name was
2295 given in a host list; looking up its IP address could give an abbreviated
2296 form, whereas the current host's name might or might not be abbreviated.
2297 The same fix has been applied.
2303 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2304 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2307 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2308 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2310 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2312 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2313 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2314 Also, for "Auto-Submitted", treat anything other than "no" as
2315 non-personal, in accordance with RFC 3834. (Previously it treated
2316 anything starting "auto-" as non-personal.)
2318 TF/01 The control=submission/name=... option had a problem with syntax
2319 errors if the name included a slash character. The /name= option
2320 now slurps the rest of the string, so it can include any characters
2321 but it must come last in the list of options (after /sender_retain
2324 PH/05 Some modifications to the interface to the fake nameserver for the new
2332 TK/01 Added the "success_on_redirect" address verification option. See
2333 NewStuff for rationale and an example.
2335 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2337 PH/02 Patch to exigrep to allow it to work on syslog lines.
2339 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2340 fread() to skip over the body file's header line, because in Cygwin the
2341 header line is locked and is inaccessible.
2343 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2344 co-exist for some time) to make it clear that it is the Exim ID that is
2345 referenced, not the Message-ID: header line.
2347 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2348 string_format() function, because snprintf() does not exist on all
2351 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2352 ${stat: expansion item.
2354 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2355 protocol synchronization error", to keep the pedants happy.
2357 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2358 well as for IRIX systems, when gcc is being used. See the host.c source
2361 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2363 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2366 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2367 transport and to the smtp transport in LMTP mode.
2369 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2371 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2372 run a filter in a subprocess. This could lead to confusion in subsequent
2373 lookups in the parent process. There should also be a search_tidyup() at
2374 the end of the subprocess.
2376 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2377 only if the host matched helo_try_verify_hosts, which caused the
2378 verification to occur when the EHLO/HELO command was issued. The ACL just
2379 tested the remembered result. Now, if a previous verification attempt has
2380 not happened, "verify = helo" does it there and then.
2382 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2384 TK/03 Fix log output including CR from clamd.
2386 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2387 exist provoked a memory error which could cause a segfault.
2389 PH/15 Installed PCRE 6.2
2391 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2393 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2394 of the problem. Specifically, suggested +O2 rather than +O1 for the
2397 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2399 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2400 record, the port was not correctly logged when the outgoing_port log
2401 selector was set (it logged the transort's default port).
2403 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2404 fallback_hosts, and "hosts" in the smtp transport.
2406 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2407 host errors such as "Connection refused".
2409 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2410 authentication with radiusclient 0.4.9:
2412 - Error returned from rc_read_config was caught wrongly
2413 - Username/password not passed on to radius server due to wrong length.
2415 The presumption is that some radiusclient API changes for 4.51/PH/17
2416 were not taken care of correctly. The code is still untested by me (my
2417 Linux distribution still has 0.3.2 of radiusclient), but it was
2418 contributed by a Radius user.
2420 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2421 expanding the "port" option of the smtp transport.
2423 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2424 while reading a MIME header. Thanks to Tom Hughes for a patch.
2426 PH/24 Include config.h inside local_scan.h so that configuration settings are
2429 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2430 that in an ACL for RCPT (for example), you can examine exactly what was
2433 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2434 commands, but it was not correctly comparing the address with the actual
2435 client host address. Thus, it would show the EHLO address in Received:
2436 header lines when this was not necessary.
2438 PH/27 Added the % operator to ${eval:}.
2440 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2441 it should be ignoring failures (because with -C, for example, it has lost
2442 privilege). It wasn't ignoring creation failures other than "already
2445 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2448 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2449 because an input line was too long, either on its own, or by virtue of
2450 too many continuations, the temporary file was not being removed, and the
2451 return code was incorrect.
2453 PH/31 Missing "BOOL" in function definition in filtertest.c.
2455 PH/32 Applied Sieve patches from the maintainer.
2457 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2459 PH/33 Added "verify = not_blind".
2461 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2462 Local/Makefile (with some defaults set). These are used in built scripts
2463 such as exicyclog, but they have never been used in the exim_install
2464 script (though there are many overriding facilities there). I have
2465 arranged that the exim_install script now takes note of these two
2468 PH/35 Installed configuration files for Dragonfly.
2470 PH/36 When a locally submitted message by a trusted user did not contain a
2471 From: header, and the sender address was obtained from -f or from an SMTP
2472 MAIL command, and the trusted user did not use -F to supply a sender
2473 name, $originator_name was incorrectly used when constructing a From:
2474 header. Furthermore, $originator_name was used for submission mode
2475 messages from external hosts without From: headers in a similar way,
2476 which is clearly wrong.
2478 PH/37 Added control=suppress_local_fixups.
2480 PH/38 When log_selector = +received_sender was set, and the addition of the
2481 sender made the log line's construction buffer exactly full, or one byte
2482 less than full, an overflow happened when the terminating "\n" was
2485 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2486 when the result of a list match is failure because a DNS lookup failed.
2488 PH/40 RM_COMMAND is now used in the building process.
2490 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2491 the "build-* directories that it finds.
2493 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2494 address in a domain literal was a prefix of an interface address.
2496 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2497 when verifying a sender address, unless rewrite_headers is false.
2499 PH/44 Wrote a long comment about why errors_to addresses are verified as
2500 recipients, not senders.
2502 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2503 the ratelimit ACL was added.
2505 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2507 PH/46 Added extra information about PostgreSQL errors to the error string.
2509 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2510 avoiding the need to install special zones in a real server. This is
2511 backwards compatible; if it can't find the fake resolver, it drops back.
2512 Thus, both old and new test suites can be run.
2514 TF/02 Added util/ratelimit.pl
2516 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2517 clock is set back in time.
2519 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2520 Candler <B.Candler@pobox.com>.
2522 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2523 OK for addresses that are the result of redirections.
2525 PH/48 A number of further additions for the benefit of the new test suite,
2526 including a fake gethostbyname() that interfaces to the fake DNS resolver
2529 TF/06 The fix for widen_domains has also been applied to qualify_single and
2530 search_parents which are the other dnslookup options that can cause
2533 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2536 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2538 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2539 service" if it encountered an MX record with an empty target hostname.
2540 The message is now "an MX or SRV record indicated no SMTP service".
2542 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2543 if the DNS of the sending site is misconfigured. This is quite a
2544 common situation. This change restores the behaviour of treating a
2545 helo verification defer as a failure.
2547 PH/51 If self=fail was set on a router, the bounce message did not include the
2548 actual error message.
2554 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2556 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2557 command itself ended in error, the underlying message about the transport
2558 filter timeout was being overwritten with the pipe command error. Now the
2559 underlying error message should be appended to the second error message.
2561 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2563 PH/02 Reduce the amount of output that "make" produces by default. Full output
2564 can still be requested.
2566 PH/03 The warning log line about a condition test deferring for a "warn" verb
2567 was being output only once per connection, rather than after each
2568 occurrence (because it was using the same function as for successful
2569 "warn" verbs). This seems wrong, so I have changed it.
2571 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2572 it should not have, which might have caused a crash in the right
2573 circumstances, but probably never did.
2575 PH/04 Installed a modified version of Tony Finch's patch to make submission
2576 mode fix the return path as well as the Sender: header line, and to
2577 add a /name= option so that you can make the user's friendly name appear
2580 TF/03 Added the control = fakedefer ACL modifier.
2582 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2583 Mark Lowes for thorough testing.
2585 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2587 TK/03 Merged latest SRS patch from Miles Wilton.
2589 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2590 with the definition in sysexits.h (which is #included earlier).
2591 Fortunately, Exim does not actually use EX_OK. The code used to try to
2592 preserve the sysexits.h value, by assumimg that macro definitions were
2593 scanned for macro replacements. I have been disabused of this notion,
2594 so now the code just undefines EX_OK before #including unistd.h.
2596 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2597 in the smtp transport. When a block could not be written in a single
2598 write() function, the timeout was being re-applied to each part-write.
2599 This seems wrong - if the receiver was accepting one byte at a time it
2600 would take for ever. The timeout is now adjusted when this happens. It
2601 doesn't have to be particularly precise.
2603 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2604 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2606 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2607 without a domain if the check to <postmaster@domain> fails.
2609 SC/01 Eximstats: added -xls and the ability to specify output files
2610 (patch written by Frank Heydlauf).
2612 SC/02 Eximstats: use FileHandles for outputing results.
2614 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2616 SC/04 Eximstats: fixed display of large numbers with -nvr option
2618 SC/05 Eximstats: fixed merging of reports with empty tables.
2620 SC/06 Eximstats: added the -include_original_destination flag
2622 SC/07 Eximstats: removed tabs and trailing whitespace.
2624 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2626 TK/06 MBOX spool code: Add real "From " MBOX separator line
2627 so the .eml file is really in mbox format (even though
2628 most programs do not really care). Patch from Alex Miller.
2630 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2631 The latter is generated from $received_to and is only set if the
2632 message has one envelope recipient. SA can use these headers,
2633 obviously out-of-the-box. Patch from Alex Miller.
2635 PH/08 The ${def test on a variable was returning false if the variable's
2636 value was "0", contrary to what the specification has always said!
2637 The result should be true unless the variable is empty.
2639 PH/09 The syntax error of a character other than { following "${if
2640 def:variable_name" (after optional whitespace) was not being diagnosed.
2641 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2642 accidental colon was present, for example, could give incorrect results.
2644 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2645 result is used (not including appendfile, where other changes are about
2648 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2649 This involved changing a lot of size variables from int to off_t. It
2650 should work with maildirs and everything.
2652 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2653 spamd dying while we are connected to it.
2655 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2658 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2659 function for BDB 4.3.
2661 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2663 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2664 the sources for PCRE have been re-arranged and more files are now
2667 PH/15 The code I had for printing potentially long long variables in PH/11
2668 above was not the best (it lost precision). The length of off_t variables
2669 is now inspected at build time, and an appropriate printing format (%ld
2670 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2671 to be "long long int" or "long int". This is needed for the internal
2672 formatting function string_vformat().
2674 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2675 the configuration file to be ":syslog", then the script "guesses" where
2676 the logs files are, rather than using the compiled in default. In our
2677 case the guess is not the same as the compiled default, so the script
2678 suddenly stopped working when I started to use syslog. The patch checks
2679 to see if log_file_path is "". If so, it attempts to read it from exim
2680 with no configuration file to get the compiled in version, before it
2681 falls back to the previous guessing code."
2683 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2684 implementing BATV in an Exim configuration. See NewStuff for the gory
2687 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2688 Makefile that are specific to HP-UX.
2690 PH/18 If the "use_postmaster" option was set for a recipient callout together
2691 with the "random" option, the postmaster address was used as the MAIL
2692 FROM address for the random test, but not for the subsequent recipient
2693 test. It is now used for both.
2695 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2696 patch removes a few documentation additions to RFC 3028, because the
2697 latest draft now contains them. It adds the new en;ascii-case comparator
2698 and a new error check for 8bit text in MIME parts. Comparator and
2699 require names are now matched exactly. I enabled the subaddress
2700 extension, but it is not well tested yet (read: it works for me)."
2702 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2703 rework some of the code of TK/09 above to avoid the hardwired use of
2704 "%lld" and "long long". Replaced the call to snprintf() with a call to
2707 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2708 records point to non-existent hosts", "retry timeout exceeded", and
2709 "retry time not reached for any host after a long failure period".
2711 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2712 experimental DomainKeys support:
2714 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2715 (2) On an error such as an illegally used "control", the wrong name for
2716 the control was given.
2718 These problems did NOT occur unless DomainKeys support was compiled.
2720 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2722 PH/24 Added ${if match_ip condition.
2724 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2725 databases so that it will be absolutely obvious if a crash occurs in the
2726 DB library. This is a regular occurrence (often caused by mis-matched
2729 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2730 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2731 picked up on a user's system that detects such things. There doesn't seem
2732 to be a gcc warning option for this - only an attribute that has to be
2733 put on the function's prototype. It seems that in Fedora Core 4 they have
2734 set this on a number of new functions. No doubt there will be more in due
2737 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2738 specify a transport. However, if an address that was verified by such a
2739 router was the subject of a callout, Exim crashed because it tried to
2740 read the rcpt_include_affixes from the non-existent transport. Now it
2741 just assumes that the setting of that option is false. This bug was
2742 introduced by 4.51/PH/31.
2744 PH/28 Changed -d+all to exclude +memory, because that information is very
2745 rarely of interest, but it makes the output a lot bigger. People tend to
2746 do -d+all out of habit.
2748 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2749 code in os-type was giving problems when libc.so lives in lib64, like on
2752 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2753 aren't the modern standard, and it seems that some systems' include files
2754 don't always have them. Exim was already checking for some of the newer
2755 ones like T_AAAA, and defining it itself. I've added checks for all the
2756 record types that Exim uses.
2758 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2759 not automatically generating a new one, as it is supposed to. This
2760 prevented TLS from working. If the file did exist, but contained invalid
2761 data, a new version was generated, as expected. It was only the case of a
2762 non-existent file that was broken.
2764 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2765 with a change in libdomainkeys > 0.64.
2767 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2768 from DNS. If the selector record carries the flag, it now has
2769 precedence over the domain-wide flag.
2771 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2773 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2774 the use of a port name that isn't defined in /etc/services) occurred, the
2775 message was deferred as in a normal delivery, and thus remained on the
2776 spool, instead of being failed because of the mua_wrapper setting. This
2777 is now fixed, and I tidied up some of the mua_wrapper messages at the
2780 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2781 the messages in a hash of arrays rather than using individual hashes.
2782 This is a bit cleaner and results in dramatic memory savings, albeit
2783 at a slight CPU cost.
2785 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2786 as requested by Marc Sherman.
2788 SC/10 Eximstats: added histograms for user specified patterns as requested
2791 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2793 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2794 fopen() in the content-scanning modules that did not already have it.
2800 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2801 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2803 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2805 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2807 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2808 to test Sieve filters that use "vacation".
2810 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2811 that changes the way the GnuTLS parameters are stored in the cache file.
2812 The new format can be generated externally. For backward compatibility,
2813 if the data in the cache doesn't make sense, Exim assumes it has read an
2814 old-format file, and it generates new data and writes a new file. This
2815 means that you can't go back to an older release without removing the
2818 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2819 work if there are any delivery delays because "one_time" forces the
2820 parent to be marked "delivered", so its unseen clone is never tried
2821 again. For this reason, Exim now forbids the simultaneous setting of
2824 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2825 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2826 entry if you want to know the details. The fix, however, neglected to
2827 consider the case when local delivery batching is involved. The test for
2828 "previously delivered" was not happening when checking to see if an
2829 address could be batched with a previous (undelivered) one; under
2830 certain circumstances this could lead to multiple deliveries to the same
2833 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2834 in its include files, and this causes problems building Exim.
2836 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2837 header_syntax) but Exim was just ignoring anything given after a slash.
2838 In particular, this caused confusion with an attempt to use "verify =
2839 reverse_host_lookup/defer_ok". An error is now given when options are
2840 supplied for verify items that do not have them. (Maybe reverse_host_
2841 lookup should have a defer_ok option, but that's a different point.)
2843 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2844 defined by RFC 821) to 2048, because there were problems with some AUTH
2845 commands, and RFC 1869 says the size should be increased for extended
2846 SMTP commands that take arguments.
2848 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2851 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2852 "unknown" error; now it says that the functionality isn't in the binary.
2854 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2855 an address' error message when a string expansion fails (syntax or
2856 whatever). Otherwise the password may appear in the log. Following change
2857 PH/42 below, there is no longer a chance of it appearing in a bounce
2860 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2862 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2863 of its addresses was ever tried. (Bugzilla bug #2.)
2865 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2866 the result incorrectly in the debug output. (It correctly added a newline
2867 to what was transported.)
2869 TF/01 Added $received_time.
2871 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2872 commented out examples of how to interface to a virus scanner and to
2873 SpamAssassin. Also added commented examples of av_scanner and
2874 spamd_address settings.
2876 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2877 and controls are allowed in which ACLs. There were a couple of minor
2878 errors. Some of the entries in the conditions table (which is a table of
2879 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2880 negation of where the condition IS allowed.
2882 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2884 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2885 header file does not have a version number, so I've had to invent a new
2886 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2887 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2888 radiusclient), but it was contributed by a Radius user.
2890 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2891 files or directories, for OpenSSL.
2893 PH/19 When an Exim process that is running as root has to create an Exim log
2894 file, it does so in a subprocess that runs as exim:exim so as to get the
2895 ownership right at creation (otherwise, other Exim processes might see
2896 the file with the wrong ownership). There was no test for failure of this
2897 fork() call, which would lead to the process getting stuck as it waited
2898 for a non-existent subprocess. Forks do occasionally fail when resources
2899 run out. I reviewed all the other calls to fork(); they all seem to check
2902 PH/20 When checking for unexpected SMTP input at connect time (before writing
2903 the banner), Exim was not dealing correctly with a non-positive return
2904 from the read() function. If the client had disconnected by this time,
2905 the result was a log entry for a synchronization error with an empty
2906 string after "input=" when read() returned zero. If read() returned -1
2907 (an event I could not check), uninitialized data bytes were printed.
2908 There were reports of junk text (parts of files, etc) appearing after
2911 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2913 PH/22 Added support for macro redefinition, and (re)definition in between
2914 driver and ACL definitions.
2916 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2917 forgetting to use the resulting value; it was using the unexpanded value.
2919 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2920 hadn't been configured. The fix is from Juergen Kreileder, who
2921 understands it better than I do:
2923 "Here's what I see happening with three configured cyrus_sasl
2924 authenticators configured (plain, login, cram-md5):
2926 On startup auth_cyrus_sasl_init() gets called for each of these.
2927 This means three calls to sasl_listmech() without a specified mech_list.
2928 => SASL tests which mechs of all available mechs actually work
2929 => three warnings about OTP not working
2930 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2932 With the patch, sasl_listmech() also gets called three times. But now
2933 SASL's mech_list option is set to the server_mech specified in the the
2934 authenticator. Or in other words, the answer from sasl_listmech()
2935 gets limited to just the mech you're testing for (which is different
2937 => the return list contains just 'plain' or 'login', 'cram-md5' or
2938 nothing depending on the value of ob->server_mech.
2940 I've just tested the patch: Authentication still works fine,
2941 unavailable mechs specified in the exim configuration are still
2942 caught, and the auth.log warnings about OTP are gone."
2944 PH/25 When debugging is enabled, the contents of the command line are added
2945 to the debugging output, even when log_selector=+arguments is not
2948 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2949 answer is "GNU", and only if the return is "GNU/something" is the answer
2952 PH/27 $acl_verify_message is now set immediately after the failure of a
2953 verification in an ACL, and so is available in subsequent modifiers. In
2954 particular, the message can be preserved by coding like this:
2956 warn !verify = sender
2957 set acl_m0 = $acl_verify_message
2959 Previously, $acl_verify_message was set only while expanding "message"
2960 and "log_message" when a very denied access.
2962 PH/28 Modified OS/os.c-Linux with
2964 -#ifndef OS_LOAD_AVERAGE
2965 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2967 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2968 nomenclature these days.)
2970 PH/29 Installed patch from the Sieve maintainer that adds the options
2971 sieve_useraddress and sieve_subaddress to the redirect router.
2973 PH/30 In these circumstances:
2974 . Two addresses routed to the same list of hosts;
2975 . First host does not offer TLS;
2976 . First host accepts first address;
2977 . First host gives temporary error to second address;
2978 . Second host offers TLS and a TLS session is established;
2979 . Second host accepts second address.
2980 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2981 and peerdn, if requested) that were in fact used only for the second
2984 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2985 attention to any local part prefix or suffix that was matched by the
2986 router that accepted the address. It now behaves in the same way as it
2987 does for delivery: the affixes are removed from the local part unless
2988 rcpt_include_affixes is set on the transport.
2990 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2991 timeout during the DATA phase of an incoming message.
2993 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2994 applied a patch sanctioned by the Sieve maintainer.
2996 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2997 the uid or gid is negative. A case of a negative gid caused this to be
2998 noticed. The fix allows for either to be negative.
3000 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3001 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3004 PH/36 The change PH/12 above was broken. Fixed it.
3006 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3007 the grounds that routing the same address twice would always produce the
3008 same answer. This might have been true once, but it is certainly no
3009 longer true now. Routing a child address may depend on the previous
3010 routing that produced that child. Some complicated redirection strategies
3011 went wrong when messages had multiple recipients, and made Exim's
3012 behaviour dependent on the order in which the addresses were given.
3014 I have moved the duplicate checking until after the routing is complete.
3015 Exim scans the addresses that are assigned to local and remote
3016 transports, and removes any duplicates. This means that more work will be
3017 done, as duplicates will always all be routed, but duplicates are
3018 presumably rare, so I don't expect this is of any significance.
3020 For deliveries to pipes, files, and autoreplies, the duplicate checking
3021 still happens during the routing process, since they are not going to be
3024 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3025 It corrects a timeout issue with spamd. This is Ian's comment: "The
3026 background is that sometimes spamd either never reads data from a
3027 connection it has accepted, or it never writes response data. The exiscan
3028 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3029 blindly assumes that writes won't block so it may never time out."
3031 PH/39 Allow G after quota size as well as K and M.
3033 PH/40 The value set for $authenticated_id in an authenticator may not contain
3034 binary zeroes or newlines because the value is written to log lines and
3035 to spool files. There was no check on this. Now the value is run through
3036 the string_printing() function so that such characters are converted to
3037 printable escape sequences.
3039 PH/41 $message_linecount is a new variable that contains the total number of
3040 lines in the message. Compare $body_linecount, which is the count for the
3043 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3044 bounce and delay warning messages, except in certain special cases, which
3047 (a) An SMTP error message from a remote host;
3048 (b) A message specified in a :fail: redirection;
3049 (c) A message specified in a "fail" command in a system filter;
3050 (d) A message specified in a FAIL return from the queryprogram router;
3051 (e) A message specified by the cannot_route_message router option.
3053 In these cases only, Exim does include the error details in bounce and
3054 warning messages. There are also a few cases where bland messages such
3055 as "unrouteable address" or "local delivery error" are given.
3057 PH/43 $value is now also set for the "else" part of a ${run expansion.
3059 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3060 being worked on, but at least Exim now implements the latest version to
3063 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3064 process to complete was treated as a delivery failure, a timeout while
3065 writing the message to the pipe was logged, but erroneously treated as a
3066 successful delivery. Such timeouts include transport filter timeouts. For
3067 consistency with the overall process timeout, these timeouts are now
3068 treated as errors, giving rise to delivery failures by default. However,
3069 there is now a new Boolean option for the pipe transport called
3070 timeout_defer, which, if set TRUE, converts the failures into defers for
3071 both kinds of timeout. A transport filter timeout is now identified in
3074 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3075 systems where "make" and "gmake" are different, calling "gmake" at top
3076 level broke things. I've arranged for the value of $(MAKE) to be passed
3077 from the Makefile to this script so that it can call the same version of
3081 A note about Exim versions 4.44 and 4.50
3082 ----------------------------------------
3084 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3085 changes of various kinds. As a consequence, a big documentation update was
3086 needed. This delayed the release for rather longer than seemed good, especially
3087 in the light of a couple of (minor) security issues. Therefore, the changes
3088 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3089 release. So 4.44 and 4.50 are in effect two different branches that both start
3092 I have left the 4.50 change log unchanged; it contains all the changes since
3093 4.43. The change log for 4.44 is below; many of its items are identical to
3094 those for 4.50. This seems to be the most sensible way to preserve the
3095 historical information.
3101 1. Minor wording change to the doc/README.SIEVE file.
3103 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3104 computation of the current number of files was incorrect.
3106 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3107 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3110 4. Give more explanation in the error message when the command for a transport
3111 filter fails to execute.
3113 5. There are several places where Exim runs a non-Exim command in a
3114 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3115 was being done only for the command run by the queryprogram router. It is
3116 now done for all such subprocesses. The other cases are: ${run, transport
3117 filters, and the commands run by the lmtp and pipe transports.
3119 6. Added CONFIGURE_GROUP build-time option.
3121 7. Some older OS have a limit of 256 on the maximum number of file
3122 descriptors. Exim was using setrlimit() to set 1000 as a large value
3123 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3124 systems. I've change it so that if it can't get 1000, it tries for 256.
3126 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3127 was an oversight, and furthermore, ever since the addition of extra
3128 controls (e.g. 4.43/32), the checks on when to allow different forms of
3129 "control" were broken. There should now be diagnostics for all cases when a
3130 control that does not make sense is encountered.
3132 9. Added the /retain_sender option to "control=submission".
3134 10. $recipients is now available in the predata ACL (oversight).
3136 11. Tidy the search cache before the fork to do a delivery from a message
3137 received from the command line. Otherwise the child will trigger a lookup
3138 failure and thereby defer the delivery if it tries to use (for example) a
3139 cached ldap connection that the parent has called unbind on.
3141 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3142 of $address_data from the recipient verification was clobbered by the
3143 sender verification.
3145 13. The value of address_data from a sender verification is now available in
3146 $sender_address_data in subsequent conditions in the ACL statement.
3148 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3150 15. Added a new option "connect=<time>" to callout options, to set a different
3153 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3154 was its contents. (It was OK if the option was not defined at all.)
3156 17. A "Completed" log line is now written for messages that are removed from
3157 the spool by the -Mrm option.
3159 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3160 information about exactly what failed.
3162 19. Added -dd to debug only the daemon process.
3164 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3165 handles timeouts, both on the server side and network timeouts. Renamed the
3166 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3168 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3169 It is now set to "smtps".
3171 22. $host_address is now set to the target address during the checking of
3172 ignore_target_hosts.
3174 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3175 being passed; this would have caused $sender_host_name to have been used if
3176 matching the list had actually called for a host name (not very likely,
3177 since this list is usually IP addresses). A host name is now passed as
3180 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3181 code that specifies a non-restarting handler (typically sigaction() in
3182 modern systems) in an attempt to fix a rare and obscure crash bug.
3184 25. Narrowed the window for a race in the daemon that could cause it to ignore
3185 SIGCHLD signals. This is not a major problem, because they are used only to
3186 wake it up if nothing else does.
3188 26. A malformed maildirsize file could cause Exim to calculate negative values
3189 for the mailbox size or file count. Odd effects could occur as a result.
3190 The maildirsize information is now recalculated if the size or filecount
3193 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3194 support for a long time. Removed HAVE_SYS_VFS_H.
3196 28. Installed the latest version of exipick from John Jetmore.
3198 29. In an address list, if the pattern was not a regular expression, an empty
3199 subject address (from a bounce message) matched only if the pattern was an
3200 empty string. Non-empty patterns were not even tested. This was the wrong
3201 because it is perfectly reasonable to use an empty address as part of a
3202 database query. An empty address is now tested by patterns that are
3203 lookups. However, all the other forms of pattern expect the subject to
3204 contain a local part and a domain, and therefore, for them, an empty
3205 address still always fails if the pattern is not itself empty.
3207 30. Exim went into a mad DNS loop when attempting to do a callout where the
3208 host was specified on an smtp transport, and looking it up yielded more
3209 than one IP address.
3211 31. Re-factored the code for checking spool and log partition space into a
3212 function that finds that data and another that does the check. The former
3213 is then used to implement four new variables: $spool_space, $log_space,
3214 $spool_inodes, and $log_inodes.
3216 32. The RFC2047 encoding function was originally intended for short strings
3217 such as real names; it was not keeping to the 75-character limit for
3218 encoded words that the RFC imposes. It now respects the limit, and
3219 generates multiple encoded words if necessary. To be on the safe side, I
3220 have increased the buffer size for the ${rfc2047: expansion operator from
3223 33. It is now permitted to omit both strings after an "if" condition; if the
3224 condition is true, the result is "true". As before, when the second string
3225 is omitted, a false condition yields an empty string. This makes it less
3226 cumbersome to write custom ACL and router conditions.
3228 34. Failure to deliver a bounce message always caused it to be frozen, even if
3229 there was an errors_to setting on the router. The errors_to setting is now
3232 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3233 canonical form (fully expanded) before being placed in
3234 $sender_host_address.
3236 36. The table in the code that translates DNS record types into text (T_A to
3237 "A" for instance) was missing entries for NS and CNAME. It is just possible
3238 that this could have caused confusion if both these types were looked up
3239 for the same domain, because the text type is used as part of Exim's
3240 per-process caching. But the chance of anyone hitting this buglet seems
3243 37. The dnsdb lookup has been extended in a number of ways.
3245 (1) There is a new type, "zns", which walks up the domain tree until it
3246 finds some nameserver records. It should be used with care.
3248 (2) There is a new type, "mxh", which is like "mx" except that it returns
3249 just the host names, not the priorities.
3251 (3) It is now possible to give a list of domains (or IP addresses) to be
3252 looked up. The behaviour when one of the lookups defers can be
3253 controlled by a keyword.
3255 (4) It is now possible to specify the separator character for use when
3256 multiple records are returned.
3258 38. The dnslists ACL condition has been extended: it is now possible to supply
3259 a list of IP addresses and/or domains to be looked up in a particular DNS
3262 39. Added log_selector=+queue_time_overall.
3264 40. When running the queue in the test harness, wait just a tad after forking a
3265 delivery process, to get repeatability of debugging output.
3267 41. Include certificate and key file names in error message when GnuTLS fails
3268 to set them up, because the GnuTLS error message doesn't include the name
3269 of the failing file when there is a problem reading it.
3271 42. Allow both -bf and -bF in the same test run.
3273 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3275 44. The "Exiscan patch" is now merged into the mainline Exim source.
3277 45. Sometimes the final signoff response after QUIT could fail to get
3278 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3279 < 0 before doing a fflush(). This bug looks as though it goes back to the
3280 introduction of TLS in release 3.20, but "sometimes" must have been rare
3281 because the tests only now provoked it.
3283 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3284 (this can affect the format of dates).
3286 47. exim_tidydb, when checking for the continued existence of a message for
3287 which it has found a message-specific retry record, was not finding
3288 messages that were in split spool directories. Consequently, it was
3289 deleting retry records that should have stayed in existence.
3291 48. Steve fixed some bugs in eximstats.
3293 49. The SPA authentication driver was not abandoning authentication and moving
3294 on to the next authenticator when an expansion was forced to fail,
3295 contradicting the general specification for all authenticators. Instead it
3296 was generating a temporary error. It now behaves as specified.
3298 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3299 (the order specifies the preference for clients). The order is now AES256,
3300 AES128, 3DES, ARCFOUR128.
3302 51. Small patch to Sieve code - explicitly set From: when generating an
3305 52. Exim crashed if a remote delivery caused a very long error message to be
3306 recorded - for instance if somebody sent an entire SpamAssassin report back
3307 as a large number of 550 error lines. This bug was coincidentally fixed by
3308 increasing the size of one of Exim's internal buffers (big_buffer) that
3309 happened as part of the Exiscan merge. However, to be on the safe side, I
3310 have made the code more robust (and fixed the comments that describe what
3313 53. Now that there can be additional text after "Completed" in log lines (if
3314 the queue_time_overall log selector is set), a one-byte patch to exigrep
3315 was needed to allow it to recognize "Completed" as not the last thing in
3318 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3319 patch that reportedly fixes this has been added. I am not expert enough to
3320 create a test for it. This is what the patch creator wrote:
3322 "I found a little strange behaviour of ldap code when working with
3323 Windows 2003 AD Domain, where users was placed in more than one
3324 Organization Units. When I tried to give exim partial DN, the exit code
3325 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3326 But simultaneously result of request was absolutely normal ldap result,
3327 so I produce this patch..."
3329 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3330 REFERENCE, so I have modified the code to exclude the patch when that macro
3333 55. Some experimental protocols are using DNS PTR records for new purposes. The
3334 keys for these records are domain names, not reversed IP addresses. The
3335 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3336 leaves it alone. Component reversal etc. now happens only for IP addresses.
3339 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3341 57. Double the size of the debug message buffer (to 2048) so that more of very
3342 long debug lines gets shown.
3344 58. The exicyclog utility now does better if the number of log files to keep
3345 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3347 59. Two changes related to the smtp_active_hostname option:
3349 (1) $smtp_active_hostname is now available as a variable.
3350 (2) The default for smtp_banner uses $smtp_active_hostname instead
3351 of $primary_hostname.
3353 60. The host_aton() function is supposed to be passed a string that is known
3354 to be a valid IP address. However, in the case of IPv6 addresses, it was
3355 not checking this. This is a hostage to fortune. Exim now panics and dies
3356 if the condition is not met. A case was found where this could be provoked
3357 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3358 components; fortuitously, this particular loophole had already been fixed
3359 by change 4.50/55 above.
3361 If there are any other similar loopholes, the new check in host_aton()
3362 itself should stop them being exploited. The report I received stated that
3363 data on the command line could provoke the exploit when Exim was running as
3364 exim, but did not say which command line option was involved. All I could
3365 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3366 running as the user.
3369 61. There was a buffer overflow vulnerability in the SPA authentication code
3370 (which came originally from the Samba project). I have added a test to the
3371 spa_base64_to_bits() function which I hope fixes it.
3374 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3375 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3377 63. The daemon start-up calls getloadavg() while still root for those OS that
3378 need the first call to be done as root, but it missed one case: when
3379 deliver_queue_load_max is set with deliver_drop_privilege. This is
3380 necessary for the benefit of the queue runner, because there is no re-exec
3381 when deliver_drop_privilege is set.
3383 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3384 This has been fixed.
3386 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3387 was in use, was not putting the data itself into the right store pool;
3388 consequently, it could be overwritten for a subsequent message in the same
3389 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3392 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3394 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3395 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3396 the function were treating the return as a boolean value, which happened to
3397 work because 0=false and not-0=true, but is not correct code.
3399 68. The host_aton() function was not handling scoped IPv6 addresses (those
3400 with, for example, "%eth0" on the end) correctly.
3402 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3403 negated items (that is, ~something) in unsigned ints. Some compilers
3404 apparently mutter when there is no cast.
3406 70. If an address verification called from an ACL failed, and did not produce a
3407 user-specific message (i.e. there was only a "system" message), nothing was
3408 put in $acl_verify_message. In this situation, it now puts the system
3411 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3412 session; change 4.31/43 added the unwanted input to the log line - except
3413 that it did not do this in the start of session case. It now does.
3415 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3416 This is wrong and can cause the other end to generate a synchronization
3417 error if it is another Exim or anything else that does the synchronization
3418 check. A QUIT command is no longer sent after a timeout.
3420 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3421 during host lookups.
3423 74. The defer_ok option of callout verification was not working if it was used
3424 when verifying addresses in header lines, that is, for this case:
3426 verify = header_sender/callout=defer_ok
3428 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3429 those file descriptors could be used for SMTP connections. If anything
3430 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3431 could be sent to the SMTP client, causing chaos. The daemon now opens
3432 stdin, stdout, and stderr to /dev/null when it puts itself into the
3435 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3436 log by default. The user can override this with suitable Perl magic.
3438 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3439 the log message when discard triggers, was not working for the DATA ACL or
3440 for the non-SMTP ACL.
3442 78. Error message wording change in sieve.c.
3444 79. If smtp_accept_max_per_host was set, the number of connections could be
3445 restricted to fewer than expected, because the daemon was trying to set up
3446 a new connection before checking whether the processes handling previous
3447 connections had finished. The check for completed processes is now done
3448 earlier. On busy systems, this bug wouldn't be noticed because something
3449 else would have woken the daemon, and it would have reaped the completed
3452 80. If a message was submitted locally by a user whose login name contained one
3453 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3454 It caused a spool format error. I have fixed the spool reading code. A
3455 related problem was that the "from" clause in the Received: line became
3456 illegal because of the space(s). It is now covered by ${quote_local_part.
3458 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3461 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3463 83. Patch from Sieve maintainer to fix unterminated string problem in
3464 "vacation" handling.
3466 84. Some minor changes to the Linux configuration files to help with other
3467 OS variants using glibc.
3469 85. One more patch for Sieve to update vacation handling to latest spec.
3472 ----------------------------------------------------
3473 See the note above about the 4.44 and 4.50 releases.
3474 ----------------------------------------------------
3480 1. Change 4.43/35 introduced a bug that caused file counts to be
3481 incorrectly computed when quota_filecount was set in an appendfile
3484 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3485 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3488 3. Give more explanation in the error message when the command for a transport
3489 filter fails to execute.
3491 4. There are several places where Exim runs a non-Exim command in a
3492 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3493 was being done only for the command run by the queryprogram router. It is
3494 now done for all such subprocesses. The other cases are: ${run, transport
3495 filters, and the commands run by the lmtp and pipe transports.
3497 5. Some older OS have a limit of 256 on the maximum number of file
3498 descriptors. Exim was using setrlimit() to set 1000 as a large value
3499 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3500 systems. I've change it so that if it can't get 1000, it tries for 256.
3502 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3503 was an oversight, and furthermore, ever since the addition of extra
3504 controls (e.g. 4.43/32), the checks on when to allow different forms of
3505 "control" were broken. There should now be diagnostics for all cases when a
3506 control that does not make sense is encountered.
3508 7. $recipients is now available in the predata ACL (oversight).
3510 8. Tidy the search cache before the fork to do a delivery from a message
3511 received from the command line. Otherwise the child will trigger a lookup
3512 failure and thereby defer the delivery if it tries to use (for example) a
3513 cached ldap connection that the parent has called unbind on.
3515 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3516 of $address_data from the recipient verification was clobbered by the
3517 sender verification.
3519 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3520 was its contents. (It was OK if the option was not defined at all.)
3522 11. A "Completed" log line is now written for messages that are removed from
3523 the spool by the -Mrm option.
3525 12. $host_address is now set to the target address during the checking of
3526 ignore_target_hosts.
3528 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3529 being passed; this would have caused $sender_host_name to have been used if
3530 matching the list had actually called for a host name (not very likely,
3531 since this list is usually IP addresses). A host name is now passed as
3534 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3535 code that specifies a non-restarting handler (typically sigaction() in
3536 modern systems) in an attempt to fix a rare and obscure crash bug.
3538 15. Narrowed the window for a race in the daemon that could cause it to ignore
3539 SIGCHLD signals. This is not a major problem, because they are used only to
3540 wake it up if nothing else does.
3542 16. A malformed maildirsize file could cause Exim to calculate negative values
3543 for the mailbox size or file count. Odd effects could occur as a result.
3544 The maildirsize information is now recalculated if the size or filecount
3547 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3548 support for a long time. Removed HAVE_SYS_VFS_H.
3550 18. Updated exipick to current release from John Jetmore.
3552 19. Allow an empty sender to be matched against a lookup in an address list.
3553 Previously the only cases considered were a regular expression, or an
3556 20. Exim went into a mad DNS lookup loop when doing a callout where the
3557 host was specified on the transport, if the DNS lookup yielded more than
3560 21. The RFC2047 encoding function was originally intended for short strings
3561 such as real names; it was not keeping to the 75-character limit for
3562 encoded words that the RFC imposes. It now respects the limit, and
3563 generates multiple encoded words if necessary. To be on the safe side, I
3564 have increased the buffer size for the ${rfc2047: expansion operator from
3567 22. Failure to deliver a bounce message always caused it to be frozen, even if
3568 there was an errors_to setting on the router. The errors_to setting is now
3571 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3572 canonical form (fully expanded) before being placed in
3573 $sender_host_address.
3575 24. Updated eximstats to version 1.33
3577 25. Include certificate and key file names in error message when GnuTLS fails
3578 to set them up, because the GnuTLS error message doesn't include the name
3579 of the failing file when there is a problem reading it.
3581 26. Expand error message when OpenSSL has problems setting up cert/key files.
3584 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3585 (this can affect the format of dates).
3587 28. exim_tidydb, when checking for the continued existence of a message for
3588 which it has found a message-specific retry record, was not finding
3589 messages that were in split spool directories. Consequently, it was
3590 deleting retry records that should have stayed in existence.
3592 29. eximstats updated to version 1.35
3593 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3594 1.35 - bugfix such that pie charts by volume are generated correctly
3596 30. The SPA authentication driver was not abandoning authentication and moving
3597 on to the next authenticator when an expansion was forced to fail,
3598 contradicting the general specification for all authenticators. Instead it
3599 was generating a temporary error. It now behaves as specified.
3601 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3602 (the order specifies the preference for clients). The order is now AES256,
3603 AES128, 3DES, ARCFOUR128.
3605 31. Small patch to Sieve code - explicitly set From: when generating an
3608 32. Exim crashed if a remote delivery caused a very long error message to be
3609 recorded - for instance if somebody sent an entire SpamAssassin report back
3610 as a large number of 550 error lines. This bug was coincidentally fixed by
3611 increasing the size of one of Exim's internal buffers (big_buffer) that
3612 happened as part of the Exiscan merge. However, to be on the safe side, I
3613 have made the code more robust (and fixed the comments that describe what
3616 33. Some experimental protocols are using DNS PTR records for new purposes. The
3617 keys for these records are domain names, not reversed IP addresses. The
3618 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3619 leaves it alone. Component reversal etc. now happens only for IP addresses.
3622 34. The host_aton() function is supposed to be passed a string that is known
3623 to be a valid IP address. However, in the case of IPv6 addresses, it was
3624 not checking this. This is a hostage to fortune. Exim now panics and dies
3625 if the condition is not met. A case was found where this could be provoked
3626 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3627 components; fortuitously, this particular loophole had already been fixed
3628 by change 4.50/55 or 4.44/33 above.
3630 If there are any other similar loopholes, the new check in host_aton()
3631 itself should stop them being exploited. The report I received stated that
3632 data on the command line could provoke the exploit when Exim was running as
3633 exim, but did not say which command line option was involved. All I could
3634 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3635 running as the user.
3638 35. There was a buffer overflow vulnerability in the SPA authentication code
3639 (which came originally from the Samba project). I have added a test to the
3640 spa_base64_to_bits() function which I hope fixes it.
3643 36. The daemon start-up calls getloadavg() while still root for those OS that
3644 need the first call to be done as root, but it missed one case: when
3645 deliver_queue_load_max is set with deliver_drop_privilege. This is
3646 necessary for the benefit of the queue runner, because there is no re-exec
3647 when deliver_drop_privilege is set.
3649 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3650 was in use, was not putting the data itself into the right store pool;
3651 consequently, it could be overwritten for a subsequent message in the same
3652 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3655 38. Sometimes the final signoff response after QUIT could fail to get
3656 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3657 < 0 before doing a fflush(). This bug looks as though it goes back to the
3658 introduction of TLS in release 3.20, but "sometimes" must have been rare
3659 because the tests only now provoked it.
3665 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3666 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3667 Now it returns an integer. A number of calls were still expecting a T/F
3668 return. Fortuitously, in all cases, the tests worked in OK situations,
3669 which is the norm. However, things would have gone wrong on any write
3670 failures on the smtp file descriptor. This function is used when sending
3671 messages over SMTP and also when doing verify callouts.
3673 2. When Exim is called to do synchronous delivery of a locally submitted
3674 message (the -odf or -odi options), it no longer closes stderr before doing
3677 3. Implemented the mua_wrapper option.
3679 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3681 5. Implemented the functions header_remove(), header_testname(),
3682 header_add_at_position(), and receive_remove_recipient(), and exported them
3685 6. If an ACL "warn" statement specified the addition of headers, Exim already
3686 inserted X-ACL-Warn: at the start if there was no header name. However, it
3687 was not making this test for the second and subsequent header lines if
3688 there were newlines in the string. This meant that an invalid header could
3689 be inserted if Exim was badly configured.
3691 7. Allow an ACL "warn" statement to add header lines at the start or after all
3692 the Received: headers, as well as at the end.
3694 8. Added the rcpt_4xx retry error code.
3696 9. Added postmaster_mailfrom=xxx to callout verification option.
3698 10. Added mailfrom=xxxx to the callout verification option, for verify=
3701 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3702 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3704 12. Inserted some casts to stop certain compilers warning when using pointer
3705 differences as field lengths or precisions in printf-type calls (mostly
3706 affecting debugging statements).
3708 13. Added optional readline() support for -be (dynamically loaded).
3710 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3711 same clock tick as a message's arrival, so that its received time was the
3712 same as the "first fail" time on the retry record, and that message
3713 remained on the queue past the ultimate address timeout, every queue runner
3714 would try a delivery (because it was past the ultimate address timeout) but
3715 after another failure, the ultimate address timeout, which should have then
3716 bounced the address, did not kick in. This was a "< instead of <=" error;
3717 in most cases the first failure would have been in the next clock tick
3718 after the received time, and all would be well.
3720 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3721 being recognized when the domain list was tested by the match_domain
3722 condition in an expansion string.
3724 16. Added the ${str2b64: operator.
3726 17. Exim was always calling setrlimit() to set a large limit for the number of
3727 processes, without checking whether the existing limit was already
3728 adequate. (It did check for the limit on file descriptors.) Furthermore,
3729 errors from getrlimit() and setrlimit() were being ignored. Now they are
3730 logged to the main and panic logs, but Exim does carry on, to try to do its
3731 job under whatever limits there are.
3733 18. Imported PCRE 5.0.
3735 19. Trivial typo in log message " temporarily refused connection" (the leading
3738 20. If the log selector return_path_on_delivery was set and an address was
3739 redirected to /dev/null, the delivery process crashed because it assumed
3740 that a return path would always be set for a "successful" delivery. In this
3741 case, the whole delivery is bypassed as an optimization, and therefore no
3744 21. Internal re-arrangement: the function for sending a challenge and reading
3745 a response while authentication was assuming a zero-terminated challenge
3746 string. It's now changed to take a pointer and a length, to allow for
3747 binary data in such strings.
3749 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3751 23. Exim was not respecting finduser_retries when seeking the login of the
3752 uid under which it was called; it was always trying 10 times. (The default
3753 setting of finduser_retries is zero.) Also, it was sleeping after the final
3754 failure, which is pointless.
3756 24. Implemented tls_on_connect_ports.
3758 25. Implemented acl_smtp_predata.
3760 26. If the domain in control=submission is set empty, Exim assumes that the
3761 authenticated id is a complete email address when it generates From: or
3762 Sender: header lines.
3764 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3765 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3766 chown and chgrp in /bin and hostname in /usr/bin.
3768 28. Exim was keeping the "process log" file open after each use, just as it
3769 does for the main log. This opens the possibility of it remaining open for
3770 long periods when the USR1 signal hits a daemon. Occasional processlog
3771 errors were reported, that could have been caused by this. Anyway, it seems
3772 much more sensible not to leave this file open at all, so that is what now
3775 29. The long-running daemon process does not normally write to the log once it
3776 has entered its main loop, and it closes the log before doing so. This is
3777 so that log files can straightforwardly be renamed and moved. However,
3778 there are a couple of unusual error situations where the daemon does write
3779 log entries, and I had neglected to close the log afterwards.
3781 30. The text of an SMTP error response that was received during a remote
3782 delivery was being truncated at 512 bytes. This is too short for some of
3783 the long messages that one sometimes sees. I've increased the limit to
3786 31. It is now possible to make retry rules that apply only when a message has a
3787 specific sender, in particular, an empty sender.
3789 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3790 it possible to be selective about when SMTP synchronization is enforced.
3792 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3794 32. Implemented hosts_connection_nolog.
3796 33. Added an ACL for QUIT.
3798 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3801 35. Added mailbox_size and mailbox_filecount to appendfile.
3803 36. Added control = no_multiline_responses to ACLs.
3805 37. There was a bug in the logic of the code that waits for the clock to tick
3806 in the case where the clock went backwards by a substantial amount such
3807 that the microsecond fraction of "now" was more than the microsecond
3808 fraction of "then" (but the whole seconds number was less).
3810 38. Added support for the libradius Radius client library this is found on
3811 FreeBSD (previously only the radiusclient library was supported).
3817 1. When certain lookups returned multiple values in the form name=value, the
3818 quoting of the values was not always being done properly. Specifically:
3819 (a) If the value started with a double quote, but contained no whitespace,
3821 (b) If the value contained whitespace other than a space character (i.e.
3822 tabs or newlines or carriage returns) it was not quoted.
3823 This fix has been applied to the mysql and pgsql lookups by writing a
3824 separate quoting function and calling it from the lookup code. The fix
3825 should probably also be applied to nisplus, ibase and oracle lookups, but
3826 since I cannot test any of those, I have not disturbed their existing code.
3828 2. A hit in the callout cache for a specific address caused a log line with no
3829 reason for rejecting RCPT. Now it says "Previous (cached) callout
3830 verification failure".
3832 3. There was an off-by-one bug in the queryprogram router. An over-long
3833 return line was truncated at 256 instead of 255 characters, thereby
3834 overflowing its buffer with the terminating zero. As well as fixing this, I
3835 have increased the buffer size to 1024 (and made a note to document this).
3837 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3838 when Exim is waiting for an SMTP response from a remote server, Exim
3839 restarts its select() call on the socket, thereby resetting its timeout.
3840 This is not a problem when such interrupts are rare. Somebody set up a cron
3841 job to run exiwhat every 2 minutes, which is less than the normal select()
3842 timeout (5 or 10 minutes). This meant that the select() timeout never
3843 kicked in because it was always reset. I have fixed this by comparing the
3844 time when an interrupt arrives with the time at the start of the first call
3845 to select(). If more time than the timeout has elapsed, the interrupt is
3846 treated as a timeout.
3848 5. Some internal re-factoring in preparation for the addition of Sieve
3849 extensions (by MH). In particular, the "personal" test is moved to a
3850 separate function, and given an option for scanning Cc: and Bcc: (which is
3851 not set for Exim filters).
3853 6. When Exim created an email address using the login of the caller as the
3854 local part (e.g. when creating a From: or Sender: header line), it was not
3855 quoting the local part when it contained special characters such as @.
3857 7. Installed new OpenBSD configuration files.
3859 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3860 try to make them clearer.
3862 9. Callout options, other than the timeout value, were being ignored when
3863 verifying sender addresses in header lines. For example, when using
3865 verify = header_sender/callout=no_cache
3867 the cache was (incorrectly) being used.
3869 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3870 only the Cygwin environment.
3872 11. When return_path_on_delivery was set as a log selector, if different remote
3873 addresses in the same message used different return paths and parallel
3874 remote delivery occurred, the wrong values would sometimes be logged.
3875 (Whenever a remote delivery process finished, the return path value from
3876 the most recently started remote delivery process was logged.)
3878 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3879 lines when AUTH and/or TLS are in use. This is the "received protocol"
3880 field. Exim used to use "asmtp" for authenticated SMTP, without any
3881 indication (in the protocol name) for TLS use. Now it follows the RFC and
3882 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3883 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3884 names appear in log lines as well as in Received: header lines.
3886 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3887 extensions, and comparison tests, and to fix some bugs.
3889 14. Changes to the "personal" filter test:
3891 (1) The test was buggy in that it was just doing the equivalent of
3892 "contains" tests on header lines. For example, if a user's address was
3893 anne@some.where, the "personal" test would incorrectly be true for
3895 To: susanne@some.where
3897 This test is now done by extracting each address from the header in turn,
3898 and checking the entire address. Other tests that are part of "personal"
3899 are now done using regular expressions (for example, to check local parts
3900 of addresses in From: header lines).
3902 (2) The list of non-personal local parts in From: addresses has been
3903 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3904 taken from the Sieve specification recommendations.
3906 (3) If the message contains any header line starting with "List-" it is
3907 treated as non-personal.
3909 (4) The test for "circular" in the Subject: header line has been removed
3910 because it now seems ill-conceived.
3912 15. Minor typos in src/EDITME comments corrected.
3914 16. Installed latest exipick from John Jetmore.
3916 17. If headers_add on a router specified a text string that was too long for
3917 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3918 of string_sprintf() is now avoided.
3920 18. $message_body_size was not set (it was always zero) when running the DATA
3921 ACL and the local_scan() function.
3923 19. For the "mail" command in an Exim filter, no default was being set for
3924 the once_repeat time, causing a random time value to be used if "once" was
3925 specified. (If the value happened to be <= 0, no repeat happened.) The
3926 default is now 0s, meaning "never repeat". The "vacation" command was OK
3927 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3928 (I found it when inspecting the code).
3930 20. There is now an overall timeout for performing a callout verification. It
3931 defaults to 4 times the callout timeout, which applies to individual SMTP
3932 commands during the callout. The overall timeout applies when there is more
3933 than one host that can be tried. The timeout is checked before trying the
3934 next host. This prevents very long delays if there are a large number of
3935 hosts and all are timing out (e.g. when the network connections are timing
3936 out). The value of the overall timeout can be changed by specifying an
3937 additional sub-option for "callout", called "maxwait". For example:
3939 verify = sender/callout=5s,maxwait=20s
3941 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3942 to the end before writing, but this should make it even safer).
3944 22. Exim was forgetting that it had advertised PIPELINING for the second and
3945 subsequent messages on an SMTP connection. It was also not resetting its
3946 memory on STARTTLS and an internal HELO.
3948 23. When Exim logs an SMTP synchronization error within a session, it now
3949 records whether PIPELINING has been advertised or not.
3951 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3952 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3953 rather than long int.
3955 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3957 26. Added the never_mail option to autoreply.
3963 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3964 crash if the getsockname() call failed; this can happen if a connection is
3965 closed very soon after it is established. The problem was simply in the
3966 order in which certain operations were done, causing Exim to try to write
3967 to the SMTP stream before it had set up the file descriptor. The bug has
3968 been fixed by making things happen in the correct order.
3974 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3975 before the connection was closed, thus losing the rejection response.
3977 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3978 some early Solaris releases, but causes trouble in current releases where
3979 socklen_t is defined.
3981 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3984 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3987 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3988 that releasing the top of it at the end releases what was used for sub-
3989 expansions (unless the block got too big). However, discard this block if
3990 the first thing is a variable or header, so that we can use its block when
3991 it is dynamic (useful for very large $message_headers, for example).
3993 6. Lookups now cache *every* query, not just the most recent. A new, separate
3994 store pool is used for this. It can be recovered when all lookup caches are
3995 flushed. Lookups now release memory at the end of their result strings.
3996 This has involved some general refactoring of the lookup sources.
3998 7. Some code has been added to the store_xxx() functions to reduce the amount
3999 of flapping under certain conditions.
4001 8. log_incoming_interface used to affect only the <= reception log lines. Now
4002 it causes the local interface and port to be added to several more SMTP log
4003 lines, for example "SMTP connection from", and rejection lines.
4005 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4007 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4009 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4010 could be overwritten at the end of the current message (or the start of a
4011 new message if it was set in a HELO ACL). The value is now preserved for
4012 the duration of the SMTP connection.
4014 12. If a transport had a headers_rewrite setting, and a matching header line
4015 contained an unqualified address, that address was qualified, even if it
4016 did not match any rewriting rules. The underlying bug was that the values
4017 of the flags that permit the existence of unqualified sender and recipient
4018 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4019 non-local messages, and by -bnq for local messages) were not being
4020 preserved with the message after it was received.
4022 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4023 "next input=" as part of the text comprising the host identity instead of
4024 the correct text. The code was using the same buffer for two different
4025 strings. However, depending on which order the printing function evaluated
4026 its arguments, the bug did not always show up. Under Linux, for example, my
4027 test suite worked just fine.
4029 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4030 doesn't work with some older versions of Perl. It has been changed to "my",
4031 which in any case is probably the better facility to use.
4033 15. A really picky compiler found some instances of statements for creating
4034 error messages that either had too many or two few arguments for the format
4037 16. The size of the buffer for calls to the DNS resolver has been increased
4038 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4039 for addresses that have a lot of PTR records. This alleviates a problem; it
4040 does not fully solve it.
4042 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4043 buffer now truncates the list and logs the incident, which is the same
4044 action as happens when Exim is looking up a host name and its aliases.
4045 Previously in this situation something unpredictable would happen;
4046 sometimes it was "internal error: store_reset failed".
4048 18. If a server dropped the connection unexpectedly when an Exim client was
4049 using GnuTLS and trying to read a response, the client delivery process
4050 crashed while trying to generate an error log message.
4052 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4053 string, for example:
4055 warn message = H1: something\nH2: something
4057 the text was added as a single header line from Exim's point of view
4058 though it ended up OK in the delivered message. However, searching for the
4059 second and subsequent header lines using $h_h2: did not work. This has been
4060 fixed. Similarly, if a system filter added multiple headers in this way,
4061 the routers could not see them.
4063 20. Expanded the error message when iplsearch is called with an invalid key to
4064 suggest using net-iplsearch in a host list.
4066 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4067 ACLs are no longer actually imposed (and a message to that effect is
4070 22. If a "gecos" field in a passwd entry contained escaped characters, in
4071 particular, if it contained a \" sequence, Exim got it wrong when building
4072 a From: or a Sender: header from that name. A second bug also caused
4073 incorrect handling when an unquoted " was present following a character
4074 that needed quoting.
4076 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4077 was not being matched caselessly.
4079 24. Arranged for all hyphens in the exim.8 source to be escaped with
4082 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4083 properly. Recipient callout cache records were still being keyed to include
4084 the sender, even when use_sender was set false. This led to far more
4085 callouts that were necessary. The sender is no longer included in the key
4086 when use_sender is false.
4088 26. Added "control = submission" modifier to ACLs.
4090 27. Added the ${base62d: operator to decode base 62 numbers.
4092 28. dnsdb lookups can now access SRV records.
4094 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4095 the configuration file.
4097 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4098 (-v) mode. This makes the output for a verbose queue run more intelligible.
4100 31. Added a use_postmaster feature to recipient callouts.
4102 32. Added the $body_zerocount variable, containing the number of binary zero
4103 bytes in the message body.
4105 33. The time of last modification of the "new" subdirectory is now used as the
4106 "mailbox time last read" when there is a quota error for a maildir
4109 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4111 35. Added +ignore_unknown as a special item in host lists.
4113 36. Code for decoding IPv6 addresses in host lists is now included, even if
4114 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4115 address was recognized as an IP address, but was then not correctly decoded
4116 into binary, causing unexpected and incorrect effects when compared with
4123 1. Very minor rewording of debugging text in manualroute to say "list of
4124 hosts" instead of "hostlist".
4126 2. If verify=header_syntax was set, and a header line with an unqualified
4127 address (no domain) and a large number of spaces between the end of the
4128 name and the colon was received, the reception process suffered a buffer
4129 overflow, and (when I tested it) crashed. This was caused by some obsolete
4130 code that should have been removed. The fix is to remove it!
4132 3. When running in the test harness, delay a bit after writing a bounce
4133 message to get a bit more predictability in the log output.
4135 4. Added a call to search_tidyup() just before forking a reception process. In
4136 theory, someone could use a lookup in the expansion of smtp_accept_max_
4137 per_host which, without the tidyup, could leave open a database connection.
4139 5. Added the variables $recipient_data and $sender_data which get set from a
4140 lookup success in an ACL "recipients" or "senders" condition, or a router
4141 "senders" option, similar to $domain_data and $local_part_data.
4143 6. Moved the writing of debug_print from before to after the "senders" test
4146 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4147 problems for message scanning, either using a data ACL, or using
4148 local_scan() because the Received: header was not generated till after they
4149 were called (in order to set the time as the time of reception completion).
4150 I have revised the way this works. The header is now generated after the
4151 body is received, but before the ACL or local_scan() are called. After they
4152 are run, the timestamp in the header is updated.
4158 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4159 before starting a queue runner without re-exec. This happened only when
4160 deliver_drop_privilege was set or when the Exim user was set to root. The
4161 effect of the bug was that timeouts during subsequent deliveries caused
4162 crashes instead of being properly handled. The handler is now left at its
4163 default (and expected) setting.
4165 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4166 message, again when deliver_drop_privilege is set or Exim is run as root.
4167 The bug described in (1) was not present in this case, but the tidying up
4168 of the other signals was missing. I have made the two cases consistent.
4170 3. The ignore_target_hosts setting on a manualroute router was being ignored
4171 for hosts that were looked up using the /MX notation.
4173 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4176 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4177 operated on the sender address. After changing the $sender_address to <>
4178 for the sender address verify, Exim was re-instated it as the original
4179 (before rewriting) address, but remembering that it had rewritten it, so it
4180 wasn't rewriting it again. This bug also had the effect of breaking the
4181 sender address verification caching when the sender address was rewritten.
4183 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4184 This has been changed so that if the ip literal address matches
4185 ignore_target_hosts, the router declines.
4187 7. Added expansion conditions match_domain, match_address, and match_local_
4188 part (NOT match_host).
4190 8. The placeholder for the Received: header didn't have a length field set.
4192 9. Added code to Exim itself and to exim_lock to test for a specific race
4193 condition that could lead to file corruption when using MBX delivery. The
4194 issue is with the lockfile that is created in /tmp. If this file is removed
4195 after a process has opened it but before that process has acquired a lock,
4196 there is the potential for a second process to recreate the file and also
4197 acquire a lock. This could lead to two Exim processes writing to the file
4198 at the same time. The added code performs the same test as UW imapd; it
4199 checks after acquiring the lock that its file descriptor still refers to
4200 the same named file.
4202 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4203 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4206 11. Added the smtp_active_hostname option. If used, this will typically be made
4207 to depend on the incoming interface address. Because $interface_address is
4208 not set up until the daemon has forked a reception process, error responses
4209 that can happen earlier (such as "too many connections") no longer contain
4212 12. If an expansion in a condition on a "warn" statement fails because a lookup
4213 defers, the "warn" statement is abandoned, and the next ACL statement is
4214 processed. Previously this caused the whole ACL to be aborted.
4216 13. Added the iplsearch lookup type.
4218 14. Added ident_timeout as a log selector.
4220 15. Added tls_certificate_verified as a log selector.
4222 16. Added a global option tls_require_ciphers (compare the smtp transport
4223 option of the same name). This controls incoming TLS connections.
4225 17. I finally figured out how to make tls_require_ciphers do a similar thing
4226 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4227 before starting the TLS session.
4229 18. Tabs are now shown as \t in -bP output.
4231 19. If the log selector return_path_on_delivery was set, Exim crashed when
4232 bouncing a message because it had too many Received: header lines.
4234 20. If two routers both had headers_remove settings, and the first one included
4235 a superfluous trailing colon, the final name in the first list and the
4236 first name in the second list were incorrectly joined into one item (with a
4237 colon in the middle).
4243 1. Added -C and -D options to the exinext utility, mainly to make it easier
4244 to include in the automated testing, but these could be helpful when
4245 multiple configurations are in use.
4247 2. The exinext utility was not formatting the output nicely when there was
4248 an alternate port involved in the retry record key, nor when there was a
4249 message id as well (for retries that were specific to a specific message
4250 and a specific host). It was also confused by IPv6 addresses, because of
4251 the additional colons they contain. I have fixed the IPv4 problem, and
4252 patched it up to do a reasonable job for IPv6.
4254 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4255 delivery, the log line now contains "pipelined" if PIPELINING was used.
4257 4. An SMTP transport process used to panic and die if the bind() call to set
4258 an explicit outgoing interface failed. This has been changed; it is now
4259 treated in the same way as a connect() failure.
4261 5. A reference to $sender_host_name in the part of a conditional expansion
4262 that was being skipped was still causing a DNS lookup. This no longer
4265 6. The def: expansion condition was not recognizing references to header lines
4266 that used bh_ and bheader_.
4268 7. Added the _cache feature to named lists.
4270 8. The code for checking quota_filecount in the appendfile transport was
4271 allowing one more file than it should have been.
4273 9. For compatibility with Sendmail, the command line option
4281 and sets the incoming protocol and host name (for trusted callers). The
4282 host name and its colon can be omitted when only the protocol is to be set.
4283 Note the Exim already has two private options, -pd and -ps, that refer to
4284 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4285 "s", but I don't think that's a major issue.
4287 10. A number of refactoring changes to the code, none of which should affect
4290 (a) The number of logging options was getting close to filling up the
4291 32-bit word that was used as a bit map. I have split them into two classes:
4292 those that are passed in the argument to log_write(), and those that are
4293 only ever tested independently outside of that function. These are now in
4294 separate 32-bit words, so there is plenty of room for expansion again.
4295 There is no change in the user interface or the logging behaviour.
4297 (b) When building, for example, log lines, the code previously used a
4298 macro that called string_cat() twice, in order to add two strings. This is
4299 not really sufficiently general. Furthermore, there was one instance where
4300 it was actually wrong because one of the argument was used twice, and in
4301 one call a function was used. (As it happened, calling the function twice
4302 did not affect the overall behaviour.) The macro has been replaced by a
4303 function that can join an arbitrary number of extra strings onto a growing
4306 (c) The code for expansion conditions now uses a table and a binary chop
4307 instead of a serial search (which was left over from when there were very
4308 few conditions). Also, it now recognizes conditions like "pam" even when
4309 the relevant support is not compiled in: a suitably worded error message is
4310 given if an attempt is made to use such a condition.
4312 11. Added ${time_interval:xxxxx}.
4314 12. A bug was causing one of the ddress fields not to be passed back correctly
4315 from remote delivery subprocesses. The field in question was not being
4316 subsequently used, so this caused to problems in practice.
4318 13. Added new log selectors queue_time and deliver_time.
4320 14. Might have fixed a bug in maildirsizefile handling that threw up
4321 "unexpected character" debug warnings, and recalculated the data
4322 unnecessarily. In any case, I expanded the warning message to give more
4325 15. Added the message "Restricted characters in address" to the statements in
4326 the default ACL that block characters like @ and % in local parts.
4328 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4329 Three changes have been made:
4331 (a) There was a serious bug; a negative response to MAIL caused the whole
4332 recipient domain to be cached as invalid, thereby blocking all messages
4333 to all local parts at the same domain, from all senders. This bug has
4334 been fixed. The domain is no longer cached after a negative response to
4335 MAIL if the sender used is not empty.
4337 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4340 (c) A new callout option, "use_sender" has been added for people who want
4341 the modified behaviour.
4347 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4350 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4351 indeed breaks things for older releases.
4353 3. Added additional logging to the case where there is a problem reading data
4354 from a filter that is running in a subprocess using a pipe, in order to
4355 try to track down a specific problem.
4357 4. Testing facility fudge: when running in the test harness and attempting
4358 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4359 getting "No route to host". Convert this to a timeout.
4361 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4364 6. Some OS don't have socklen_t but use size_t instead. This affects the
4365 fifth argument of getsockopt() amongst other things. This is now
4366 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4367 can be set for individual OS. I have set it for SunOS5, OSF1, and
4368 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4369 some earlier ones do not.
4371 7. Change 4.30/15 was not doing the test caselessly.
4373 8. The standard form for an IPv6 address literal was being rejected by address
4374 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4375 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4376 this, as well as the form without the "IPv6" on the front (but only when
4377 address literals are enabled, of course).
4379 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4381 10. Exim crashed if a message with an empty sender address specified by -f
4382 encountered a router with an errors_to setting. This could be provoked only
4383 by a command such as
4387 where an empty string was supplied; "<>" did not hit this bug.
4389 11. Installed PCRE release 4.5.
4391 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4392 remained set. It is now erased.
4394 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4395 times from message ids (which are base 36 rather than the normal 62).
4397 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4398 were being counted as actual protocol errors, and logged if the log
4399 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4400 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4401 and DATA following a set of rejected RCPTs do not count as protocol errors.
4402 In other words, Exim assumes they were pipelined, though this may not
4403 actually be the case. Of course, in all cases the client gets an
4404 appropriate error code.
4406 15. If a lookup fails in an ACL condition, a message about the failure may
4407 be available; it is used if testing the ACL cannot continue, because most
4408 such messages specify what the cause of the deferral is. However, some
4409 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4410 that caused an old message to be retained and used if a later statement
4411 caused a defer, replacing the real cause of the deferral.
4413 16. If an IP address had so many PTR records that the DNS lookup buffer
4414 was not large enough to hold them, Exim could crash while trying to process
4415 the truncated data. It now detects and logs this case.
4417 17. Further to 4.21/58, another change has been made: if (and only if) the
4418 first line of a message (the first header line) ends with CRLF, a bare LF
4419 in a subsequent header line has a space inserted after it, so as not to
4420 terminate the header.
4422 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4423 unnecessarily, used atoi() instead of strtol(), and didn't check the
4424 termination when getting file sizes from file names by regex.
4426 19. Completely re-implemented the support for maildirsize files, in the light
4427 of a number of problems with the previous contributed implementation
4428 (4.30/29). In particular:
4430 . If the quota is zero, the maildirsize file is maintained, but no quota is
4433 . If the maildir directory does not exist, it is created before any attempt
4434 to write a maildirsize file.
4436 . The quota value in the file is just a cache; if the quota is changed in
4437 the transport, the new value overrides.
4439 . A regular expression is available for excluding directories from the
4442 20. The autoreply transport checks the characters in options that define the
4443 message's headers; it allows continued headers, but it was checking with
4444 isspace() after an embedded newline instead of explicitly looking for a
4447 21. If all the "regular" hosts to which an address was routed had passed their
4448 expiry times, and had not reached their retry times, the address was
4449 bounced, even if fallback hosts were defined. Now Exim should go on to try
4452 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4453 equivalent code in the SMTP transport. Some hosts send humungous responses
4454 to HELO/EHLO, more than 1024 it seems.
4456 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4457 gives compiler warnings in some environments. I've now done it "properly",
4460 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4461 (because of problems with the built-in one) was declared to return uschar *
4462 instead of char *, causing compiler failure.
4464 25. Fixed a file descriptor leak when processing alias/forward files.
4466 26. Fixed a minor format string issue in dbfn.c.
4468 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4470 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4471 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4474 29. When a sender address is verified, it is cached, to save repeating the test
4475 when there is more than one recipient in a message. However, when the
4476 verification involves a callout, it is possible for different callout
4477 options to be set for different recipients. It is too complicated to keep
4478 track of this in the cache, so now Exim always runs a verification when a
4479 callout is required, relying on the callout cache for the optimization.
4480 The overhead is duplication of the address routing, but this should not be
4483 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4484 to be verified with callout=postmaster, and the main callout worked but the
4485 postmaster check failed, the verification correctly failed. However, if a
4486 subsequent RCPT command asked for sender verification *without* the
4487 postmaster check, incorrect caching caused this verification also to fail,
4490 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4491 it was not caching the DNS options (qualify_single, search_parents) that
4492 were used when the lookup failed. A subsequent lookup with different
4493 options therefore always gave the same answer, though there were cases
4494 where it should not have. (Example: a "domains = !$mx_any" option on a
4495 dnslookup router: the "domains" option is always processed without any
4496 widening, but the router might have qualify_single set.) Now Exim uses the
4497 cached value only when the same options are set.
4499 32. Added John Jetmore's "exipick" utility to the distribution.
4501 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4502 than a timeout (e.g. a certificate is required, and is not provided), an
4503 Exim server now closes the connection immediately. Previously it waited for
4504 the client to close - but if the client is SSL, it seems that they each
4505 wait for each other, leading to a delay before one of them times out.
4507 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4508 maintained 0.8.x compatibility because I don't think many are using it, and
4509 it is clearly obsolete.
4511 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4514 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4515 client certificate was expired. A simple patch fixes this, though I don't
4516 understand the full logic of why the verify callback is called multiple
4519 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4520 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4521 which causes problems with some clients (such as the Certicom SSL Plus
4522 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4523 disables the coutermeasure allowing Eudora to connect."
4525 38. Exim was not checking that a write() to a log file succeeded. This could
4526 lead to Bad Things if a log got too big, in particular if it hit a file
4527 size limit. Exim now panics and dies if it cannot write to a log file, just
4528 as it does if it cannot open a log file.
4530 39. Modified OS/Makefile-Linux so that it now contains
4532 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4534 The two -D definitions ensure that Exim is compiled with large file
4535 support, which makes it possible to handle log files that are bigger than
4538 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4539 instance) a domain was checked against a named list that involved a lookup,
4540 causing $domain_data to be set, then another domain was checked against the
4541 same list, then the first domain was re-checked, the value of $domain_data
4542 after the final check could be wrong. In particular, if the second check
4543 failed, it could be set empty. This bug probably also applied to
4546 41. The strip_trailing_dot option was not being applied to the address given
4547 with the -f command-line option.
4549 42. The code for reading a message's header from the spool was incrementing
4550 $received_count, but never initializing it. This meant that the value was
4551 incorrect (doubled) while delivering a message in the same process in which
4552 it was received. In the most common configuration of Exim, this never
4553 happens - a fresh exec is done - but it can happen when
4554 deliver_drop_privilege is set.
4556 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4557 it now includes up to 150 characters of the unexpected data in the log
4560 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4561 and building data strings. The size of both of these buffers was 10 000
4562 bytes - far larger than anybody would *ever* want, thought I. Needless to
4563 say, somebody hit the limit. I have increased the maximum line length to
4564 20 000 and the maximum data length of concatenated lines to 100 000. I have
4565 also fixed two bugs, because there was no checking on these buffers. Tsk,
4566 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4567 buffer is too small.
4569 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4570 lsearch lookups. Now it does.
4572 46. When parsing a route_list item in a manualroute router, a fixed-length
4573 buffer was used for the list of hosts. I made this 1024 bytes long,
4574 thinking that nobody would ever have a list of hosts that long. Wrong.
4575 Somebody had a whole pile of complicated expansion conditions, and the
4576 string was silently truncated, leading to an expansion error. It turns out
4577 that it is easier to change to an unlimited length (owing to other changes
4578 that have happened since this code was originally written) than to build
4579 structure for giving a limitation error. The length of the item that
4580 expands into the list of hosts is now unlimited.
4582 47. The lsearch lookup could not handle data where the length of text line was
4583 more than 4095 characters. Such lines were truncated, leading to shortened
4584 data being returned. It should now handle lines of any length.
4586 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4587 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4590 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4591 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4592 no longer happens in comment lines. A list of replacements is now placed
4593 at the head of all of the source files, except those whose only change is
4594 to replace PERL_COMMAND in the very first #! line.
4596 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4597 messages on the queue, with a bottom-up merge sort, using code contributed
4598 by Michael Haardt. This should make operations like -bp somewhat faster on
4599 large queues. It won't affect queue runners, except when queue_run_in_order
4602 51. Installed eximstats 1.31 in the distribution.
4604 52. Added support for SRV lookups to the dnslookup router.
4606 53. If an ACL referred to $message_body or $message_body_end, the value was not
4607 reset for any messages that followed in the same SMTP session.
4609 54. The store-handling optimization for building very long strings was not
4610 differentiating between the different store pools. I don't think this
4611 actually made any difference in practice, but I've tidied it.
4613 55. While running the routers to verify a sender address, $sender_address
4614 was still set to the sender address. This is wrong, because when routing to
4615 send a bounce to the sender, it would be empty. Therefore, I have changed
4616 it so that, while verifying a sender address, $sender_address is set to <>.
4617 (There is no change to what happens when verifying a recipient address.)
4619 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4620 target A or AAAA records (if not already returned) without resetting the
4621 qualify_single or search_parents options of the DNS resolver. These are
4622 inappropriate in this case because the targets of MX and SRV records must
4623 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4624 target that, when qualified, matched something in the local domain. These
4625 two options are now turned off when doing these lookups.
4627 57. It seems that at least some releases of Reiserfs (which does not have the
4628 concept of a fixed number of inodes) returns zero and not -1 for the
4629 number of available inodes. This interacted badly with check_spool_inodes,
4630 which assumed that -1 was the "no such thing" setting. What I have done is
4631 to check that the total number of inodes is greater than zero before doing
4632 the test of how many are available.
4634 58. When a "warn" ACL statement has a log_message modifier, the message is
4635 remembered, and not repeated. This is to avoid a lot of repetition when a
4636 message has many recipients that cause the same warning to be written.
4637 Howewer, Exim was preserving the list of already written lines for an
4638 entire SMTP session, which doesn't seem right. The memory is now reset if a
4639 new message is started.
4641 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4642 debugging output unless log_rewrite was also set.
4644 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4645 of (int)(handle) when we know that handle contains (void *)(-1).
4647 61. The Exim daemon panic-logs an error return when it closes the incoming
4648 connection. However "connection reset by peer" seems to be common, and
4649 isn't really an error worthy of noting specially, so that particular error
4652 62. When Exim is trying to find all the local interfaces, it used to panic and
4653 die if the ioctl to get the interface flags failed. However, it seems that
4654 on at least one OS (Solaris 9) it is possible to have an interface that is
4655 included in the list of interfaces, but for which you get a failure error
4656 for this call. This happens when the interface is not "plumbed" into a
4657 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4658 failure of the "get flags" call assumes that the interface is down.
4660 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4661 makes life easier for people who are doing arithmetic on fields extracted
4662 from dates, where you often get leading zeros that should not be
4663 interpreted as octal.
4665 64. Added qualify_domain to the redirect router, to override the global
4668 65. If a pathologically long header line contained very many addresses (the
4669 report of this problem mentioned 10 000) and each of them was rewritten,
4670 Exim could use up a very large amount of memory. (It kept on making new
4671 copies of the header line as it rewrote, and never released the old ones.)
4672 At the expense of a bit more processing, the header rewriting function has
4673 been changed so that it no longer eats memory in this way.
4675 66. The generation of the Received: header has been moved from the time that a
4676 message starts to be received, to the time that it finishes. The timestamp
4677 in the Received: header should now be very close to that of the <= log
4678 line. There are two side-effects of this change:
4680 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4681 logged header lines no longer include the local Received: line, because
4682 it has not yet been created. The same applies to a copy of the message
4683 that is returned to a non-SMTP sender when a message is rejected.
4685 (b) When a filter file is tested using -bf, no additional Received: header
4686 is added to the test message. After some thought, I decided that this
4689 This change does not affect the value of $received_for. It is still set
4690 after address rewriting, but before local_scan() is called.
4692 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4694 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4695 gave an unhelpful panic error message, and a defer error. I have managed to
4696 change this behaviour so that it now rejects any supplied certificate,
4697 which seems right, as the list of acceptable certificates is empty.
4699 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4700 gave an unhelpful defer error. I have not managed to make this reject any
4701 supplied certificates, but the error message it gives is "no certificate
4702 supplied", which is not helpful.
4704 70. exigrep's output now also includes lines that are not associated with any
4705 message, but which match the given pattern. Implemented by a patch from
4706 Martin Sluka, which also tidied up the Perl a bit.
4708 71. Recipient callout verification, like sender verification, was using <> in
4709 the MAIL FROM command. This isn't really the right thing, since the actual
4710 sender may affect whether the remote host accepts the recipient or not. I
4711 have changed it to use the actual sender in the callout; this means that
4712 the cache record is now keyed on a recipient/sender pair, not just the
4713 recipient address. There doesn't seem to be a real danger of callout loops,
4714 since a callout by the remote host to check the sender would use <>.
4715 [SEE ABOVE: changed after hitting problems.]
4717 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4718 temporary errors. However, in the case of such a code being given after
4719 the end of a data transmission (i.e. after ".") Exim was failing to write
4720 a retry record for the message. (Yes, there was some broken host that was
4721 actually sending 8xx at this point.)
4723 73. An unknown lookup type in a host list could cause Exim to panic-die when
4724 the list was checked. (An example that provoked this was putting <; in the
4725 middle of a list instead of at the start.) If this happened during a DATA
4726 ACL check, a -D file could be left lying around. This kind of configuration
4727 error no longer causes Exim to die; instead it causes a defer errror. The
4728 incident is still logged to the main and panic logs.
4730 74. Buglet left over from Exim 3 conversion. The message "too many messages
4731 in one connection" was written to the rejectlog but not the mainlog, except
4732 when address rewriting (yes!) was being logged.
4734 75. Added write_rejectlog option.
4736 76. When a system filter was run not as root (that is, when system_filter_user
4737 was set), the values of the $n variables were not being returned to the
4738 main process; thus, they were not subsequently available in the $sn
4741 77. Added +return_path_on_delivery log selector.
4743 78. A connection timeout was being treated differently from recipients deferred
4744 when testing hosts_max_try with a message that was older than the host's
4745 retry timeout. (The host should not be counted, thus allowing all hosts to
4746 be tried at least once before bouncing.) This may have been the cause of an
4747 occasionally reported bug whereby a message would remain on the queue
4748 longer than the retry timeout, but would be bounced if a delivery was
4749 forced. I say "may" because I never totally pinned down the problem;
4750 setting up timeout/retry tests is difficult. See also the next item.
4752 79. The ultimate address timeout was not being applied to errors that involved
4753 a combination of host plus message (for example, a timeout on a MAIL
4754 command). When an address resolved to a number of possible hosts, and they
4755 were not all tried for each delivery (e.g. because of hosts_max_try), a
4756 message could remain on the queue longer than the retry timeout.
4758 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4761 81. Fixed an obscure SMTP outgoing bug which required at least the following
4762 conditions: (a) there was another message waiting for the same server;
4763 (b) the server returned 5xx to all RCPT commands in the first message so
4764 that the message was not completed; (c) the server dropped the connection
4765 or gave a negative response to the RSET that Exim sends to abort the
4766 transaction. The observed case was a dropped connection after DATA that had
4767 been sent in pipelining mode. That is, the server had advertised PIPELINING
4768 but was not implementing it correctly. The effect of the bug was incorrect
4769 behaviour, such as trying another host, and this could lead to a crash.
4775 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4776 and daemon.c were passed as pointers to ints; they should have been
4777 pointers to socklen_t variables (which are typically unsigned ints).
4779 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4782 3. Fixed a really odd bug that affected only the testing scheme; patching a
4783 certain fixed string in the binary changed the value of another string that
4784 happened to be identical to the end of the original first string.
4786 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4787 name", it returns that address as the IP address. On some operating
4788 systems (e.g. Solaris), it also passes back the IP address string as the
4789 "host name". However, on others (e.g. Linux), it passes back an empty
4790 string. Exim wasn't checking for this, and was changing the host name to an
4791 empty string, assuming it had been canonicized.
4793 5. Although rare, it is permitted to have more than one PTR record for a given
4794 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4795 all the names associated with an address, because they do in Solaris.
4796 However, it seems that they do not in Linux for data that comes from the
4797 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4798 I found this out when I moved to a new Linux workstation and tried to run
4799 the Exim test suite.
4801 To get round this problem I have changed the code so that it now does its
4802 own call to the DNS to look up PTR records when searching for a host name.
4803 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4804 addresses that are only in /etc/hosts are still found.
4806 This behaviour is, however, controlled by an option called host_lookup_
4807 order, which defaults to "bydns:byaddr". If people want to use the other
4808 order, or indeed, just use one or the other means of lookup, they can
4809 specify it in this variable.
4811 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4812 some operating systems, this comes back from gethostbyaddr() as an empty
4813 string, and this is what Exim used to test for. However, it seems that in
4814 other systems, "." is yielded. Exim now tests for this case too.
4816 7. The values of check_spool_space and check_log_space are now held internally
4817 as a number of kilobytes instead of an absolute number of bytes. If a
4818 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4819 kilobyte. This means that much larger values can be stored.
4821 8. Exim monitor: an attempt to get the action menu when not actually pointing
4822 at a message produces an empty menu entitled "No message selected". This
4823 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4824 no entries in it ("Shell widget menu has zero width and/or height"). So I
4825 have added a single, blank menu entry in this case.
4827 9. Added ${quote_local_part.
4829 10. MIME decoding is now applied to the contents of Subject: header lines when
4832 11. Now that a reference to $sender_host_address automatically causes a reverse
4833 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4834 host lookup before query-style lookups in lists that might use this
4835 variable. This has therefore been abolished, and the "net-" prefix is no
4836 longer necessary for query-style lookups.
4838 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4839 have been a typo for LFLAGS, so it has been changed.
4841 13. The install script calls Exim with "-C /dev/null" in order to find the
4842 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4843 to be output. Howeve, since Exim outputs its version number before the
4844 error, it didn't break the script. It just looked ugly. I fixed this by
4845 always allowing "-C /dev/null" if the caller is root.
4847 14. Ignore overlarge ACL variable number when reading spool file - insurance
4848 against a later release with more variables having written the file.
4850 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4851 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4852 this, as well as the form without the "IPv6" on the front.
4854 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4855 OS/Makefile-Darwin file.
4857 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4858 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4860 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4861 variable. However, some OS use ints for inodes. I've added cast to long int
4862 to get rid of the compiler warning.
4864 19. I had forgotten to lock out "/../" in configuration file names when
4865 ALT_CONFIG_PREFIX was set.
4867 20. Routers used for verification do not need to specify transports. However,
4868 if such a router generated a host list, and callout was configured, Exim
4869 crashed, because it could not find a port number from the (non-existent)
4870 transport. It now assumes port 25 in this circumstance.
4872 21. Added the -t option to exigrep.
4874 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4875 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4876 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4877 them is set, LOOKUP_LSEARCH is forced.
4879 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4881 24. Added sender and host information to the "rejected by local_scan()" log
4882 line; previously there was no indication of these.
4884 25. Added .include_if_exists.
4886 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4887 Linux. It turns out that not all file systems support this. Apparently some
4888 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4889 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4890 not supported on the file descriptor, is now ignored when Exim is trying to
4891 sync a directory. This applies only to Linux.
4893 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4895 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4896 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4899 29. Added support for maildirsize files from supplied patch (modified a bit).
4901 30. The use of :fail: followed by an empty string could lead Exim to respond to
4902 sender verification failures with (e.g.):
4904 550 Verification failed for <xxx>
4905 550 Sender verify failed
4907 where the first response line was missing the '-' that indicates it is not
4908 the final line of the response.
4910 31. The loop for finding the name of the user that called Exim had a hardwired
4911 limit of 10; it now uses the value of finduser_retries, which is used for
4912 all other user lookups.
4914 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4917 33. Exim was neglecting to zero errno before one call of strtol() when
4918 expanding a string and expecting an integer value. On some systems this
4919 resulted in spurious "integer overflow" errors. Also, it was casting the
4920 result into an int without checking.
4922 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4923 did not work. The code looks as if it has *never* worked, though it appears
4924 to have been documented since at least releast 1.62. I have made it work.
4926 35. The "timeout_DNS" error in retry rules, also documented since at least
4927 1.62, also never worked. As it isn't clear exactly what this means, and
4928 clearly it isn't a major issue, I have abolished the feature by treating it
4929 as "timeout", and writing a warning to the main and panic logs.
4931 36. The display of retry rules for -brt wasn't always showing the error code
4934 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4935 timeout_connect_A, timeout_connect_MX.
4937 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4938 to the empty sender.
4940 39. The daemon was not analysing the content of -oX till after it had closed
4941 stderr and disconnected from the controlling terminal. This meant that any
4942 syntax errors were only noted on the panic log, and the return code from
4943 the command was 0. By re-arranging the code a little, I've made the
4944 decoding happen first, so such errors now appear on stderr, and the return
4945 code is 1. However, the actual setting up of the sockets still happens in
4946 the disconnected process, so errors there are still only recorded on the
4949 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4950 connections (as happens on some IP stacks) was logged at start up time as
4951 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4952 it from "IPv6 and IPv4", which means that two separate sockets are being
4955 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4956 says whether AF_INET or AF_INET6 was passed as an argument.
4958 42. Exiwhat output was messed up when time zones were included in log
4961 43. Exiwhat now gives more information about the daemon's listening ports,
4962 and whether -tls-on-connect was used.
4964 44. The "port" option of the smtp transport is now expanded.
4966 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4967 silently ignored. Now an error message is written to the main and panic
4970 46. There's a new ACL modifier called "logwrite" which writes to a log file
4971 as soon as it is encountered.
4973 47. Added $local_user_uid and $local_user_gid at routing time.
4975 48. Exim crashed when trying to verify a sender address that was being
4978 49. Exim was recognizing only a space character after ".include". It now also
4979 recognizes a tab character.
4981 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4982 extracting the relevant information from the specification. The man page no
4983 longer contains scrambled data for the -d option, and I've added a section
4984 at the front about calling Exim under different names.
4986 51. Added "extra_headers" argument to the "mail" command in filter files.
4988 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4991 53. Installed eximstats 1.29.
4993 54. Added transport_filter_timeout as a generic transport option.
4995 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4996 Cc: header lines. This was required by RFC 822, but it not required by RFC
4999 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5000 incoming messages that did not have them. Now it does so only if the
5001 message originates locally, that is, if there is no associated remote host
5002 address. When Resent- header lines are present, this applies to the Resent-
5003 lines rather than the non-Resent- lines.
5005 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5006 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5008 58. Messages for configuration errors now include the name of the main
5009 configuration file - useful now that there may be more than one file in a
5010 list (.included file names were always shown).
5012 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5013 for those rare installations that do not start the daemon as root or run it
5014 setuid root. I've cut out the call to initgroups() if the daemon is not
5017 60. The Exim user and group can now be bound into the binary as text strings
5018 that are looked up at the start of Exim's processing.
5020 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5022 62. Added $mailstore_basename variable.
5024 63. Installed patch to sieve.c from Michael Haardt.
5026 64. When Exim failed to open the panic log after failing to open the main log,
5027 the original message it was trying to log was written to stderr and debug
5028 output, but if they were not available (the usual case in production), it
5029 was lost. Now it is written to syslog before the two lines that record the
5030 failures to open the logs.
5032 65. Users' Exim filters run in subprocesses under the user's uid. It is
5033 possible for a "deliver" command or an alias in a "personal" command to
5034 provoke an address rewrite. If logging of address rewriting is configured,
5035 this fails because the process is not running as root or exim. There may be
5036 a better way of dealing with this, but for the moment (because 4.30 needs
5037 to be released), I have disabled address rewrite logging when running a
5038 filter in a non-root, non-exim process.
5044 1. The buildconfig auxiliary program wasn't quoting the value set for
5045 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5046 not defined. This bug was masked in 4.22 by the effect that was fixed in
5049 2. Some messages that were rejected after a message id was allocated were
5050 shown as "incomplete" by exigrep. It no longer does this for messages that
5051 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5053 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5054 have allow_domain_literals set, the ID did not get logged in the <= line.
5055 Domain literals are now always recognized in Message-ID: header lines.
5057 4. The first argument for a ${extract expansion item is the key name or field
5058 number. Leading and trailing spaces in this item were not being ignored,
5059 causing some misleading effects.
5061 5. When deliver_drop_privilege was set, single queue runner processes started
5062 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5063 same command in the process it spins off) were not dropping privilege.
5065 6. When the daemon running as "exim" started a queue runner, it always
5066 re-executed Exim in the spun-off process. This is a waste of effort when
5067 deliver_drop_privilege is set. The new process now just calls the
5068 queue-runner function directly.
5074 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5077 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5078 ignored. Though the use of -oP was forcing the writing of a pid file, it
5079 was always written to the default place.
5081 3. If the message "no IP address found for host xxxx" is generated during
5082 incoming verification, it is now followed by identification of the incoming
5083 connection (so you can more easily find what provoked it).
5085 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5087 5. Added some features to "harden" Exim a bit more against certain attacks:
5089 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5090 be put in Local/Makefile. This is like the never_users runtime option,
5091 but it cannot be overridden. The default setting is "root".
5093 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5094 prefix string with which any file named in a -C command line option
5097 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5098 is retained for -C and -D only if the caller of Exim is root. Without
5099 it, the exim user may also use -C and -D and retain privilege.
5101 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5102 command line option is disabled.
5104 6. Macro names set by the -D option must start with an upper case letter, just
5105 like macro names defined in the configuration file.
5107 7. Added "dereference=" facility to LDAP.
5109 8. Two instances of the typo "uknown" in the source files are fixed.
5111 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5112 the Configure-Makefile script screwed up while processing it.
5114 10. Incorporated PCRE 4.4.
5116 11. The SMTP synchronization check was not operating right at the start of an
5117 SMTP session. For example, it could not catch a HELO sent before the client
5118 waited for the greeting. There is now a check for outstanding input at the
5119 point when the greeting is written. Because of the duplex, asynchronous
5120 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5121 way, but not yet received, when the check is performed.
5123 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5124 on TCP/IP sockets, because this apparently causes some broken clients to
5127 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5128 unchanged) from the Cygwin maintainer.
5130 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5131 was supported instead of testing for mbx. Effectively a typo.
5133 15. The spa authenticator server code was not checking that the input it
5134 received was valid base64.
5136 16. The debug output line for the "set" modifier in ACLs was not showing the
5137 name of the variable that was being set.
5139 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5141 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5142 lookup on its own. Something else was needed to trigger the lookup. For
5143 example, a match in host_lookup or the need for a host name in a host list.
5144 Now, if $sender_host_name is referenced and the host name has not yet been
5145 looked up, a lookup is performed. If the lookup fails, the variable remains
5146 empty, and $host_lookup_failed is set to "1".
5148 19. Added "eqi" as a case-independent comparison operator.
5150 20. The saslauthd authentication condition could segfault if neither service
5151 nor realm was specified.
5153 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5154 error message that was logged was misleading, and incoming SMTP
5155 connections were dropped. The message is now more accurate, and temporary
5156 errors are given to SMTP connections.
5158 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5159 (except RSET) with a 421 error, until QUIT is received. However, it was
5160 failing to send a response to QUIT.
5162 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5163 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5164 ACL is now run after the helo_try_verify_hosts code.
5166 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5167 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5168 case-independent, so other case variants are also recognized). Apparently
5169 some systems use these upper case variants.
5171 25. If more than two messages were waiting for the same host, and a transport
5172 filter was specified for the transport, Exim sent two messages over the
5173 same TCP/IP connection, and then failed with "socket operation on non-
5174 socket" when it tried to send the third.
5176 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5178 27. The extern definition of crypt16() in expand.c was not being excluded when
5179 the OS had its own crypt16() function.
5181 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5182 as a preferred synonym for return_size_limit, both as an option and as an
5185 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5187 30. Changed the default configuration ACL to relax the local part checking rule
5188 for addresses that are not in any local domains. For these addresses,
5189 slashes and pipe symbols are allowed within local parts, but the sequence
5190 /../ is explicitly forbidden.
5192 31. SPA server authentication was not clearing the challenge buffer before
5195 32. log_message in a "warn" ACL statement was writing to the reject log as
5196 well as to the main log, which contradicts the documentation and doesn't
5197 seem right (because no rejection is happening). So I have stopped it.
5199 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5200 However, I am unable to do any testing of this.
5202 34. Fixed an infelicity in the appendfile transport. When checking directories
5203 for a mailbox, to see if any needed to be created, it was accidentally
5204 using path names with one or more superfluous leading slashes; tracing
5205 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5207 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5208 contents are added to the log line that is written for every discarded
5209 recipient. (Previously a log_message setting was ignored.)
5211 36. The ${quote: operator now quotes the string if it is empty.
5213 37. The install script runs exim in order to find its version number. If for
5214 some reason other than non-existence or emptiness, which it checks, it
5215 could not run './exim', it was installing it with an empty version number,
5216 i.e. as "exim-". This error state is now caught, and the installation is
5219 38. An argument was missing from the function that creates an error message
5220 when Exim fails to connect to the socket for saslauthd authentication.
5221 This could cause Exim to crash, or give a corrupted message.
5223 39. Added isip, isip4, and isip6 to ${if conditions.
5225 40. The ACL variables $acl_xx are now saved with the message, and can be
5226 accessed later in routers, transports, and filters.
5228 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5229 strings in the file are not string-expanded.
5231 42. If a MAIL command specified a SIZE value that was too large to fit into an
5232 int variable, the check against message_size_limit failed. Such values are
5233 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5234 day this will have to be increased, but I don't think I want to be around
5235 when emails are that large.
5242 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5243 iconv() is not standard in FreeBSD.
5245 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5246 IPv6 enabled. The observed symptom was a segmentation fault on return from
5247 the function os_common_find_running_interfaces() in src/os.c.
5249 3. In the check_special_case() function in daemon.c I had used "errno" as an
5250 argument name, which causes warnings on some systems. This was basically a
5251 typo, since it was named "eno" in the comments!
5253 4. The code that waits for the clock to tick (at a resolution of some fraction
5254 of a second) so as to ensure message-id uniqueness was always waiting for
5255 at least one whole tick, when it could have waited for less. [This is
5256 almost certainly not relevant at current processor speeds, where it is
5257 unlikely to ever wait at all. But we try to future-proof.]
5259 5. The function that sleeps for a time interval that includes fractions of a
5260 second contained a race. It did not block SIGALRM between setting the
5261 timer, and suspending (a couple of lines later). If the interval was short
5262 and the sigsuspend() was delayed until after it had expired, the suspension
5263 never ended. On busy systems this could lead to processes getting stuck for
5266 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5267 process, before it forks any delivery processes. The open lookup caching
5268 mechanism meant that the open file or database connection was passed into
5269 the delivery process. The problem was that delivery processes always tidy
5270 up cached lookup data. This could cause a problem for the next delivery
5271 process started by the queue runner, because the external queue runner
5272 process does not know about the closure. So the next delivery process
5273 still has data in the lookup cache. In the case of a file lookup, there was
5274 no problem because closing a file descriptor in a subprocess doesn't affect
5275 the parent. However, if the lookup was caching a connection to a database,
5276 the connection was closed, and the second delivery process was likely to
5277 see errors such as "PGSQL: query failed: server closed the connection
5278 unexpectedly". The problem has been fixed by closing all cached lookups
5279 in a queue runner before running a delivery process.
5281 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5282 seem to have the "const" qualifier which it has on other OS. I've
5285 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5286 *of the same type* (client or server) with the same public name that an
5287 error should be diagnosed.
5289 9. When Exim looked up a host name for an IP address, but failed to find the
5290 original IP address when looking up the host name (a safety check), it
5291 output the message "<ip address> does not match any IP for NULL", which was
5292 confusing, to say the least. The bug was that the host name should have
5293 appeared instead of "NULL".
5295 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5296 user that is built into the binary, and the -C or -D options is used, root
5297 privilege is dropped before the configuration file is read. In addition,
5298 logging is switched to stderr instead of the normal log files. If the
5299 configuration then re-defines the Exim user, the unprivileged environment
5300 is probably not what is expected, so Exim logs a panic warning message (but
5303 However, if deliver_drop_privilege is set, the unprivileged state may well
5304 be exactly what is intended, so the warning has been cut out in that case,
5305 and Exim is allowed to try to write to its normal log files.
5311 1. smtp_return_error_details was not giving details for temporary sender
5312 or receiver verification errors.
5314 2. Diagnose a configuration error if two authenticators have the same public
5317 3. Exim used not to create the message log file for a message until the first
5318 delivery attempt. This could be confusing when incoming messages were held
5319 for policy or load reasons. The message log file is now created at the time
5320 the message is received, and an initial "Received" line is written to it.
5322 4. The automatically generated man page for command line options had a minor
5323 bug that caused no ill effects; however, a more serious problem was that
5324 the procedure for building the man page automatically didn't always
5325 operate. Consequently, release 4.20 contains an out-of-date version. This
5326 shouldn't happen again.
5328 5. When building Exim with embedded Perl support, the script that builds the
5329 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5330 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5332 6. The freeze_tell option was not being used for messages that were frozen on
5333 arrival, either by an ACL or by local_scan().
5335 7. Added the smtp_incomplete_transaction log selector.
5337 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5338 was accepting AUTH without a new EHLO.
5340 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5341 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5344 10. Logging of TCP/IP connections (when configured) now happens in the main
5345 daemon process instead of the child process, so that the TCP/IP connection
5346 count is more accurate (but it can never be perfect).
5348 11. The use of "drop" in a nested ACL was not being handled correctly in the
5349 outer ACL. Now, if condition failure induced by the nested "drop" causes
5350 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5351 or "require"), the connection is dropped.
5353 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5354 that yield "discard" can now be used with an "accept" or a "discard" verb,
5355 but an error is generated for any others (because I can't see a useful way
5356 to define what should happen).
5358 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5359 are now processed in the same way as lines in the Exim configuration file.
5360 In particular, continuation lines are supported.
5362 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5364 15. Added -ti meaning -t -i.
5366 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5367 domains, and warn by logging if others are found.
5369 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5370 returned from GIFCONF when Exim is trying to find the list of interfaces on
5371 a host. The code in os.c has been modified to copy each ifreq to an aligned
5372 structure in all cases.
5374 Also, in some cases, the returned ifreq's were being copied to a 'struct
5375 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5376 means the last couple of bytes of an IPv6 address could be chopped if the
5377 ifreq contained only a normal sockaddr (14 bytes storage).
5379 18. Named domain lists were not supported in the hosts_treat_as_local option.
5380 An entry such as +xxxx was not recognized, and was treated as a literal
5383 19. Ensure that header lines added by a DATA ACL are included in the reject log
5384 if the ACL subsequently rejects the message.
5386 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5387 MD5 (which is deprecated).
5389 21. When testing a filter file using -bf, Exim was writing a message when it
5390 took the sender from a "From " line in the message, but it was not doing so
5391 when it took $return_path from a Return-Path: header line. It now does.
5393 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5394 with a valid header line field name (a series of printing characters
5395 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5397 23. Changed "disc" in the source to "disk" to conform to the documentation and
5398 the book and for uniformity.
5400 24. Ignore Sendmail's -Ooption=value command line item.
5402 25. When execve() failed while trying to run a command in a pipe transport,
5403 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5404 could be confused with a return value of 69 from the command itself. This
5405 has been changed to 127, the value the shell returns if it is asked to run
5406 a non-existent command. The wording for the related log line suggests a
5407 non-existent command as the problem.
5409 26. If received_header_text expands to an empty string, do not add a Received:
5410 header line to the message. (Well, it adds a token one on the spool, but
5411 marks it "old" so that it doesn't get used or transmitted.)
5413 27. Installed eximstats 1.28 (addition of -nt option).
5415 28. There was no check for failure on the call to getsockname() in the daemon
5416 code. This can fail if there is a shortage of resources on the system, with
5417 ENOMEM, for example. A temporary error is now given on failure.
5419 29. Contrary to the C standard, it seems that in some environments, the
5420 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5421 program. Exim now does this explicitly; it affects the formatting of
5422 timestamps using strftime().
5424 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5425 complaints. I've now initialized all the variables, to avoid this.
5427 32. Header lines added by a system filter were not being "seen" during
5428 transport-time rewrites.
5430 33. The info_callback() function passed to OpenSSL is set up with type void
5431 (*)(SSL *, int, int), as described somewhere. However, when calling the
5432 function (actually a macro) that sets it up, the type void(*)() is
5433 expected. I've put in a cast to prevent warnings from picky compilers.
5435 34. If a DNS black list lookup found a CNAME record, but there were no A
5436 records associated with the domain it pointed at, Exim crashed.
5438 35. If a DNS black list lookup returned more than one A record, Exim ignored
5439 all but the first. It now scans all returned addresses if a particular IP
5440 value is being sought. In this situation, the contents of the
5441 $dnslist_value variable are a list of all the addresses, separated by a
5444 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5445 to accept a host name and all its aliases if the forward lookup for any of
5446 them yielded the IP address of the incoming connection. Now it accepts only
5447 those names whose forward lookup yields the correct IP address. Any other
5448 names are discarded. This closes a loophole whereby a rogue DNS
5449 administrator could create reverse DNS records to break through a
5450 wildcarded host restriction in an ACL.
5452 37. If a user filter or a system filter that ran in a subprocess used any of
5453 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5454 the wrong values were passed to the pipe command ($thisaddress had the
5455 value of $0, $0 had the value of $1, etc). This bug was introduced by
5456 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5458 38. Improved the line breaking for long SMTP error messages from ACLs.
5459 Previously, if there was no break point between 40 and 75 characters, Exim
5460 left the rest of the message alone. Two changes have been made: (a) I've
5461 reduced the minimum length to 35 characters; (b) if it can't find a break
5462 point between 35 and 75 characters, it looks ahead and uses the first one
5463 that it finds. This may give the occasional overlong line, but at least the
5464 remaining text gets split now.
5466 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5467 file descriptors might be low, and that setting 1000 would always raise it.
5468 It turns out that in some environments, the limit is already over 1000 and
5469 that lowering it causes trouble. So now Exim takes care not to decrease it.
5471 40. When delivering a message, the value of $return_path is set to $sender_
5472 address at the start of routing (routers may change the value). By an
5473 oversight, this default was not being set up when an address was tested by
5474 -bt or -bv, which affected the outcome if any router or filter referred to
5477 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5478 log only when "message" or "log_message" are set. However, if one of the
5479 conditions was an address verification, or a call to a nested ACL, the
5480 messages generated by the underlying test were being passed through. This
5481 no longer happens. The underlying message is available in $acl_verify_
5482 message for both "message" and "log_message" expansions, so it can be
5483 passed through if needed.
5485 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5486 new expansion $bh_ to give the encoded byte string without charset
5487 translation. Translation happens only if iconv() is available; HAVE_ICONV
5488 indicates this at build time. HEADERS_CHARSET gives the charset to
5489 translate to; headers_charset can change it in the configuration, and
5490 "headers charset" can change it in an individual filter file.
5492 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5493 that creates RFC 2047 encoded "words" labels them as that charset instead
5494 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5495 expansion operator; (ii) when Exim creates a From: line for a local
5496 message; (iii) when a header line is rewritten to include a "phrase" part.
5498 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5499 buggy, causing it to skip the first lines of messages whose message ID
5500 ended in 'D'. This would not have bitten before Exim release 4.14, because
5501 message IDs were unlikely to end in 'D' before then. The effect was to have
5502 incorrect size information for certain domains.
5504 45. #include "config.h" was missing at the start of the crypt16.c module. This
5505 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5508 46. If there was a timeout during a "random" callout check, Exim treated it as
5509 a failure of the random address, and carried on sending RSET and the real
5510 address. If the delay was just some slowness somewhere, the response to the
5511 original RCPT would be taken as a response to RSET and so on, causing
5512 mayhem of various kinds.
5514 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5515 when I implemented it. It didn't allow for the fact that some option values
5516 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5517 the right test for positive values.
5519 48. Domain names in DNS records are case-independent. Exim always looks them up
5520 in lower case. Some resolvers return domain names in exactly the case they
5521 appear in the zone file, that is, they may contain uppercase letters. Not
5522 all resolvers do this - some return always lower case. Exim was treating a
5523 change of case by a resolver as a change of domain, similar to a widening
5524 of a domain abbreviation. This triggered its re-routing code and so it was
5525 trying to route what was effectively the same domain again. This normally
5526 caused routing to fail (because the router wouldn't handle the domain
5527 twice). Now Exim checks for this case specially, and just changes the
5528 casing of the domain that it ultimately uses when it transmits the message
5531 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5534 50. If a filter generated a file delivery with a non-absolute name (possible if
5535 no home directory exists for the router), the forbid_file option was not
5538 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5539 the existing equality matching.
5541 52. Exim was using ints instead of ino_t variables in some places where it was
5542 dealing with inode numbers.
5544 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5545 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5546 called TMPDIR, and if it finds it is different, it changes its value.
5548 54. The smtp_printf() function is now made available to local_scan() so
5549 additional output lines can be written before returning. There is also an
5550 smtp_fflush() function to enable the detection of a dropped connection.
5551 The variables smtp_input and smtp_batched_input are exported to
5554 55. Changed the default runtime configuration: the message "Unknown user"
5555 has been removed from the ACL, and instead placed on the localuser router,
5556 using the cannot_route_message feature. This means that any verification
5557 failures that generate their own messages won't get overridden. Similarly,
5558 the "Unrouteable address" message that was in the ACL for unverifiable
5559 relay addresses has also been removed.
5561 56. Added hosts_avoid_esmtp to the smtp transport.
5563 57. The exicyclog script was not checking for the esoteric option
5564 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5565 will work only if exicyclog is run under the appropriate euid.
5567 58. Following a discussion on the list, the rules by which Exim recognises line
5568 endings on incoming messages have been changed. The -dropcr and drop_cr
5569 options are now no-ops, retained only for backwards compatibility. The
5570 following line terminators are recognized: LF CRLF CR. However, special
5571 processing applies to CR:
5573 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5574 nor a local message in the state where . is a terminator.
5576 (ii) If a bare CR is encountered in a header line, an extra space is added
5577 after the line terminator so as not to end the header. The reasoning
5578 behind this is that bare CRs in header lines are most likely either
5579 to be mistakes, or people trying to play silly games.
5581 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5582 was being incorrectly given as 18 bytes larger than it should have been.
5583 This is a VOB (very old bug).
5585 60. This may never have affected anything current, but just in case it has:
5586 When the local host is found other than at the start of a list of hosts,
5587 the local host, those with the same MX, and any that follow, are discarded.
5588 When the list in question was part of a longer list of hosts, the following
5589 hosts (not currently being processed) were also being discarded. This no
5590 longer happens. I'm not sure if this situation could ever has previously
5593 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5596 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5597 header. This is something that is recommended in a new Internet Draft, and
5598 is something that is documented as being done by Sendmail. There are two
5599 possible values. For messages generated by the autoreply transport, Exim
5602 Auto-Submitted: auto-replied
5604 whereas for all other generated messages (e.g. bounces) it adds
5606 Auto-Submitted: auto-generated
5608 63. The "personal" condition in filters now includes a test for the
5609 Auto-Submitted: header. If it contains the string "auto-" the message it
5610 not considered personal.
5612 64. Added rcpt_include_affixes as a generic transport option.
5614 65. Added queue_only_override (default true).
5616 66. Added the syslog_duplication option.
5618 67. If what should have been the first header line of a message consisted of
5619 a space followed by a colon, Exim was mis-interpreting it as a header line.
5620 It isn't of course - it is syntactically invalid and should therefore be
5621 treated as the start of the message body. The misbehaviour could have
5622 caused a number of strange effects, including loss of data in subsequent
5623 header lines, and spool format errors.
5625 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5626 client host had authenticated. This control can now be exercised by an ACL
5627 for more flexibility.
5629 69. By default, callouts do not happen when testing with -bh. There is now a
5630 variant, -bhc, which does actually run the callout code, including
5631 consulting and updating the callout cache.
5633 70. Added support for saslauthd authentication, courtesy of Alexander
5636 71. If statvfs() failed on the spool or log directories while checking their
5637 size for availability, Exim confusingly gave the error "space shortage".
5638 Furthermore, in debugging mode it crashed with a floating point exception.
5639 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5640 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5641 serious problem, Exim now writes to the main and panic logs when this
5642 happens, with details of the failure. It then refuses to accept the
5643 incoming message, giving the message "spool directory problem" or "log
5644 directory problem" with a 421 code for SMTP messages.
5646 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5647 0, 1, and 2 exist, because some OS complain for execs without them (see
5648 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5649 descriptors. However, the code omitted to check that the open succeeded,
5650 causing mysterious errors if for some reason the permissions on /dev/null
5651 got screwed. Now Exim writes a message to the main and panic logs, and
5652 bombs out if it can't open /dev/null.
5654 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5655 interact so that it is all more flexible. It is supposed to remain
5656 backwards compatible. Also added extra_local_interfaces.
5658 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5659 to bomb out with an assertion failure - to the client this appears as a
5660 connection drop. This problem occurs in the part of the code that was taken
5661 from the Samba project. Fortunately, the assertion is in a very simple
5662 function, so I have fixed this by reproducing the function inline in the
5663 one place where it is called, and arranging for authentication to fail
5664 instead of killing the process with assert().
5666 75. The SPA client code was not working when the server requested OEM rather
5667 than Unicode encoding.
5669 76. Added code to make require_files with a specific uid setting more usable in
5670 the case where statting the file as root fails - usually a non-root-mounted
5671 NFS file system. When this happens and the failure is EACCES, Exim now
5672 forks a subprocess and does the per-uid checking as the relevant uid.
5674 77. Added process_log_path.
5676 78. If log_file_path was not explicitly set, a setting of check_log_space or
5677 check_log_inodes was ignored.
5679 79. If a space check for the spool or log partitions fails, the incident is now
5680 logged. Of course, in the latter case the data may get lost...
5682 80. Added the %p formatting code to string_format() so that it can be used to
5683 print addresses in debug_print(). Adjusted all the address printing in the
5684 debugging in store.c to use %p rather than %d.
5686 81. There was a concern that a line of code in smtp_in.c could overflow a
5687 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5688 initially expressed, the concern was not well-founded, because trailing
5689 spaces are removed early. However, if the trailing spaces were followed by
5690 a NULL, they did not get removed, so the overflow was possible. Two fixes
5693 (a) I re-wrote the offending code in a cleaner fashion.
5694 (b) If an incoming SMTP command contains a NULL character, it is rejected
5697 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5698 runs initgroups(), so that if the Exim user is in any additional groups,
5699 they will be used during message reception.
5705 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob