1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
46 TL/01 When running the test suite, the README says that variables such as
47 no_msglog_check are global and can be placed anywhere in a specific
48 test's script, however it was observed that placement needed to be near
49 the beginning for it to behave that way. Changed the runtest perl
50 script to read through the entire script once to detect and set these
51 variables, reset to the beginning of the script, and then run through
52 the script parsing/test process like normal.
54 TL/02 The BSD's have an arc4random API. One of the functions to induce
55 adding randomness was arc4random_stir(), but it has been removed in
56 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
57 function when detected.
59 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
60 cause callback expansion.
62 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
63 syntax errors in an expansion can be treated as a string instead of
64 logging or causing an error, due to the internal use of bool_lax
65 instead of bool when processing it.
67 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
68 server certificates when making smtp deliveries.
70 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
72 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
74 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
76 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
77 Merged patch from Sebastian Wiedenroth.
79 JH/05 Fix results-pipe from transport process. Several recipients, combined
80 with certificate use, exposed issues where response data items split
81 over buffer boundaries were not parsed properly. This eventually
82 resulted in duplicates being sent. This issue only became common enough
83 to notice due to the introduction of conection certificate information,
84 the item size being so much larger. Found and fixed by Wolfgang Breyha.
86 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
87 size buffer was used, resulting in syntax errors when an expansion
90 JH/07 Add support for directories of certificates when compiled with a GnuTLS
91 version 3.3.6 or later.
93 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
94 is EXPERIMENTAL_EVENT, the main-configuration and transport options
95 both become "event_action", the variables become $event_name, $event_data
96 and $event_defer_errno. There is a new variable $verify_mode, usable in
97 routers, transports and related events. The tls:cert event is now also
98 raised for inbound connections, if the main configuration event_action
101 TL/06 In test suite, disable OCSP for old versions of openssl which contained
102 early OCSP support, but no stapling (appears to be less than 1.0.0).
104 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
105 server certificate names available under the smtp transport option
106 "tls_verify_cert_hostname" now do not permit multi-component wildcard
109 JH/10 Time-related extraction expansions from certificates now use the main
110 option "timezone" setting for output formatting, and are consistent
111 between OpenSSL and GnuTLS compilations. Bug 1541.
113 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
116 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
117 include certificate info, eximon was claiming there were spoolfile
120 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
122 JH/14 Log delivery-related information more consistently, using the sequence
123 "H=<name> [<ip>]" wherever possible.
128 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
129 checkers that were complaining about end of non-void function with no
132 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
133 This was a regression intruduced in 4.83 by another bugfix.
135 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
137 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
138 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
144 TF/01 Correctly close the server side of TLS when forking for delivery.
146 When a message was received over SMTP with TLS, Exim failed to clear up
147 the incoming connection properly after forking off the child process to
148 deliver the message. In some situations the subsequent outgoing
149 delivery connection happened to have the same fd number as the incoming
150 connection previously had. Exim would try to use TLS and fail, logging
151 a "Bad file descriptor" error.
153 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
154 utilities have not been installed.
156 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
157 temporary space as the ACL may create new global variables.
159 TL/01 LDAP support uses per connection or global context settings, depending
160 upon the detected version of the libraries at build time.
162 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
163 to extract and use the src ip:port in logging and expansions as if it
164 were a direct connection from the outside internet. PPv2 support was
165 updated based on HAProxy spec change in May 2014.
167 JH/02 Add ${listextract {number}{list}{success}{fail}}.
169 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
170 Properly escape header and check for NULL return.
172 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
175 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
177 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
178 characters in header names, implemented as a verify condition.
179 Contributed by Michael Fischer v. Mollard.
181 TL/05 Rename SPF condition results err_perm and err_temp to standardized
182 results permerror and temperror. Previous values are deprecated but
183 still accepted. In a future release, err_perm and err_temp will be
184 completely removed, which will be a backward incompatibility if the
185 ACL tests for either of these two old results. Patch contributed by
186 user bes-internal on the mailing list.
188 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
190 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
191 selectors, in both main and reject logs.
193 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
196 JH/07 Add malware type "sock" for talking to simple daemon.
198 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
200 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
201 routers/transports under cutthrough routing.
203 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
204 numbers. Touch up "bool" conditional to keep the same definition.
206 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
208 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
210 JH/12 Expand items in router/transport headers_add or headers_remove lists
211 individually rather than the list as a whole. Bug 1452.
213 Required for reasonable handling of multiple headers_ options when
214 they may be empty; requires that headers_remove items with embedded
215 colons must have them doubled (or the list-separator changed).
217 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
218 view the policy declared in the DMARC record. Currently, $dmarc_status
219 is a combined value of both the record presence and the result of the
222 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
224 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
225 dnslookup router and the smtp transport (applying to the forward
228 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
229 of ldap servers used for a specific lookup. Patch provided by Heiko
232 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
233 New variable $lookup_dnssec_authenticated for observability.
235 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
236 Patch submitted by Lars Timman.
238 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
240 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
241 Requires trusted mode and valid format message id, aborts otherwise.
242 Patch contributed by Heiko Schlichting.
244 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
245 certextract with support for various fields. Bug 1358.
247 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
248 is requested by default, modifiable by smtp transport option
251 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
252 operate on certificate variables to give certificate fingerprints
253 Also new ${sha256:cert_variable}.
255 JH/23 The PRDR feature is moved from being Experimental into the mainline.
257 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
258 Christian Aistleitner.
260 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
262 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
263 file. Patch from Wolfgang Breyha.
265 JH/25 Expand the coverage of the delivery $host and $host_address to
266 client authenticators run in verify callout. Bug 1476.
268 JH/26 Port service names are now accepted for tls_on_connect_ports, to
269 align with daemon_smtp_ports. Bug 72.
271 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
272 support and error reporting did not work properly.
274 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
275 and is readable. Patch from Andrew Colin Kissa.
277 TL/14 Enhance documentation of ${run expansion and how it parses the
278 commandline after expansion, particularly in the case when an
279 unquoted variable expansion results in an empty value.
281 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
283 PP/02 Fix internal collision of T_APL on systems which support RFC3123
284 by renaming away from it. Addresses GH issue 15, reported by
287 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
289 TL/15 SECURITY: prevent double expansion in math comparison functions
290 (can expand unsanitized data). Not remotely exploitable.
297 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
299 PP/02 Make -n do something, by making it not do something.
300 When combined with -bP, the name of an option is not output.
302 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
305 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
306 $sender_host_name and config options to manage this, and basic check
309 PP/05 DSCP support for outbound connections and control modifier for inbound.
311 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
312 (Only plugin which currently uses this is kerberos4, which nobody should
313 be using, but we should make it available and other future plugins might
314 conceivably use it, even though it would break NAT; stuff *should* be
315 using channel bindings instead).
317 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
318 name; added for Sendmail compatibility; requires admin caller.
319 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
320 just ignore it); requires trusted caller.
321 Also parse but ignore: -Ac -Am -X<logfile>
324 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
326 TL/02 Add +smtp_confirmation as a default logging option.
328 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
329 Patch by Magnus Holmgren from 2007-02-20.
331 TL/04 Bugzilla 1281 - Spec typo.
332 Bugzilla 1283 - Spec typo.
333 Bugzilla 1290 - Spec grammar fixes.
335 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
337 TL/06 Add Experimental DMARC support using libopendmarc libraries.
339 TL/07 Fix an out of order global option causing a segfault. Reported to dev
340 mailing list by by Dmitry Isaikin.
342 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
344 JH/02 Support "G" suffix to numbers in ${if comparisons.
346 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
348 NM/01 Bugzilla 1197 - Spec typo
349 Bugzilla 1196 - Spec examples corrections
351 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
353 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
354 gnutls_enable_pkcs11, but renamed to more accurately indicate its
357 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
358 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
360 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
361 "acl {{name}{arg}...}", and optional args on acl condition
364 JH/05 Permit multiple router/transport headers_add/remove lines.
366 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
368 JH/07 Avoid using a waiting database for a single-message-only transport.
369 Performance patch from Paul Fisher. Bugzilla 1262.
371 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
374 JH/09 Add $headers_added variable, with content from use of ACL modifier
375 add_header (but not yet added to the message). Bugzilla 199.
377 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
378 Pulled from Bugzilla 817 by Wolfgang Breyha.
380 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
382 (nb: this is the same fix as in Exim 4.80.1)
384 JH/11 Add A= logging on delivery lines, and a client_set_id option on
387 JH/12 Add optional authenticated_sender logging to A= and a log_selector
390 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
392 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
393 advertise SMTP AUTH mechanism to us, instead of a generic
394 protocol violation error. Also, make Exim more robust to bad
395 data from the Dovecot auth socket.
397 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
399 When a queue runner is handling a message, Exim first routes the
400 recipient addresses, during which it prunes them based on the retry
401 hints database. After that it attempts to deliver the message to
402 any remaining recipients. It then updates the hints database using
405 So if a recipient address works intermittently, it can get repeatedly
406 deferred at routing time. The retry hints record remains fresh so the
407 address never reaches the final cutoff time.
409 This is a fairly common occurrence when a user is bumping up against
410 their storage quota. Exim had some logic in its local delivery code
411 to deal with this. However it did not apply to per-recipient defers
412 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
414 This change adds a proper retry rule check during routing so that the
415 final cutoff time is checked against the message's age. We only do
416 this check if there is an address retry record and there is not a
417 domain retry record; this implies that previous attempts to handle
418 the address had the retry_use_local_parts option turned on. We use
419 this as an approximation for the destination being like a local
420 delivery, as in LMTP.
422 I suspect this new check makes the old local delivery cutoff check
423 redundant, but I have not verified this so I left the code in place.
425 TF/02 Correct gecos expansion when From: is a prefix of the username.
427 Test 0254 submits a message to Exim with the header
431 When I ran the test suite under the user fanf2, Exim expanded
432 the header to contain my full name, whereas it should have added
433 a Resent-Sender: header. It erroneously treats any prefix of the
434 username as equal to the username.
436 This change corrects that bug.
438 GF/01 DCC debug and logging tidyup
439 Error conditions log to paniclog rather than rejectlog.
440 Debug lines prefixed by "DCC: " to remove any ambiguity.
442 TF/03 Avoid unnecessary rebuilds of lookup-related code.
444 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
445 Bug spotted by Jeremy Harris; was flawed since initial commit.
446 Would have resulted in OCSP responses post-SNI triggering an Exim
447 NULL dereference and crash.
449 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
451 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
452 Bug detection, analysis and fix by Samuel Thibault.
453 Bugzilla 1331, Debian bug #698092.
455 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
457 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
458 Server implementation by Todd Lyons, client by JH.
459 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
460 config variable "prdr_enable" controls whether the server
461 advertises the facility. If the client requests PRDR a new
462 acl_data_smtp_prdr ACL is called once for each recipient, after
463 the body content is received and before the acl_smtp_data ACL.
464 The client is controlled by bolth of: a hosts_try_prdr option
465 on the smtp transport, and the server advertisement.
466 Default client logging of deliveries and rejections involving
467 PRDR are flagged with the string "PRDR".
469 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
470 fclose(). Diagnosis by Todd Lyons.
472 PP/17 Update configure.default to handle IPv6 localhost better.
473 Patch by Alain Williams (plus minor tweaks).
476 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
477 This is now consistent with GnuTLS, and is now documented: the
478 previous undocumented portable approach to treating the option as
479 unset was to force an expansion failure. That still works, and
480 an empty string is now equivalent.
482 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
483 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
484 not performing validation itself.
486 PP/20 Added force_command boolean option to pipe transport.
487 Patch from Nick Koston, of cPanel Inc.
489 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
492 TF/04 Added udpsend ACL modifer and hexquote expansion operator
494 PP/21 Fix eximon continuous updating with timestamped log-files.
495 Broken in a format-string cleanup in 4.80, missed when I repaired the
496 other false fix of the same issue.
497 Report and fix from Heiko Schlichting.
500 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
501 Report from Prashanth Katuri.
503 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
504 It's SecureTransport, so affects any MacOS clients which use the
505 system-integrated TLS libraries, including email clients.
507 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
508 using a MIME ACL for non-SMTP local injection.
509 Report and assistance in diagnosis by Warren Baker.
511 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
513 JH/16 Fix comparisons for 64b. Bugzilla 1385.
515 TL/09 Add expansion variable $authenticated_fail_id to keep track of
516 last id that failed so it may be referenced in subsequent ACL's.
518 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
521 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
522 ldap library initialization, allowing self-signed CA's to be
523 used. Also properly sets require_cert option later in code by
524 using NULL (global ldap config) instead of ldap handle (per
525 session). Bug diagnosis and testing by alxgomz.
527 TL/12 Enhanced documentation in the ratelimit.pl script provided in
528 the src/util/ subdirectory.
530 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
531 renamed to Transport Post Delivery Action by Jeremy Harris, as
534 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
535 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
536 redis_servers = needs to be configured which will be used by the redis
537 lookup. Patch from Warren Baker, of The Packet Hub.
539 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
541 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
542 hostname or reverse DNS when processing a host list. Used suggestions
543 from multiple comments on this bug.
545 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
547 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
548 Missed a few lines, added it to make the runtest require no keyboard
551 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
552 contains upper case chars. Make router use caseful_local_part.
554 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
555 support when GnuTLS has been built with p11-kit.
561 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
563 This, or similar/improved, will also be change PP/11 of 4.82.
569 PP/01 Handle short writes when writing local log-files.
570 In practice, only affects FreeBSD (8 onwards).
571 Bugzilla 1053, with thanks to Dmitry Isaikin.
573 NM/01 Bugzilla 949 - Documentation tweak
575 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
578 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
580 PP/02 Implemented gsasl authenticator.
582 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
584 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
585 `pkg-config foo` for cflags/libs.
587 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
588 with rest of GSASL and with heimdal_gssapi.
590 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
591 `pkg-config foo` for cflags/libs for the TLS implementation.
593 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
594 properties get this fed in as external SSF. A number of robustness
595 and debugging improvements to the cyrus_sasl authenticator.
597 PP/08 cyrus_sasl server now expands the server_realm option.
599 PP/09 Bugzilla 1214 - Log authentication information in reject log.
600 Patch by Jeremy Harris.
602 PP/10 Added dbmjz lookup type.
604 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
606 PP/12 MAIL args handles TAB as well as SP, for better interop with
607 non-compliant senders.
608 Analysis and variant patch by Todd Lyons.
610 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
611 Bug report from Lars Müller <lars@samba.org> (via SUSE),
612 Patch from Dirk Mueller <dmueller@suse.com>
614 PP/13 tls_peerdn now print-escaped for spool files.
615 Observed some $tls_peerdn in wild which contained \n, which resulted
616 in spool file corruption.
618 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
619 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
620 or write after TLS renegotiation, which otherwise led to messages
623 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
624 as a tracking header (ie: a signed header comes before the signature).
625 Patch from Wolfgang Breyha.
627 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
628 comma-sep list; embedded commas doubled.
630 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
632 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
634 Report and patch from Dmitry Banschikov.
636 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
637 Removed SSL_clear() after SSL_new() which led to protocol negotiation
638 failures. We appear to now support TLS1.1+ with Exim.
640 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
641 lets Exim select keys and certificates based upon TLS SNI from client.
642 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
643 before an outbound SMTP session. New log_selector, +tls_sni.
645 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
646 NULL dereference. Report and patch from Alun Jones.
648 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
649 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
650 Not seeing resolver debug output on NetBSD, but suspect this is a
651 resolver implementation change.
653 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
654 Left warnings. Added "eximon gdb" invocation mode.
656 PP/21 Defaulting "accept_8bitmime" to true, not false.
658 PP/22 Added -bw for inetd wait mode support.
660 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
661 locate the relevant includes and libraries. Made this the default.
663 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
664 Bugzilla 1246, report and most of solution from Tomasz Kusy.
666 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
667 This may cause build issues on older platforms.
669 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
670 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
671 gnutls_require_mac & gnutls_require_protocols (no longer supported).
672 Added SNI support via GnuTLS too.
673 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
675 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
677 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
678 Applied second patch from Janne, implementing suggestion to default
679 multiple-strings-in-record handling to match SPF spec.
681 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
683 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
684 read-only, out of scope).
685 Patch from Wolfgang Breyha, report from Stuart Northfield.
687 PP/29 Fix three issues highlighted by clang analyser static analysis.
688 Only crash-plausible issue would require the Cambridge-specific
689 iplookup router and a misconfiguration.
690 Report from Marcin Mirosław.
692 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
694 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
695 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
696 As part of this, removing so much warning spew let me fix some minor
697 real issues in debug logging.
699 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
700 assignment on my part. Fixed.
702 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
703 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
704 Janne Snabb (who went above and beyond: thank you).
706 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
707 string otherwise requires a connection and a bunch more work and it's
708 relatively easy to get wrong. Should also expose TLS library linkage
711 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
712 64-bit ${eval} (JH/03).
714 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
715 GNU libc to support some of the 64-bit stuff, should not lead to
716 conflicts. Defined before os.h is pulled in, so if a given platform
717 needs to override this, it can.
719 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
720 protection layer was required, which is not implemented.
721 Bugzilla 1254, patch from Wolfgang Breyha.
723 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
724 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
725 tls_dhparam take prime identifiers. Also unbreak combination of
726 OpenSSL+DH_params+TLSSNI.
728 PP/39 Disable SSLv2 by default in OpenSSL support.
734 PP/01 Solaris build fix for Oracle's LDAP libraries.
735 Bugzilla 1109, patch from Stephen Usher.
737 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
739 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
742 TF/02 Fix a couple more cases where we did not log the error message
743 when unlink() failed. See also change 4.74-TF/03.
745 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
746 lock up or crash if it happened to be inside a call to libc when it
747 got a SIGUSR1 from exiwhat.
749 The SIGUSR1 handler appends the current process status to the process
750 log which is later printed by exiwhat. It used to use the general
751 purpose logging code to do this, but several functions it calls are
752 not safe for signals.
754 The new output code in the SIGUSR1 handler is specific to the process
755 log, and simple enough that it's easy to inspect for signal safety.
756 Removing some special cases also simplifies the general logging code.
757 Removing the spurious timestamps from the process log simplifies
760 TF/04 Improved ratelimit ACL condition.
762 The /noupdate option has been deprecated in favour of /readonly which
763 has clearer semantics. The /leaky, /strict, and /readonly update modes
764 are mutually exclusive. The update mode is no longer included in the
765 database key; it just determines when the database is updated. (This
766 means that when you upgrde Exim will forget old rate measurements.)
768 Exim now checks that the per_* options are used with an update mode that
769 makes sense for the current ACL. For example, when Exim is processing a
770 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
771 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
772 must specify per_mail/readonly. If you omit the update mode it defaults to
773 /leaky where that makes sense (as before) or /readonly where required.
775 The /noupdate option is now undocumented but still supported for
776 backwards compatibility. It is equivalent to /readonly except that in
777 ACLs where /readonly is required you may specify /leaky/noupdate or
778 /strict/noupdate which are treated the same as /readonly.
780 A useful new feature is the /count= option. This is a generalization
781 of the per_byte option, so that you can measure the throughput of other
782 aggregate values. For example, the per_byte option is now equivalent
783 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
785 The per_rcpt option has been generalized using the /count= mechanism
786 (though it's more complicated than the per_byte equivalence). When it is
787 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
788 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
789 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
790 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
791 Note that using per_rcpt with a non-readonly update mode in more than
792 one ACL will cause the recipients to be double-counted. (The per_mail
793 and per_byte options don't have this problem.)
795 The handling of very low rates has changed slightly. If the computed rate
796 is less than the event's count (usually one) then this event is the first
797 after a long gap. In this case the rate is set to the same as this event's
798 count, so that the first message of a spam run is counted properly.
800 The major new feature is a mechanism for counting the rate of unique
801 events. The new per_addr option counts the number of different
802 recipients that someone has sent messages to in the last time period. It
803 behaves like per_rcpt if all the recipient addresses are different, but
804 duplicate recipient addresses do not increase the measured rate. Like
805 the /count= option this is a general mechanism, so the per_addr option
806 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
807 example, measure the rate that a client uses different sender addresses
808 with the options per_mail/unique=$sender_address. There are further
809 details in the main documentation.
811 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
813 TF/06 Removed a few PCRE remnants.
815 TF/07 Automatically extract Exim's version number from tags in the git
816 repository when doing development or release builds.
818 PP/02 Raise smtp_cmd_buffer_size to 16kB.
819 Bugzilla 879. Patch from Paul Fisher.
821 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
822 Heavily based on revision 40f9a89a from Simon Arlott's tree.
825 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
827 PP/05 Variable $av_failed, true if the AV scanner deferred.
828 Bugzilla 1078. Patch from John Horne.
830 PP/06 Stop make process more reliably on build failure.
831 Bugzilla 1087. Patch from Heiko Schlittermann.
833 PP/07 Make maildir_use_size_file an _expandable_ boolean.
834 Bugzilla 1089. Patch from Heiko Schlittermann.
836 PP/08 Handle ${run} returning more data than OS pipe buffer size.
837 Bugzilla 1131. Patch from Holger Weiß.
839 PP/09 Handle IPv6 addresses with SPF.
840 Bugzilla 860. Patch from Wolfgang Breyha.
842 PP/10 GnuTLS: support TLS 1.2 & 1.1.
844 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
847 PP/11 match_* no longer expand right-hand-side by default.
848 New compile-time build option, EXPAND_LISTMATCH_RHS.
849 New expansion conditions, "inlist", "inlisti".
851 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
853 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
855 PP/14 fix log_write() format string regression from TF/03.
856 Bugzilla 1152. Patch from Dmitry Isaikin.
862 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
864 PP/02 Harmonised TLS library version reporting; only show if debugging.
865 Layout now matches that introduced for other libraries in 4.74 PP/03.
867 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
869 PP/04 New "dns_use_edns0" global option.
871 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
874 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
875 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
877 TK/01 Updated PolarSSL code to 0.14.2.
878 Bugzilla 1097. Patch from Andreas Metzler.
880 PP/07 Catch divide-by-zero in ${eval:...}.
883 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
886 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
887 format-string attack -- SECURITY: remote arbitrary code execution.
889 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
890 time unintentionally subject to list matching rules, letting the header
891 cause arbitrary Exim lookups (of items which can occur in lists, *not*
892 arbitrary string expansion). This allowed for information disclosure.
894 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
895 INT_MIN/-1 -- value coerced to INT_MAX.
901 NM/01 Workround for PCRE version dependency in version reporting
904 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
905 This fixes portability to compilers other than gcc, notably
906 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
908 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
909 makefiles for portability to HP-UX and POSIX correctness.
911 PP/01 Permit LOOKUP_foo enabling on the make command-line.
912 Also via indented variable definition in the Makefile.
913 (Debugging by Oliver Heesakkers).
915 PP/02 Restore caching of spamd results with expanded spamd_address.
916 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
918 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
919 Improves build reliability. Fix from: Frank Elsner
921 NM/02 Fix wide character breakage in the rfc2047 coding
922 Fixes bug 1064. Patch from Andrey N. Oktyabrski
924 NM/03 Allow underscore in dnslist lookups
925 Fixes bug 1026. Patch from Graeme Fowler
927 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
928 Code patches from Adam Ciarcinski of NetBSD.
930 NM/04 Fixed exiqgrep to cope with mailq missing size issue
933 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
934 is logged, to avoid truncation. Patch from John Horne.
936 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
937 Patch from Jakob Hirsch.
939 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
940 SQL string expansion failure details.
941 Patch from Andrey Oktyabrski.
943 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
944 Patch from Simon Arlott.
946 PP/09 New lookups functionality failed to compile on old gcc which rejects
947 extern declarations in function scope.
948 Patch from Oliver Fleischmann
950 PP/10 Use sig_atomic_t for flags set from signal handlers.
951 Check getgroups() return and improve debugging.
952 Fixed developed for diagnosis in bug 927 (which turned out to be
955 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
956 Patch from Mark Zealey.
958 PP/12 Bugzilla 1056: Improved spamd server selection.
959 Patch from Mark Zealey.
961 PP/13 Bugzilla 1086: Deal with maildir quota file races.
962 Based on patch from Heiko Schlittermann.
964 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
965 Patch from Uwe Doering, sign-off by Michael Haardt.
967 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
968 variable declaration deep within a block. Bug and patch from
971 PP/15 lookups-Makefile IRIX compatibilty coercion.
973 PP/16 Make DISABLE_DKIM build knob functional.
975 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
976 Patch by Simon Arlott
978 TF/03 Fix valgrind.h portability to C89 compilers that do not support
979 variable argument macros. Our copy now differs from upstream.
985 TF/01 Failure to get a lock on a hints database can have serious
986 consequences so log it to the panic log.
988 TF/02 Log LMTP confirmation messages in the same way as SMTP,
989 controlled using the smtp_confirmation log selector.
991 TF/03 Include the error message when we fail to unlink a spool file.
993 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
994 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
995 for maintaining out-of-tree patches for some time.
997 PP/01 Bugzilla 139: Documentation and portability issues.
998 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
999 Handle per-OS dynamic-module compilation flags.
1001 PP/02 Let /dev/null have normal permissions.
1002 The 4.73 fixes were a little too stringent and complained about the
1003 permissions on /dev/null. Exempt it from some checks.
1004 Reported by Andreas M. Kirchwitz.
1006 PP/03 Report version information for many libraries, including
1007 Exim version information for dynamically loaded libraries. Created
1008 version.h, now support a version extension string for distributors
1009 who patch heavily. Dynamic module ABI change.
1011 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1012 privilege escalation vulnerability whereby the Exim run-time user
1013 can cause root to append content of the attacker's choosing to
1016 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1019 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1020 If dropping privileges for untrusted macros, we disabled normal logging
1021 on the basis that it would fail; for the Exim run-time user, this is not
1022 the case, and it resulted in successful deliveries going unlogged.
1023 Fixed. Reported by Andreas Metzler.
1029 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1030 only prepend for the Resent-* case. Fixes regression introduced in
1031 Exim 4.70 by NM/22 for Bugzilla 607.
1033 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1034 increasing numbers of administrators be bitten by this.
1036 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1038 PP/03 Bugzilla 994: added openssl_options main configuration option.
1040 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1042 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1044 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1046 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1047 without a peer certificate, leading to a segfault because of an
1048 assumption that peers always have certificates. Be a little more
1049 paranoid. Problem reported by Martin Tscholak.
1051 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1052 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1053 NB: ClamAV planning to remove STREAM in "middle of 2010".
1054 CL also introduces -bmalware, various -d+acl logging additions and
1055 more caution in buffer sizes.
1057 PP/09 Implemented reverse_ip expansion operator.
1059 PP/10 Bugzilla 937: provide a "debug" ACL control.
1061 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1063 PP/12 Bugzilla 973: Implement --version.
1065 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1067 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1069 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1071 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1072 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1073 ignore trailing whitespace.
1075 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1077 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1080 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1081 Notification from Dr Andrew Aitchison.
1083 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1084 ExtendedDetectionInfo response format.
1085 Notification from John Horne.
1087 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1090 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1091 XSL and documented dependency on system catalogs, with examples of how
1094 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1097 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1098 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1099 configuration file which is writeable by the Exim user or group.
1101 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1102 of configuration files to cover files specified with the -C option if
1103 they are going to be used with root privileges, not just the default
1106 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1107 option (effectively making it always true).
1109 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1110 files to be used while preserving root privileges.
1112 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1113 that rogue child processes cannot use them.
1115 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1116 run-time user, instead of root.
1118 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1119 Exim run-time user without dropping privileges.
1121 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1122 result string, instead of calling string_vformat() twice with the same
1125 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1126 for other users. Others should always drop root privileges if they use
1127 -C on the command line, even for a whitelisted configure file.
1129 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1131 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1137 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1138 $data_path, and $header_path variables; fixed documentation bugs and
1141 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1142 exipick to access non-standard spools, including the "frozen" queue
1145 NM/01 Bugzilla 965: Support mysql stored procedures.
1146 Patch from Alain Williams
1148 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1150 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1151 Patch from Andreas Metzler
1153 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1154 Patch from Kirill Miazine
1156 NM/05 Bugzilla 671: Added umask to procmail example.
1158 JJ/03 installed exipick 20100323.0, fixing doc bug
1160 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1161 directory. Notification and patch from Dan Rosenberg.
1163 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1165 TK/02 Improve log output when DKIM signing operation fails.
1167 MH/01 Treat the transport option dkim_domain as a colon separated
1168 list, not as a single string, and sign the message with each element,
1169 omitting multiple occurences of the same signer.
1171 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1172 Bugzilla 985, 986. Patch by Simon Arlott
1174 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1175 Patch by Simon Arlott
1177 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1178 MBX locking. Notification from Dan Rosenberg.
1184 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1186 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1188 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1190 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1192 NM/04 Bugzilla 915: Flag broken perl installation during build.
1198 TK/01 Added patch by Johannes Berg that expands the main option
1199 "spamd_address" if it starts with a dollar sign.
1201 TK/02 Write list of recipients to X-Envelope-Sender header when building
1202 the mbox-format spool file for content scanning (suggested by Jakob
1205 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1206 (http://www.dcc-servers.net/) support via dccifd. Activated by
1207 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1209 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1210 by Mark Daniel Reidel <mr@df.eu>.
1212 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1213 When building exim an external PCRE library is now needed -
1214 PCRE is a system library on the majority of modern systems.
1215 See entry on PCRE_LIBS in EDITME file.
1217 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1218 conversation. Added nologin parameter to request.
1219 Patch contributed by Kirill Miazine.
1221 TF/01 Do not log submission mode rewrites if they do not change the address.
1223 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1225 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1226 log files in place. Contributed by Roberto Lima.
1228 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1230 TF/03 Bugzilla 615: When checking the local_parts router precondition
1231 after a local_part_suffix or local_part_prefix option, Exim now
1232 does not use the address's named list lookup cache, since this
1233 contains cached lookups for the whole local part.
1235 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1236 Robert Millan. Documentation is in experimental-spec.txt.
1238 TF/04 Bugzilla 668: Fix parallel build (make -j).
1240 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1242 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1243 Patch provided by Jan Srzednicki.
1245 TF/05 Leading white space used to be stripped from $spam_report which
1246 wrecked the formatting. Now it is preserved.
1248 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1249 that they are available at delivery time.
1251 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1253 TF/08 TLS error reporting now respects the incoming_interface and
1254 incoming_port log selectors.
1256 TF/09 Produce a more useful error message if an SMTP transport's hosts
1257 setting expands to an empty string.
1259 NM/06 Bugzilla 744: EXPN did not work under TLS.
1260 Patch provided by Phil Pennock.
1262 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1263 Patch provided by Richard Godbee.
1265 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1266 acl_smtp_notquit, added index entry.
1268 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1269 Patch provided by Eugene Bujak.
1271 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1272 accept(). Patch provided by Maxim Dounin.
1274 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1275 Patch provided by Phil Pennock.
1277 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1279 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1280 Patch provided by Brad "anomie" Jorsch.
1282 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1283 Patch provided by Dean Brooks.
1285 TK/05 Add native DKIM support (does not depend on external libraries).
1287 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1288 Patch provided by Graeme Fowler.
1290 NM/16 Bugzilla 851: Documentation example syntax fix.
1292 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1294 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1297 NM/19 Bugzilla 745: TLS version reporting.
1298 Patch provided by Phil Pennock.
1300 NM/20 Bugzilla 167: bool: condition support.
1301 Patch provided by Phil Pennock.
1303 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1304 clients. Patch provided by Phil Pennock.
1306 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1307 Patch provided by Brad "anomie" Jorsch.
1309 NM/23 Bugzilla 687: Fix misparses in eximstats.
1310 Patch provided by Heiko Schlittermann.
1312 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1313 Patch provided by Heiko Schlittermann.
1315 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1316 plus update to original patch.
1318 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1320 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1321 Patch provided by David Brownlee.
1323 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1325 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1327 NM/30 Bugzilla 888: TLS documentation bugfixes.
1329 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1331 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1332 Unlike the original bugzilla I have changed all shell scripts in src tree.
1334 NM/33 Bugzilla 898: Transport filter timeout fix.
1335 Patch by Todd Rinaldo.
1337 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1338 Patch by Serge Demonchaux.
1340 NM/35 Bugzilla 39: Base64 decode bug fixes.
1341 Patch by Jakob Hirsch.
1343 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1345 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1347 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1349 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1355 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1356 ALT-N's libdkim that I have put here:
1357 http://duncanthrax.net/exim-experimental/
1359 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1360 were called 'true' and it seems that C99 defines that as a reserved
1361 keyword to be used with 'bool' variable types. That means you could
1362 not include C99-style headers which use bools without triggering
1363 build errors in sieve.c.
1365 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1366 as mailq or other aliases. Changed the --help handling significantly
1367 to do whats expected. exim_usage() emits usage/help information.
1369 SC/01 Added the -bylocaldomain option to eximstats.
1371 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1373 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1375 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1381 PH/01 Another patch from the Sieve maintainer.
1383 PH/02 When an IPv6 address is converted to a string for single-key lookup
1384 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1385 dots are used instead of colons so that keys in lsearch files need not
1386 contain colons. This was done some time before quoting was made available
1387 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1388 (notated using the quote facility) so as to distinguish them from IPv4
1389 keys. This meant that lookups for IP addresses in host lists did not work
1390 for iplsearch lookups.
1392 This has been fixed by arranging for IPv6 addresses to be expressed with
1393 colons if the lookup type is iplsearch. This is not incompatible, because
1394 previously such lookups could never work.
1396 The situation is now rather anomolous, since one *can* have colons in
1397 ordinary lsearch keys. However, making the change in all cases is
1398 incompatible and would probably break a number of configurations.
1400 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1403 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1404 conversion specification without a maximum field width, thereby enabling
1405 a rogue spamd server to cause a buffer overflow. While nobody in their
1406 right mind would setup Exim to query an untrusted spamd server, an
1407 attacker that gains access to a server running spamd could potentially
1408 exploit this vulnerability to run arbitrary code as the Exim user.
1410 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1411 $primary_hostname instead of what libspf2 thinks the hosts name is.
1413 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1414 a directory entry by the name of the lookup key. Previously, if a
1415 symlink pointed to a non-existing file or a file in a directory that
1416 Exim lacked permissions to read, a lookup for a key matching that
1417 symlink would fail. Now it is enough that a matching directory entry
1418 exists, symlink or not. (Bugzilla 503.)
1420 PH/03 The body_linecount and body_zerocount variables are now exported in the
1423 PH/04 Added the $dnslist_matched variable.
1425 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1426 This means they are set thereafter only if the connection becomes
1429 PH/06 Added the client_condition to authenticators so that some can be skipped
1430 by clients under certain conditions.
1432 PH/07 The error message for a badly-placed control=no_multiline_responses left
1433 "_responses" off the end of the name.
1435 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1437 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1438 (without spaces) instead of just copying the configuration text.
1440 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1442 PH/11 Added $max_received_linelength.
1444 PH/12 Added +ignore_defer and +include_defer to host lists.
1446 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1447 way in which PCRE > 7.0 is built.
1449 PH/14 Implemented queue_only_load_latch.
1451 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1452 MAIL command. The effect was to mangle the value on 64-bit systems.
1454 PH/16 Another patch from the Sieve maintainer.
1456 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1458 PH/18 If a system quota error occurred while trying to create the file for
1459 a maildir delivery, the message "Mailbox is full" was not appended to the
1460 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1461 only to a quota excession during the actual writing of the file.
1463 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1464 characters?) which causes problems in log lines. The DN values are now
1465 passed through string_printing() before being added to log lines.
1467 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1468 and InterBase are left for another time.)
1470 PH/21 Added message_body_newlines option.
1472 PH/22 Guard against possible overflow in moan_check_errorcopy().
1474 PH/23 POSIX allows open() to be a macro; guard against that.
1476 PH/24 If the recipient of an error message contained an @ in the local part
1477 (suitably quoted, of course), incorrect values were put in $domain and
1478 $local_part during the evaluation of errors_copy.
1484 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1485 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1488 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1489 issue a MAIL command.
1491 PH/02 In an ACL statement such as
1493 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1495 if a client was not listed at all, or was listed with a value other than
1496 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1497 the condition was not true (as it should be), so access was not denied.
1498 The bug was that the ! inversion was incorrectly passed on to the second
1499 item. This has been fixed.
1501 PH/03 Added additional dnslists conditions == and =& which are different from
1502 = and & when the dns lookup returns more than one IP address.
1504 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1505 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1507 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1508 FSYNC, which compiles an option called disable_fsync that allows for
1509 bypassing fsync(). The documentation is heavily laced with warnings.
1511 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1513 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1514 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1515 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1516 including adding "make clean"; (3) Added -fPIC when compiling the test
1517 dynamically loaded module, to get rid of a warning.
1519 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1520 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1521 = 0s. The bug is otherwise harmless.
1523 PH/07 There was a bug in the dovecot authenticator such that the value of
1524 $auth1 could be overwritten, and so not correctly preserved, after a
1525 successful authentication. This usually meant that the value preserved by
1526 the server_setid option was incorrect.
1528 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1530 PH/09 Installed PCRE release 7.0.
1532 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1533 run for batched SMTP input. It is now run at the start of every message
1534 in the batch. While fixing this I discovered that the process information
1535 (output by running exiwhat) was not always getting set for -bs and -bS
1536 input. This is fixed, and it now also says "batched" for BSMTP.
1538 PH/11 Added control=no_pipelining.
1540 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1541 patch, slightly modified), and move the expansion of helo_data till after
1542 the connection is made in the smtp transport (so it can use these
1545 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1547 PH/14 Added log_selector = +pid.
1549 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1551 PH/16 Add ${if forany and ${if forall.
1553 PH/17 Added dsn_from option to vary the From: line in DSNs.
1555 PH/18 Flush SMTP output before performing a callout, unless control =
1556 no_callout_flush is set.
1558 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1559 was true (the default) a successful delivery failed to delete the retry
1560 item, thus causing premature timeout of the address. The bug is now
1563 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1565 PH/21 Long custom messages for fakedefer and fakereject are now split up
1566 into multiline reponses in the same way that messages for "deny" and
1567 other ACL rejections are.
1569 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1570 with slight modification.
1572 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1573 draft, changing the syntax and factoring some duplicate code".
1575 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1576 for deliveries of the second and subsequent messages over the same SMTP
1579 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1580 ${reduce, with only minor "tidies".
1582 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1584 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1585 expansion side effects.
1587 PH/27 When a message times out after an over-quota error from an Exim-imposed
1588 quota, the bounce message says "mailbox is full". This message was not
1589 being given when it was a system quota that was exceeded. It now should
1592 MH/03 Made $recipients available in local_scan(). local_scan() already has
1593 better access to the recipient list through recipients_list[], but
1594 $recipients can be useful in postmaster-provided expansion strings.
1596 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1597 in the case of a MAIL command with additional options following the
1598 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1599 were accidentally chopped off.
1601 PH/29 SMTP synchronization checks are implemented when a command is read -
1602 there is a check that no more input is waiting when there shouldn't be
1603 any. However, for some commands, a delay in an ACL can mean that it is
1604 some time before the response is written. In this time, more input might
1605 arrive, invalidly. So now there are extra checks after an ACL has run for
1606 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1607 pipelining has not been advertised.
1609 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1611 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1612 returned in the SMTP response when smtp_return_error_details was false.
1613 This has been fixed.
1615 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1616 instead of the C I/O that was originally supplied, because problems were
1617 reported on Solaris.
1619 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1620 Exim which did not show up earlier: it was assuming that a call to
1621 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1622 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1623 was a macro that became an assignment, so it seemed to work. This has
1624 changed to a proper function call with a void return, hence the compile
1625 error. Exim's code has been fixed.
1627 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1630 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1632 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1634 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1635 because it turns out to be too big or there is a timeout) there is no
1636 "Completed" line in the log. When some messages of this type were
1637 selected by exigrep, they were listed as "not completed". Others were
1638 picked up by some special patterns. I have improved the selection
1639 criteria to be more general.
1641 PH/38 The host_find_failed option in the manualroute router can now be set
1642 to "ignore", to completely ignore a host whose IP address cannot be
1643 found. If all hosts are ignored, the behaviour is controlled by the new
1644 host_all_ignored option.
1646 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1647 homing or because of multiple MX records with /mx) generated more than
1648 one IP address, and the following item turned out to be the local host,
1649 all the secondary addresses of the first item were incorrectly removed
1650 from the list, along with the local host and any following hosts (which
1651 is what is supposed to happen).
1653 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1654 whoever called Exim into the -H file. In the case of the daemon it was
1655 behaving confusingly. When first started, it used values for whoever
1656 started the daemon, but after a SIGHUP it used the Exim user (because it
1657 calls itself on a restart). I have changed the code so that it now always
1660 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1661 message are rejected with the same error (e.g. no authentication or bad
1662 sender address), and a DATA command is nevertheless sent (as can happen
1663 with PIPELINING or a stupid MUA), the error message that was given to the
1664 RCPT commands is included in the rejection of the DATA command. This is
1665 intended to be helpful for MUAs that show only the final error to their
1668 PH/42 Another patch from the Sieve maintainer.
1670 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1671 Eximstats - Fixed some broken HTML links and added missing column headers
1673 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1674 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1676 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1678 PH/43 Yet another patch from the Sieve maintainer.
1680 PH/44 I found a way to check for a TCP/IP connection going away before sending
1681 the response to the final '.' that terminates a message, but only in the
1682 case where the client has not sent further data following the '.'
1683 (unfortunately, this is allowed). However, in many cases there won't be
1684 any further data because there won't be any more messages to send. A call
1685 to select() can be used: if it shows that the input is "ready", there is
1686 either input waiting, or the socket has been closed. An attempt to read
1687 the next input character can distinguish the two cases. Previously, Exim
1688 would have sent an OK response which the client would never have see.
1689 This could lead to message repetition. This fix should cure that, at
1690 least in a lot of common cases.
1692 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1693 advertised in response to EHLO.
1699 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1700 fixed by 4.65/MH/01 (is this a record?) are fixed:
1702 (i) An empty string was always treated as zero by the numeric comparison
1703 operators. This behaviour has been restored.
1705 (ii) It is documented that the numeric comparison operators always treat
1706 their arguments as decimal numbers. This was broken in that numbers
1707 starting with 0 were being interpreted as octal.
1709 While fixing these problems I realized that there was another issue that
1710 hadn't been noticed. Values of message_size_limit (both the global option
1711 and the transport option) were treated as octal if they started with 0.
1712 The documentation was vague. These values are now always treated as
1713 decimal, and I will make that clear in the documentation.
1719 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1720 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1723 MH/01 Don't check that the operands of numeric comparison operators are
1724 integers when their expansion is in "skipping" mode (fixes bug
1725 introduced by 4.64-PH/07).
1727 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1728 child addresses, Exim now panics and dies. Previously, because the count
1729 is held in a short int, deliveries were likely to be lost. As such a
1730 large number of recipients for a single message is ridiculous
1731 (performance will be very, very poor), I have chosen to impose a limit
1732 rather than extend the field.
1738 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1739 leftover -K file (the existence of which was triggered by #402).
1740 While we were at it, introduced process PID as part of the -K
1741 filename. This should rule out race conditions when creating
1744 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1745 processing considerably. Previous code took too long for large mails,
1746 triggering a timeout which in turn triggers #401.
1748 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1749 in the DK code in transports.c. sendfile() is not really portable,
1750 hence the _LINUX specificness.
1752 TF/01 In the add_headers option to the mail command in an Exim filter,
1753 there was a bug that Exim would claim a syntax error in any
1754 header after the first one which had an odd number of characters
1757 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1758 callout verification, Exim cached a "reject" for the entire domain. This
1759 is correct for most verifications, but it is not correct for a recipient
1760 verification with use_sender or use_postmaster set, because in that case
1761 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1762 case of MAIL FROM:<> rejection from other early rejections (e.g.
1763 rejection of HELO). When verifying a recipient using a non-null MAIL
1764 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1765 Whatever the result of the callout, the value of the domain cache is
1766 left unchanged (for any other kind of callout, getting as far as trying
1767 RCPT means that the domain itself is ok).
1769 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1772 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1773 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1774 session, the daemon ignores ECONNECT errors and logs others; it now
1775 ignores EPIPE as well.
1777 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1778 (quoted-printable decoding).
1780 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1781 later the small subsequent patch to fix an introduced bug.
1783 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1785 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1787 PH/08 An error is now given if message_size_limit is specified negative.
1789 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1790 to be given (somewhat) arbitrary names.
1792 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1795 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1796 miscellaneous code fixes
1798 PH/10 Added the log_reject_target ACL modifier to specify where to log
1801 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1802 hostname. This is wrong, because it relates to the incoming message (and
1803 probably the interface on which it is arriving) and not to the outgoing
1804 callout (which could be using a different interface). This has been
1805 changed to use the value of the helo_data option from the smtp transport
1806 instead - this is what is used when a message is actually being sent. If
1807 there is no remote transport (possible with a router that sets up host
1808 addresses), $smtp_active_hostname is used.
1810 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1811 tweaks were necessary in order to get it to work (see also 21 below):
1812 (a) The code assumed that strncpy() returns a negative number on buffer
1813 overflow, which isn't the case. Replaced with Exim's string_format()
1815 (b) There were several signed/unsigned issues. I just did the minimum
1816 hacking in of casts. There is scope for a larger refactoring.
1817 (c) The code used strcasecmp() which is not a standard C function.
1818 Replaced with Exim's strcmpic() function.
1819 (d) The code set only $1; it now sets $auth1 as well.
1820 (e) A simple test gave the error "authentication client didn't specify
1821 service in request". It would seem that Dovecot has changed its
1822 interface. Fortunately there's a specification; I followed it and
1823 changed what the client sends and it appears to be working now.
1825 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1828 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1829 address A is aliased to B and C, where B exists and C does not. Without
1830 -v the output is "A verified" because verification stops after a
1831 successful redirection if more than one address is generated. However,
1832 with -v the child addresses are also verified. Exim was outputting "A
1833 failed to verify" and then showing the successful verification for C,
1834 with its parentage. It now outputs "B failed to verify", showing B's
1835 parentage before showing the successful verification of C.
1837 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1838 look up a TXT record in a specific list after matching in a combined
1841 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1842 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1843 they consult the DNS. I had assumed they would set it the way they
1844 wanted; and indeed my experiments on Linux seem to show that in some
1845 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1846 To be on the safe side, however, I have now made the interface to
1847 host_find_byname() similar to host_find_bydns(), with an argument
1848 containing the DNS resolver options. The host_find_byname() function now
1849 sets these options at its start, just as host_find_bydns() does. The smtp
1850 transport options dns_qualify_single and dns_search_parents are passed to
1851 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1852 of host_find_byname() use the default settings of RES_DEFNAMES
1853 (qualify_single) but not RES_DNSRCH (search_parents).
1855 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1856 spool_read_header() do less string testing, by means of a preliminary
1857 switch on the second character of optional "-foo" lines. (This is
1858 overdue, caused by the large number of possibilities that now exist.
1859 Originally there were few.) While I was there, I also converted the
1860 str(n)cmp tests so they don't re-test the leading "-" and the first
1861 character, in the hope this might squeeze out yet more improvement.
1863 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1864 flag allowing group syntax was set by the header_syntax check but not
1865 turned off, possible causing trouble later; (2) The flag was not being
1866 set at all for the header_verify test, causing "group"-style headers to
1867 be rejected. I have now set it in this case, and also caused header_
1868 verify to ignore an empty address taken from a group. While doing this, I
1869 came across some other cases where the code for allowing group syntax
1870 while scanning a header line wasn't quite right (mostly, not resetting
1871 the flag correctly in the right place). These bugs could have caused
1872 trouble for malformed header lines. I hope it is now all correct.
1874 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1875 with the "reply" argument non-NULL. The code, however (which originally
1876 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1877 but it didn't always do it. This confused somebody who was copying the
1878 code for some other use. I have removed all the tests.
1880 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1881 feature that was used to support insecure browsers during the U.S. crypto
1882 embargo. It requires special client support, and Exim is probably the
1883 only MTA that supported it -- and would never use it because real RSA is
1884 always available. This code has been removed, because it had the bad
1885 effect of slowing Exim down by computing (never used) parameters for the
1886 RSA_EXPORT functionality.
1888 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1889 authenticator to fail if there's a tab character in the incoming data
1890 (there should never be unless someone is messing about, as it's supposed
1891 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1892 if the connection is using TLS or if the remote IP is the same as the
1893 local IP, and the "valid-client-cert option" if a client certificate has
1896 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1897 authenticators. This can be used for authorization after authentication
1898 succeeds. (In the case of plaintext, it servers for both authentication
1901 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1902 if any retry times were supplied.
1904 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1905 connection, where there is no client IP address to check. In this
1906 situation, the verify now always succeeds.
1908 PH/25 Applied John Jetmore's -Mset patch.
1910 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1912 PH/27 In a string expansion for a processed (not raw) header when multiple
1913 headers of the same name were present, leading whitespace was being
1914 removed from all of them, but trailing whitespace was being removed only
1915 from the last one. Now trailing whitespace is removed from each header
1916 before concatenation. Completely empty headers in a concatenation (as
1917 before) are ignored.
1919 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1920 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1922 PH/29 [Removed. This was a change that I later backed out, and forgot to
1923 correct the ChangeLog entry (that I had efficiently created) before
1924 committing the later change.]
1926 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1927 address errors (4xx response to RCPT) over the same connection as other
1928 messages routed to the same hosts. Such deliveries are always "forced",
1929 so retry times are not inspected. This resulted in far too many retries
1930 for the affected addresses. The effect occurred only when there were more
1931 hosts than the hosts_max_try setting in the smtp transport when it had
1932 the 4xx errors. Those hosts that it had tried were not added to the list
1933 of hosts for which the message was waiting, so if all were tried, there
1934 was no problem. Two fixes have been applied:
1936 (i) If there are any address or message errors in an SMTP delivery, none
1937 of the hosts (tried or untried) are now added to the list of hosts
1938 for which the message is waiting, so the message should not be a
1939 candidate for sending over the same connection that was used for a
1940 successful delivery of some other message. This seems entirely
1941 reasonable: after all the message is NOT "waiting for some host".
1942 This is so "obvious" that I'm not sure why it wasn't done
1943 previously. Hope I haven't missed anything, but it can't do any
1944 harm, as the worst effect is to miss an optimization.
1946 (ii) If, despite (i), such a delivery is accidentally attempted, the
1947 routing retry time is respected, so at least it doesn't keep
1948 hammering the server.
1950 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1951 in ${readsocket because some servers need this prod.
1953 PH/32 Added some extra debug output when updating a wait-xxx database.
1955 PH/33 The hint "could be header name not terminated by colon", which has been
1956 given for certain expansion errors for a long time, was not being given
1957 for the ${if def:h_colon_omitted{... case.
1959 PH/34 The spec says: "With one important exception, whenever a domain list is
1960 being scanned, $domain contains the subject domain." There was at least
1961 one case where this was not true.
1963 PH/35 The error "getsockname() failed: connection reset by peer" was being
1964 written to the panic log as well as the main log, but it isn't really
1965 panic-worthy as it just means the connection died rather early on. I have
1966 removed the panic log writing for the ECONNRESET error when getsockname()
1969 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1970 runs only) independently of the message's sender address. This meant
1971 that, if the 4xx error was in fact related to the sender, a different
1972 message to the same recipient with a different sender could confuse
1973 things. In particualar, this can happen when sending to a greylisting
1974 server, but other circumstances could also provoke similar problems.
1975 I have changed the default so that the retry time for these errors is now
1976 based a combination of the sender and recipient addresses. This change
1977 can be overridden by setting address_retry_include_sender=false in the
1980 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1981 remote server are returned as part of bounce messages. This was not
1982 happening for LMTP over a pipe (the lmtp transport), but now it is the
1983 same for both kinds of LMTP.
1985 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1986 in header lines that were in fact CNAMEs. This is no longer the case.
1988 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1989 and queue runs started by the daemon processed all messages. This has
1990 been fixed so that -R and -S can now usefully be given with -q<time>.
1992 PH/40 Import PCRE release 6.7 (fixes some bugs).
1994 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1996 PH/42 Give an error if -q is specified more than once.
1998 PH/43 Renamed the variables $interface_address and $interface_port as
1999 $received_ip_address and $received_port, to make it clear that these
2000 values apply to message reception, and not to the outgoing interface when
2001 a message is delivered. (The old names remain recognized, of course.)
2003 PH/44 There was no timeout on the connect() call when using a Unix domain
2004 socket in the ${readsocket expansion. There now is.
2006 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2007 be meaningful with "accept".
2009 SC/01 Eximstats V1.43
2010 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2012 SC/02 Eximstats V1.44
2013 Use a glob alias rather than an array ref in the generated
2014 parser. This improves both readability and performance.
2016 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2017 Collect SpamAssassin and rejection statistics.
2018 Don't display local sender or destination tables unless
2019 there is data to show.
2020 Added average volumes into the top table text output.
2022 SC/04 Eximstats V1.46
2023 Collect data on the number of addresses (recipients)
2024 as well as the number of messages.
2026 SC/05 Eximstats V1.47
2027 Added 'Message too big' to the list of mail rejection
2028 reasons (thanks to Marco Gaiarin).
2030 SC/06 Eximstats V1.48
2031 Mainlog lines which have GMT offsets and are too short to
2032 have a flag are now skipped.
2034 SC/07 Eximstats V1.49 (Alain Williams)
2035 Added the -emptyok flag.
2037 SC/08 Eximstats V1.50
2038 Fixes for obtaining the IP address from reject messages.
2040 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2041 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2042 whitesspace changes from 4.64-PH/27
2044 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2047 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2048 are found, allow negative numbers in numeric criteria)
2050 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2052 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2053 to match changes made in 4.64-PH/43
2055 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2057 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2058 "domain" is actually an IP address, and force a failure. This locks out
2059 those revolvers/nameservers that support "A-for-A" lookups, in
2060 contravention of the specifications.
2062 PH/48 When a host name was looked up from an IP address, and the subsequent
2063 forward lookup of the name timed out, the host name was left in
2064 $sender_host_name, contrary to the specification.
2066 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2067 restricted to single-key lookups, Exim was not diagnosing an error if
2068 * or *@ was used with a query-style lookup.
2070 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2072 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2073 long ago, but noone interested enough thought of it. Let's just say that
2074 the "1.1" means that there are some new functions that weren't there at
2075 some point in the past.
2077 PH/51 Error processing for expansion failure of helo_data from an smtp
2078 transport during callout processing was broken.
2080 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2081 tested/used via the -bh/-bhc/-bs options.
2083 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2084 bug, fixed in subsequent PCRE releases).
2086 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2087 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2089 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2095 SC/01 Use a glob alias rather than an array ref in eximstats generated
2096 parser. This improves both readability and performance.
2098 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2099 Don't display local sender or destination tables in eximstats unless
2100 there is data to show.
2101 Added average volumes into the eximstats top table text output.
2103 SC/03 Collect data on the number of addresses (recipients) as well
2104 as the number of messages in eximstats.
2106 TF/01 Correct an error in the documentation for the redirect router. Exim
2107 does (usually) call initgroups() when daemonizing.
2109 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2110 with consistent privilege compared to when running as a daemon.
2112 TF/03 Note in the spec that $authenticated_id is not set for local
2113 submissions from trusted users.
2115 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2116 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2118 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2119 by adding some example configuration directives to the default
2120 configuration file. A little bit of work is required to uncomment the
2121 directives and define how usernames and passwords are checked, but
2122 there is now a framework to start from.
2124 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2125 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2126 without this. I don't know how relevant this is to other LDAP libraries.
2128 PH/02 Add the verb name to the "unknown ACL verb" error.
2130 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2132 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2134 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2135 directory not expanded when it should be if an expanded home directory
2136 was set for the address (which is overridden by the transport).
2138 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2141 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2142 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2143 because it is too late at that time, and has no effect.
2145 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2146 security issue with \' (bugzilla #107). I could not use the
2147 PQescapeStringConn() function, because it needs a PGconn value as one of
2150 PH/08 When testing addresses using -bt, indicate those final addresses that
2151 are duplicates that would not cause an additional delivery. At least one
2152 person was confused, thinking that -bt output corresponded to deliveries.
2153 (Suppressing duplicates isn't a good idea as you lose the information
2154 about possibly different redirections that led to the duplicates.)
2156 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2157 systems where poll() doesn't work, in particular OS X.
2159 PH/10 Added more information to debugging output for retry time not reached.
2161 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2162 operations in malware.c.
2164 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2167 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2168 syslog_duplication set false, log lines that would normally be written
2169 both the the main log and to the reject log were not written to syslog at
2172 PH/14 In the default configuration, change the use of "message" in ACL warn
2173 statements to "add_header".
2175 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2176 not followed by a command (e.g. "seen endif").
2178 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2179 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2182 PH/17 Added extra conditions to the default value of delay_warning_condition
2186 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2187 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2188 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2191 The Auto-Submitted: and various List- headers are standardised, whereas I
2192 don't think Precedence: ever was.
2194 PH/18 Refactored debugging code in route_finduser() to show more information,
2195 in particular, the error code if getpwnam() issues one.
2197 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2198 This is apparently needed in addition to the PH/07 change above to avoid
2199 any possible encoding problems.
2201 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2202 but not after initializing Perl.
2204 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2205 output them only if debugging. By default they are written stderr,
2206 apparently, which is not desirable.
2208 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2211 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2214 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2216 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2217 authenticated or an ident call has been made. Suppress the default
2218 values for $authenticated_id and $authenticated_sender (but permit -oMai
2219 and -oMas) when testing with -bh.
2221 PH/24 Re-jigged the order of the tests in the default configuration so that the
2222 tests for valid domains and recipients precede the DNS black list and CSA
2223 tests, on the grounds that those ones are more expensive.
2225 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2226 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2227 This bug exists in every version of Exim that I still have, right back to
2230 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2231 However, an attempt to turn on case-sensitivity in a regex key by
2232 including (?-i) didn't work because the subject string was already
2233 lowercased, and the effects were non-intuitive. It turns out that a
2234 one-line patch can be used to allow (?-i) to work as expected.
2240 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2241 other effects) broke the use of negated acl sub-conditions.
2243 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2246 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2247 "Deny" causes Exim to reject the incoming connection with a 554 error.
2248 Unfortunately, if there is a major crisis, such as a disk failure,
2249 tcp-wrappers gives "deny", whereas what one would like would be some
2250 kind of temporary error. A kludge has been added to help with this.
2251 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2252 554 error is used if errno is still zero or contains ENOENT (which occurs
2253 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2256 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2258 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2259 errors. Otherwise a message that provokes a temporary error (when other
2260 messages do not) can cause a whole host to time out.
2262 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2263 addresses were routed directly to files or pipes from a redirect router.
2264 File deliveries just didn't batch; pipe deliveries might have suffered
2267 PH/06 A failure to get a lock for a hints database would erroneously always say
2268 "Failed to get write lock", even when it was really a read lock.
2270 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2271 of 0600. This has been changed to use the value of the lockfile_mode
2272 option (which defaults to 0600).
2274 PH/08 Applied small patch from the Sieve maintainer.
2276 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2277 folder from quota calculations, a direct delivery into this folder messed
2278 up the contents of the maildirsize file. This was because the regex was
2279 used only to exclude .Trash (or whatever) when the size of the mailbox
2280 was calculated. There was no check that a delivery was happening into an
2281 excluded directory. This bug has been fixed by ignoring all quota
2282 processing for deliveries into excluded directories.
2284 PH/10 Added the maildirfolder_create_regex option to appendfile.
2290 PH/01 The code for finding all the local interface addresses on a FreeBSD
2291 system running IPv6 was broken. This may well have applied to all BSD
2292 systems, as well as to others that have similar system calls. The broken
2293 code found IPv4 interfaces correctly, but gave incorrect values for the
2294 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2295 that it would not match correctly against @[] and not recognize the IPv6
2298 PH/02 The ipliteral router was not recognizing addresses of the form user@
2299 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2301 PH/03 Added disable_ipv6.
2303 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2304 decoded form, because it is most often used to construct To: headers
2305 lines in autoreplies, and the decoded form may well be syntactically
2306 invalid. However, $reply_address has leading white space removed, and all
2307 newlines turned into spaces so that the autoreply transport does not
2310 PH/05 If group was specified without a user on a router, and no group or user
2311 was specified on a transport, the group from the router was ignored.
2313 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2314 for visible compile-time settings that can be used to change these
2315 numbers, for those that want even more. Backwards compatibility with old
2316 spool files has been maintained. However, going back to a previous Exim
2317 release will lost any variables that are in spool files.
2319 PH/07 Two small changes when running in the test harness: increase delay when
2320 passing a TCP/IP connection to a new process, in case the original
2321 process has to generate a bounce, and remove special handling of
2322 127.0.0.2 (sic), which is no longer necessary.
2324 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2325 be the same on different OS.
2327 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2330 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2331 whether --show-vars was specified or not
2333 JJ/02 exipick: Added support for new ACL variable spool format introduced
2336 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2337 syntactically invalid From: or Reply-to: line, and a filter used this to
2338 generate an autoreply, and therefore failed to obtain an address for the
2339 autoreply, Exim could try to deliver to a non-existent relative file
2340 name, causing unrelated and misleading errors. What now happens is that
2341 it logs this as a hard delivery error, but does not attempt to create a
2344 PH/11 The exinext utility has a -C option for testing purposes, but although
2345 the given file was scanned by exinext itself; it wasn't being passed on
2346 when Exim was called.
2348 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2349 an end-of-file indication when reading a command response.
2351 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2352 compiled. In many other places in Exim, IPv6 addresses are always
2353 recognized, so I have changed this. It also means that IPv4 domain
2354 literals of the form [IPV4:n.n.n.n] are now always recognized.
2356 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2357 used if the router is not running as root, for example, when verifying at
2358 ACL time, or when using -bh. The debugging output from this situation was
2359 non-existent - all you got was a failure to exec. I have made two
2362 (a) Failures to set uid/gid, the current directory, or a process leader
2363 in a subprocess such as that created by queryprogram now generate
2364 suitable debugging ouput when -d is set.
2366 (b) The queryprogram router detects when it is not running as root,
2367 outputs suitable debugging information if -d is set, and then runs
2368 the subprocess without attempting to change uid/gid.
2370 PH/15 Minor change to Makefile for building test_host (undocumented testing
2373 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2374 additional section of a DNS packet that returns MX or SRV records.
2375 Instead, it always explicitly searches for A/AAAA records. This avoids
2376 major problems that occur when a DNS server includes only records of one
2377 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2378 fixed another bug: if SRV records were looked up and the corresponding
2379 address records were *not* found in the additional section, the port
2380 values from the SRV records were lost.
2382 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2383 using the correct key (the original address) when searching the retry
2384 rules in order to find which one to use for generating the retry hint.
2386 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2387 adding the default one. Similarly, if it contains a Reply-To: header, the
2388 errors_reply_to option, if set, is not used.
2390 PH/19 When calculating a retry time, Exim used to measure the "time since
2391 failure" by looking at the "first failed" field in the retry record. Now
2392 it does not use this if it is later than than the arrival time of the
2393 message. Instead it uses the arrival time. This makes for better
2394 behaviour in cases where some deliveries succeed, thus re-setting the
2395 "first failed" field. An example is a quota failure for a huge message
2396 when small messages continue to be delivered. Without this change, the
2397 "time since failure" will always be short, possible causing more frequent
2398 delivery attempts for the huge message than are intended.
2399 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2401 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2402 $1, $2, $3) because the numerical variables can be reset during some
2403 expansion items (e.g. "match"), thereby losing the authentication data.
2405 PH/21 Make -bV show the size of off_t variables so that the test suite can
2406 decide whether to run tests for quotas > 2G.
2408 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2409 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2410 filecount value is greater than 2G or if a quota value is greater than 2G
2411 on a system where the size of off_t is not greater than 4, a panic error
2414 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2415 never match. The debug and -bh output now contains an explicit error
2416 message indicating a malformed IPv4 address or mask.
2418 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2419 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2420 PH/23 above applies.
2422 PH/25 Do not write to syslog when running in the test harness. The only
2423 occasion when this arises is a failure to open the main or panic logs
2424 (for which there is an explicit test).
2426 PH/26 Added the /no_tell option to "control=freeze".
2428 PH/27 If a host name lookup failed very early in a connection, for example, if
2429 the IP address matched host_lookup and the reverse lookup yielded a name
2430 that did not have a forward lookup, an error message of the form "no IP
2431 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2432 could be logged. Now it outputs the IP address instead of "NULL".
2434 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2435 allows the sender and the authenticated sender to be set when
2436 submitting a message from within Exim. Since child_open_exim() is
2437 documented for local_scan(), the new function should be too.
2439 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2440 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2441 results in an empty string is now treated as unset.
2443 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2445 PH/31 Added sender_verify_fail logging option.
2447 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2448 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2449 tidied the source and removed it altogether.
2451 PH/33 When a queue run was abandoned because the load average was too high, a
2452 log line was always written; now it is written only if the queue_run log
2453 selector is set. In addition, the log line for abandonment now contains
2454 information about the queue run such as the pid. This is always present
2455 in "start" and "stop" lines but was omitted from the "abandon" line.
2457 PH/34 Omit spaces between a header name and the colon in the error message that
2458 is given when verify = headers_syntax fails (if there are lots of them,
2459 the message gets confusing).
2461 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2462 names, as there are now some PTR records that contain slashes. This check
2463 is only to protect against broken name servers that fall over on strange
2464 characters, so the fact that it applies to all lookups doesn't matter.
2466 PH/36 Now that the new test suite is complete, we can remove some of the
2467 special code in Exim that was needed for the old test suite. For example,
2468 sorting DNS records because real resolvers return them in an arbitrary
2469 order. The new test suite's fake resolver always returns records in the
2472 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2473 bounces) except when queue_only is set, to avoid logging races between
2474 the different processes.
2476 PH/38 Panic-die if .include specifies a non-absolute path.
2478 PH/39 A tweak to the "H" retry rule from its user.
2480 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2481 a label. They prevented compilation on older perls.
2483 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2484 a warning to be raised on newish perls.
2486 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2487 on queue. Changes to match documented behaviour of showing count of
2488 messages matching specified criteria.
2490 PH/40 Changed the default ident timeout from 30s to 5s.
2492 PH/41 Added support for the use of login_cap features, on those BSD systems
2493 that have them, for controlling the resources used by pipe deliveries.
2495 PH/42 The content-scanning code uses fopen() to create files in which to put
2496 message data. Previously it was not paying any attention to the mode of
2497 the files. Exim runs with umask(0) because the rest of the code creates
2498 files with open(), and sets the required mode explicitly. Thus, these
2499 files were ending up world-writeable. This was not a big issue, because,
2500 being within the spool directory, they were not world-accessible. I have
2501 created a function called modefopen, which takes an additional mode
2502 argument. It sets umask(777), creates the file, chmods it to the required
2503 mode, then resets the umask. All the relevant calls to fopen() in the
2504 content scanning code have been changed to use this function.
2506 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2507 to 24 hours. This avoids potential overflow problems when processing G
2508 and H retry rules. I suspect nobody ever tinkers with this value.
2510 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2512 PH/45 When the plaintext authenticator is running as a client, the server's
2513 challenges are checked to ensure they are valid base64 strings. By
2514 default, the authentication attempt is cancelled if an invalid string is
2515 received. Setting client_ignore_invalid_base64 true ignores these errors.
2516 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2517 they are received. Thus, the responses can be made to depend on the
2518 challenges. If an invalid string is ignored, an empty string is placed in
2521 PH/46 Messages that are created by the autoreply transport now contains a
2522 References: header, in accordance with RFCs 2822 and 3834.
2524 PH/47 Added authenticated_sender_force to the smtp transport.
2526 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2528 PH/49 Installed latest patch from the Sieve maintainer.
2530 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2531 was also set, the appendfile transport was unnecessarily scanning a
2532 directory of message files (e.g. for maildir delivery) to find the count
2533 of files (along with the size), even though it did not need this
2534 information. It now does the scan only if it needs to find either the
2535 size of the count of files.
2537 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2539 PH/52 Two bugs concerned with error handling when the smtp transport is
2542 (i) Exim was not creating retry information for temporary errors given
2543 for individual recipients after the DATA command when the smtp transport
2544 was used in LMTP mode. This meant that they could be retried too
2545 frequently, and not timed out correctly.
2547 (ii) Exim was setting the flag that allows error details to be returned
2548 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2549 recipients that were returned after the DATA command.
2551 PH/53 This is related to PH/52, but is more general: for any failing address,
2552 when detailed error information was permitted to be returned to the
2553 sender, but the error was temporary, then after the final timeout, only
2554 "retry timeout exceeded" was returned. Now it returns the full error as
2555 well as "retry timeout exceeded".
2557 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2558 do this, and (what is worse) MTAs that accept it.
2560 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2561 will now be deprecated.
2563 PH/56 New os.c-cygwin from the Cygwin maintainer.
2565 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2566 formats (previously only available in exim formats via -bpr, -bpru,
2567 and -bpra. Now also available in native and exiqgrep formats)
2569 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2570 with very large, slow to parse queues
2572 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2574 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2576 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2577 responses to authentication challenges, though it was showing the
2578 challenges; (ii) I've removed the CR characters from the debug output for
2581 PH/58 Allow for the insertion of a newline as well as a space when a string
2582 is turned into more than one encoded-word during RFC 2047 encoding. The
2583 Sieve code now uses this.
2585 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2586 data_4xx, lost_connection, tls_required.
2588 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2589 message was being sent as an SMTP response.
2591 PH/61 Add -l and -k options to exicyclog.
2593 PH/62 When verifying, if an address was redirected to one new address, so that
2594 verification continued, and the new address failed or deferred after
2595 having set something in $address_data, the value of $address_data was not
2596 passed back to the ACL. This was different to the case when no
2597 redirection occurred. The value is now passed back in both cases.
2599 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2600 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2601 use login_cap.h, so on its own it isn't the distinguishing feature. The
2602 new name refers directly to the setclassresources() function.
2604 PH/65 Added configuration files for NetBSD3.
2606 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2608 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2609 is preferred over IPv4.
2611 PH/68 The bounce_return_message and bounce_return_body options were not being
2612 honoured for bounces generated during the reception of non-SMTP messages.
2613 In particular, this applied to messages rejected by the ACL. This bug has
2614 been fixed. However, if bounce_return_message is true and bounce_return_
2615 body is false, the headers that are returned for a non-SMTP message
2616 include only those that have been read before the error was detected.
2617 (In the case of an ACL rejection, they have all been read.)
2619 PH/69 The HTML version of the specification is now built in a directory called
2620 spec_html instead of spec.html, because the latter looks like a path with
2621 a MIME-type, and this confuses some software.
2623 PH/70 Catch two compiler warnings in sieve.c.
2625 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2626 function verify_get_ident() calls ip_connect() to connect a socket, but
2627 if the "connect()" function timed out, ip_connect() used to close the
2628 socket. However, verify_get_ident() also closes the socket later, and in
2629 between Exim writes to the log, which may get opened at this point. When
2630 the socket was closed in ip_connect(), the log could get the same file
2631 descriptor number as the socket. This naturally causes chaos. The fix is
2632 not to close the socket in ip_connect(); the socket should be closed by
2633 the function that creates it. There was only one place in the code where
2634 this was missing, in the iplookup router, which I don't think anybody now
2635 uses, but I've fixed it anyway.
2637 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2638 well as to direct DNS lookups. Otherwise the handling of names in host
2639 lists is inconsistent and therefore confusing.
2645 PH/01 Two changes to the default runtime configuration:
2647 (1) Move the checks for relay_from_hosts and authenticated clients from
2648 after to before the (commented out) DNS black list checks.
2650 (2) Add control=submission to the relay_from_hosts and authenticated
2651 clients checks, on the grounds that messages accepted by these
2652 statements are most likely to be submissions.
2654 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2656 (1) Generate an error if the third argument for the ${prvs expansion is
2659 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2662 (3) Reset the variables that are obtained from the first argument of
2663 ${prvscheck and used in the second argument before leaving the code,
2664 because their memory is reclaimed, so using them afterwards may do
2667 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2668 one (it's much easier than Tom thought :-).
2670 (5) Because of (4), we can now allow for the use of $prvscheck_result
2671 inside the third argument.
2673 PH/03 For some reason, the default setting of PATH when running a command from
2674 a pipe transport was just "/usr/bin". I have changed it to
2677 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2678 anything to be listed in the output from -bV.
2680 PH/05 When a filter generated an autoreply, the entire To: header line was
2681 quoted in the delivery log line, like this:
2683 => >A.N.Other <ano@some.domain> <original@ddress> ...
2685 This has been changed so that it extracts the operative address. There
2686 may be more than one such address. If so, they are comma-separated, like
2689 => >ano@some.domain,ona@other.domain <original@ddress> ...
2691 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2692 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2693 not being looked up in the rDNS to get a host name, Exim was showing the
2694 IP address twice in Received: lines, even though the IP addresses were
2695 identical. For example:
2697 Received: from [1.2.3.4] (helo=[1.2.3.4])
2699 However, if the real host name was known, it was omitting the HELO data
2700 if it matched the actual IP address. This has been tidied up so that it
2701 doesn't show the same IP address twice.
2703 PH/07 When both +timestamp and +memory debugging was on, the value given by
2704 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2705 called by the debug printing, thereby overwriting the timestamp buffer.
2706 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2708 PH/08 When the original message was included in an autoreply transport, it
2709 always said "this is a copy of the message, including all the headers",
2710 even if body_only or headers_only was set. It now gives an appropriate
2713 PH/09 Applied a patch from the Sieve maintainer which:
2715 o fixes some comments
2716 o adds the (disabled) notify extension core
2717 o adds some debug output for the result of if/elsif tests
2718 o points to the current vacation draft in the documentation
2719 and documents the missing references header update
2723 o fixes a bug in processing the envelope test (when testing
2724 multiple envelope elements, the last element determinted the
2727 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2728 Electronic Mail") by including:
2730 Auto-submitted: auto-generated
2732 in the messages that it generates (bounce messages and others, such as
2733 warnings). In the case of bounce messages for non-SMTP mesages, there was
2734 also a typo: it was using "Auto_submitted" (underscore instead of
2735 hyphen). Since every message generated by Exim is necessarily in response
2736 to another message, thes have all been changed to:
2738 Auto-Submitted: auto-replied
2740 in accordance with these statements in the RFC:
2742 The auto-replied keyword:
2744 - SHOULD be used on messages sent in direct response to another
2745 message by an automatic process,
2747 - MUST NOT be used on manually-generated messages,
2749 - MAY be used on Delivery Status Notifications (DSNs) and Message
2750 Disposition Notifications (MDNs),
2752 - MUST NOT be used on messages generated by automatic or periodic
2753 processes, except for messages which are automatic responses to
2756 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2757 to the default Received: header definition.
2759 PH/12 Added log selector acl_warn_skipped (default on).
2761 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2762 variables because (a) they are in the wrong storage pool and (b) even if
2763 they were copied, it wouldn't work properly because of the caching.
2765 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2766 checking when decoding. Apparently there are clients that generate
2767 overlong encoded strings. Why am I not surprised?
2769 PH/15 If the first argument of "${if match_address" was not empty, but did not
2770 contain an "@" character, Exim crashed. Now it writes a panic log message
2771 and treats the condition as false.
2773 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2775 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2776 extension "envelope-auth". The code is finished and in agreement with
2777 other implementations, but there is no documentation so far and in fact,
2778 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2779 not changing the active code.
2781 Print executed "if" and "elsif" statements when debugging is used. This
2782 helps a great deal to understand what a filter does.
2784 Document more things not specified clearly in RFC3028. I had all this
2785 sorted out, when out of a sudden new issues came to my mind. Oops."
2787 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2790 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2791 canonical form (as documented). However, after a host name lookup from
2792 the IP address, check_host() was doing a simple string comparison with
2793 addresses acquired from the DNS when checking that the found name did
2794 have the original IP as one of its addresses. Since any found IPv6
2795 addresses are likely to be in abbreviated form, the comparison could
2796 fail. Luckily, there already exists a function for doing the comparison
2797 by converting both addresses to binary, so now that is used instead of
2798 the text comparison.
2800 PH/20 There was another similar case to PH/19, when a complete host name was
2801 given in a host list; looking up its IP address could give an abbreviated
2802 form, whereas the current host's name might or might not be abbreviated.
2803 The same fix has been applied.
2809 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2810 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2813 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2814 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2816 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2818 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2819 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2820 Also, for "Auto-Submitted", treat anything other than "no" as
2821 non-personal, in accordance with RFC 3834. (Previously it treated
2822 anything starting "auto-" as non-personal.)
2824 TF/01 The control=submission/name=... option had a problem with syntax
2825 errors if the name included a slash character. The /name= option
2826 now slurps the rest of the string, so it can include any characters
2827 but it must come last in the list of options (after /sender_retain
2830 PH/05 Some modifications to the interface to the fake nameserver for the new
2838 TK/01 Added the "success_on_redirect" address verification option. See
2839 NewStuff for rationale and an example.
2841 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2843 PH/02 Patch to exigrep to allow it to work on syslog lines.
2845 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2846 fread() to skip over the body file's header line, because in Cygwin the
2847 header line is locked and is inaccessible.
2849 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2850 co-exist for some time) to make it clear that it is the Exim ID that is
2851 referenced, not the Message-ID: header line.
2853 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2854 string_format() function, because snprintf() does not exist on all
2857 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2858 ${stat: expansion item.
2860 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2861 protocol synchronization error", to keep the pedants happy.
2863 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2864 well as for IRIX systems, when gcc is being used. See the host.c source
2867 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2869 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2872 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2873 transport and to the smtp transport in LMTP mode.
2875 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2877 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2878 run a filter in a subprocess. This could lead to confusion in subsequent
2879 lookups in the parent process. There should also be a search_tidyup() at
2880 the end of the subprocess.
2882 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2883 only if the host matched helo_try_verify_hosts, which caused the
2884 verification to occur when the EHLO/HELO command was issued. The ACL just
2885 tested the remembered result. Now, if a previous verification attempt has
2886 not happened, "verify = helo" does it there and then.
2888 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2890 TK/03 Fix log output including CR from clamd.
2892 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2893 exist provoked a memory error which could cause a segfault.
2895 PH/15 Installed PCRE 6.2
2897 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2899 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2900 of the problem. Specifically, suggested +O2 rather than +O1 for the
2903 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2905 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2906 record, the port was not correctly logged when the outgoing_port log
2907 selector was set (it logged the transort's default port).
2909 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2910 fallback_hosts, and "hosts" in the smtp transport.
2912 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2913 host errors such as "Connection refused".
2915 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2916 authentication with radiusclient 0.4.9:
2918 - Error returned from rc_read_config was caught wrongly
2919 - Username/password not passed on to radius server due to wrong length.
2921 The presumption is that some radiusclient API changes for 4.51/PH/17
2922 were not taken care of correctly. The code is still untested by me (my
2923 Linux distribution still has 0.3.2 of radiusclient), but it was
2924 contributed by a Radius user.
2926 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2927 expanding the "port" option of the smtp transport.
2929 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2930 while reading a MIME header. Thanks to Tom Hughes for a patch.
2932 PH/24 Include config.h inside local_scan.h so that configuration settings are
2935 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2936 that in an ACL for RCPT (for example), you can examine exactly what was
2939 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2940 commands, but it was not correctly comparing the address with the actual
2941 client host address. Thus, it would show the EHLO address in Received:
2942 header lines when this was not necessary.
2944 PH/27 Added the % operator to ${eval:}.
2946 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2947 it should be ignoring failures (because with -C, for example, it has lost
2948 privilege). It wasn't ignoring creation failures other than "already
2951 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2954 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2955 because an input line was too long, either on its own, or by virtue of
2956 too many continuations, the temporary file was not being removed, and the
2957 return code was incorrect.
2959 PH/31 Missing "BOOL" in function definition in filtertest.c.
2961 PH/32 Applied Sieve patches from the maintainer.
2963 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2965 PH/33 Added "verify = not_blind".
2967 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2968 Local/Makefile (with some defaults set). These are used in built scripts
2969 such as exicyclog, but they have never been used in the exim_install
2970 script (though there are many overriding facilities there). I have
2971 arranged that the exim_install script now takes note of these two
2974 PH/35 Installed configuration files for Dragonfly.
2976 PH/36 When a locally submitted message by a trusted user did not contain a
2977 From: header, and the sender address was obtained from -f or from an SMTP
2978 MAIL command, and the trusted user did not use -F to supply a sender
2979 name, $originator_name was incorrectly used when constructing a From:
2980 header. Furthermore, $originator_name was used for submission mode
2981 messages from external hosts without From: headers in a similar way,
2982 which is clearly wrong.
2984 PH/37 Added control=suppress_local_fixups.
2986 PH/38 When log_selector = +received_sender was set, and the addition of the
2987 sender made the log line's construction buffer exactly full, or one byte
2988 less than full, an overflow happened when the terminating "\n" was
2991 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2992 when the result of a list match is failure because a DNS lookup failed.
2994 PH/40 RM_COMMAND is now used in the building process.
2996 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2997 the "build-* directories that it finds.
2999 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3000 address in a domain literal was a prefix of an interface address.
3002 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3003 when verifying a sender address, unless rewrite_headers is false.
3005 PH/44 Wrote a long comment about why errors_to addresses are verified as
3006 recipients, not senders.
3008 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3009 the ratelimit ACL was added.
3011 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3013 PH/46 Added extra information about PostgreSQL errors to the error string.
3015 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3016 avoiding the need to install special zones in a real server. This is
3017 backwards compatible; if it can't find the fake resolver, it drops back.
3018 Thus, both old and new test suites can be run.
3020 TF/02 Added util/ratelimit.pl
3022 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3023 clock is set back in time.
3025 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3026 Candler <B.Candler@pobox.com>.
3028 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3029 OK for addresses that are the result of redirections.
3031 PH/48 A number of further additions for the benefit of the new test suite,
3032 including a fake gethostbyname() that interfaces to the fake DNS resolver
3035 TF/06 The fix for widen_domains has also been applied to qualify_single and
3036 search_parents which are the other dnslookup options that can cause
3039 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3042 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3044 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3045 service" if it encountered an MX record with an empty target hostname.
3046 The message is now "an MX or SRV record indicated no SMTP service".
3048 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3049 if the DNS of the sending site is misconfigured. This is quite a
3050 common situation. This change restores the behaviour of treating a
3051 helo verification defer as a failure.
3053 PH/51 If self=fail was set on a router, the bounce message did not include the
3054 actual error message.
3060 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3062 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3063 command itself ended in error, the underlying message about the transport
3064 filter timeout was being overwritten with the pipe command error. Now the
3065 underlying error message should be appended to the second error message.
3067 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3069 PH/02 Reduce the amount of output that "make" produces by default. Full output
3070 can still be requested.
3072 PH/03 The warning log line about a condition test deferring for a "warn" verb
3073 was being output only once per connection, rather than after each
3074 occurrence (because it was using the same function as for successful
3075 "warn" verbs). This seems wrong, so I have changed it.
3077 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3078 it should not have, which might have caused a crash in the right
3079 circumstances, but probably never did.
3081 PH/04 Installed a modified version of Tony Finch's patch to make submission
3082 mode fix the return path as well as the Sender: header line, and to
3083 add a /name= option so that you can make the user's friendly name appear
3086 TF/03 Added the control = fakedefer ACL modifier.
3088 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3089 Mark Lowes for thorough testing.
3091 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3093 TK/03 Merged latest SRS patch from Miles Wilton.
3095 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3096 with the definition in sysexits.h (which is #included earlier).
3097 Fortunately, Exim does not actually use EX_OK. The code used to try to
3098 preserve the sysexits.h value, by assumimg that macro definitions were
3099 scanned for macro replacements. I have been disabused of this notion,
3100 so now the code just undefines EX_OK before #including unistd.h.
3102 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3103 in the smtp transport. When a block could not be written in a single
3104 write() function, the timeout was being re-applied to each part-write.
3105 This seems wrong - if the receiver was accepting one byte at a time it
3106 would take for ever. The timeout is now adjusted when this happens. It
3107 doesn't have to be particularly precise.
3109 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3110 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3112 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3113 without a domain if the check to <postmaster@domain> fails.
3115 SC/01 Eximstats: added -xls and the ability to specify output files
3116 (patch written by Frank Heydlauf).
3118 SC/02 Eximstats: use FileHandles for outputing results.
3120 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3122 SC/04 Eximstats: fixed display of large numbers with -nvr option
3124 SC/05 Eximstats: fixed merging of reports with empty tables.
3126 SC/06 Eximstats: added the -include_original_destination flag
3128 SC/07 Eximstats: removed tabs and trailing whitespace.
3130 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3132 TK/06 MBOX spool code: Add real "From " MBOX separator line
3133 so the .eml file is really in mbox format (even though
3134 most programs do not really care). Patch from Alex Miller.
3136 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3137 The latter is generated from $received_to and is only set if the
3138 message has one envelope recipient. SA can use these headers,
3139 obviously out-of-the-box. Patch from Alex Miller.
3141 PH/08 The ${def test on a variable was returning false if the variable's
3142 value was "0", contrary to what the specification has always said!
3143 The result should be true unless the variable is empty.
3145 PH/09 The syntax error of a character other than { following "${if
3146 def:variable_name" (after optional whitespace) was not being diagnosed.
3147 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3148 accidental colon was present, for example, could give incorrect results.
3150 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3151 result is used (not including appendfile, where other changes are about
3154 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3155 This involved changing a lot of size variables from int to off_t. It
3156 should work with maildirs and everything.
3158 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3159 spamd dying while we are connected to it.
3161 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3164 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3165 function for BDB 4.3.
3167 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3169 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3170 the sources for PCRE have been re-arranged and more files are now
3173 PH/15 The code I had for printing potentially long long variables in PH/11
3174 above was not the best (it lost precision). The length of off_t variables
3175 is now inspected at build time, and an appropriate printing format (%ld
3176 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3177 to be "long long int" or "long int". This is needed for the internal
3178 formatting function string_vformat().
3180 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3181 the configuration file to be ":syslog", then the script "guesses" where
3182 the logs files are, rather than using the compiled in default. In our
3183 case the guess is not the same as the compiled default, so the script
3184 suddenly stopped working when I started to use syslog. The patch checks
3185 to see if log_file_path is "". If so, it attempts to read it from exim
3186 with no configuration file to get the compiled in version, before it
3187 falls back to the previous guessing code."
3189 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3190 implementing BATV in an Exim configuration. See NewStuff for the gory
3193 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3194 Makefile that are specific to HP-UX.
3196 PH/18 If the "use_postmaster" option was set for a recipient callout together
3197 with the "random" option, the postmaster address was used as the MAIL
3198 FROM address for the random test, but not for the subsequent recipient
3199 test. It is now used for both.
3201 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3202 patch removes a few documentation additions to RFC 3028, because the
3203 latest draft now contains them. It adds the new en;ascii-case comparator
3204 and a new error check for 8bit text in MIME parts. Comparator and
3205 require names are now matched exactly. I enabled the subaddress
3206 extension, but it is not well tested yet (read: it works for me)."
3208 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3209 rework some of the code of TK/09 above to avoid the hardwired use of
3210 "%lld" and "long long". Replaced the call to snprintf() with a call to
3213 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3214 records point to non-existent hosts", "retry timeout exceeded", and
3215 "retry time not reached for any host after a long failure period".
3217 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3218 experimental DomainKeys support:
3220 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3221 (2) On an error such as an illegally used "control", the wrong name for
3222 the control was given.
3224 These problems did NOT occur unless DomainKeys support was compiled.
3226 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3228 PH/24 Added ${if match_ip condition.
3230 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3231 databases so that it will be absolutely obvious if a crash occurs in the
3232 DB library. This is a regular occurrence (often caused by mis-matched
3235 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3236 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3237 picked up on a user's system that detects such things. There doesn't seem
3238 to be a gcc warning option for this - only an attribute that has to be
3239 put on the function's prototype. It seems that in Fedora Core 4 they have
3240 set this on a number of new functions. No doubt there will be more in due
3243 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3244 specify a transport. However, if an address that was verified by such a
3245 router was the subject of a callout, Exim crashed because it tried to
3246 read the rcpt_include_affixes from the non-existent transport. Now it
3247 just assumes that the setting of that option is false. This bug was
3248 introduced by 4.51/PH/31.
3250 PH/28 Changed -d+all to exclude +memory, because that information is very
3251 rarely of interest, but it makes the output a lot bigger. People tend to
3252 do -d+all out of habit.
3254 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3255 code in os-type was giving problems when libc.so lives in lib64, like on
3258 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3259 aren't the modern standard, and it seems that some systems' include files
3260 don't always have them. Exim was already checking for some of the newer
3261 ones like T_AAAA, and defining it itself. I've added checks for all the
3262 record types that Exim uses.
3264 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3265 not automatically generating a new one, as it is supposed to. This
3266 prevented TLS from working. If the file did exist, but contained invalid
3267 data, a new version was generated, as expected. It was only the case of a
3268 non-existent file that was broken.
3270 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3271 with a change in libdomainkeys > 0.64.
3273 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3274 from DNS. If the selector record carries the flag, it now has
3275 precedence over the domain-wide flag.
3277 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3279 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3280 the use of a port name that isn't defined in /etc/services) occurred, the
3281 message was deferred as in a normal delivery, and thus remained on the
3282 spool, instead of being failed because of the mua_wrapper setting. This
3283 is now fixed, and I tidied up some of the mua_wrapper messages at the
3286 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3287 the messages in a hash of arrays rather than using individual hashes.
3288 This is a bit cleaner and results in dramatic memory savings, albeit
3289 at a slight CPU cost.
3291 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3292 as requested by Marc Sherman.
3294 SC/10 Eximstats: added histograms for user specified patterns as requested
3297 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3299 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3300 fopen() in the content-scanning modules that did not already have it.
3306 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3307 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3309 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3311 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3313 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3314 to test Sieve filters that use "vacation".
3316 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3317 that changes the way the GnuTLS parameters are stored in the cache file.
3318 The new format can be generated externally. For backward compatibility,
3319 if the data in the cache doesn't make sense, Exim assumes it has read an
3320 old-format file, and it generates new data and writes a new file. This
3321 means that you can't go back to an older release without removing the
3324 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3325 work if there are any delivery delays because "one_time" forces the
3326 parent to be marked "delivered", so its unseen clone is never tried
3327 again. For this reason, Exim now forbids the simultaneous setting of
3330 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3331 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3332 entry if you want to know the details. The fix, however, neglected to
3333 consider the case when local delivery batching is involved. The test for
3334 "previously delivered" was not happening when checking to see if an
3335 address could be batched with a previous (undelivered) one; under
3336 certain circumstances this could lead to multiple deliveries to the same
3339 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3340 in its include files, and this causes problems building Exim.
3342 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3343 header_syntax) but Exim was just ignoring anything given after a slash.
3344 In particular, this caused confusion with an attempt to use "verify =
3345 reverse_host_lookup/defer_ok". An error is now given when options are
3346 supplied for verify items that do not have them. (Maybe reverse_host_
3347 lookup should have a defer_ok option, but that's a different point.)
3349 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3350 defined by RFC 821) to 2048, because there were problems with some AUTH
3351 commands, and RFC 1869 says the size should be increased for extended
3352 SMTP commands that take arguments.
3354 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3357 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3358 "unknown" error; now it says that the functionality isn't in the binary.
3360 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3361 an address' error message when a string expansion fails (syntax or
3362 whatever). Otherwise the password may appear in the log. Following change
3363 PH/42 below, there is no longer a chance of it appearing in a bounce
3366 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3368 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3369 of its addresses was ever tried. (Bugzilla bug #2.)
3371 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3372 the result incorrectly in the debug output. (It correctly added a newline
3373 to what was transported.)
3375 TF/01 Added $received_time.
3377 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3378 commented out examples of how to interface to a virus scanner and to
3379 SpamAssassin. Also added commented examples of av_scanner and
3380 spamd_address settings.
3382 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3383 and controls are allowed in which ACLs. There were a couple of minor
3384 errors. Some of the entries in the conditions table (which is a table of
3385 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3386 negation of where the condition IS allowed.
3388 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3390 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3391 header file does not have a version number, so I've had to invent a new
3392 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3393 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3394 radiusclient), but it was contributed by a Radius user.
3396 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3397 files or directories, for OpenSSL.
3399 PH/19 When an Exim process that is running as root has to create an Exim log
3400 file, it does so in a subprocess that runs as exim:exim so as to get the
3401 ownership right at creation (otherwise, other Exim processes might see
3402 the file with the wrong ownership). There was no test for failure of this
3403 fork() call, which would lead to the process getting stuck as it waited
3404 for a non-existent subprocess. Forks do occasionally fail when resources
3405 run out. I reviewed all the other calls to fork(); they all seem to check
3408 PH/20 When checking for unexpected SMTP input at connect time (before writing
3409 the banner), Exim was not dealing correctly with a non-positive return
3410 from the read() function. If the client had disconnected by this time,
3411 the result was a log entry for a synchronization error with an empty
3412 string after "input=" when read() returned zero. If read() returned -1
3413 (an event I could not check), uninitialized data bytes were printed.
3414 There were reports of junk text (parts of files, etc) appearing after
3417 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3419 PH/22 Added support for macro redefinition, and (re)definition in between
3420 driver and ACL definitions.
3422 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3423 forgetting to use the resulting value; it was using the unexpanded value.
3425 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3426 hadn't been configured. The fix is from Juergen Kreileder, who
3427 understands it better than I do:
3429 "Here's what I see happening with three configured cyrus_sasl
3430 authenticators configured (plain, login, cram-md5):
3432 On startup auth_cyrus_sasl_init() gets called for each of these.
3433 This means three calls to sasl_listmech() without a specified mech_list.
3434 => SASL tests which mechs of all available mechs actually work
3435 => three warnings about OTP not working
3436 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3438 With the patch, sasl_listmech() also gets called three times. But now
3439 SASL's mech_list option is set to the server_mech specified in the the
3440 authenticator. Or in other words, the answer from sasl_listmech()
3441 gets limited to just the mech you're testing for (which is different
3443 => the return list contains just 'plain' or 'login', 'cram-md5' or
3444 nothing depending on the value of ob->server_mech.
3446 I've just tested the patch: Authentication still works fine,
3447 unavailable mechs specified in the exim configuration are still
3448 caught, and the auth.log warnings about OTP are gone."
3450 PH/25 When debugging is enabled, the contents of the command line are added
3451 to the debugging output, even when log_selector=+arguments is not
3454 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3455 answer is "GNU", and only if the return is "GNU/something" is the answer
3458 PH/27 $acl_verify_message is now set immediately after the failure of a
3459 verification in an ACL, and so is available in subsequent modifiers. In
3460 particular, the message can be preserved by coding like this:
3462 warn !verify = sender
3463 set acl_m0 = $acl_verify_message
3465 Previously, $acl_verify_message was set only while expanding "message"
3466 and "log_message" when a very denied access.
3468 PH/28 Modified OS/os.c-Linux with
3470 -#ifndef OS_LOAD_AVERAGE
3471 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3473 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3474 nomenclature these days.)
3476 PH/29 Installed patch from the Sieve maintainer that adds the options
3477 sieve_useraddress and sieve_subaddress to the redirect router.
3479 PH/30 In these circumstances:
3480 . Two addresses routed to the same list of hosts;
3481 . First host does not offer TLS;
3482 . First host accepts first address;
3483 . First host gives temporary error to second address;
3484 . Second host offers TLS and a TLS session is established;
3485 . Second host accepts second address.
3486 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3487 and peerdn, if requested) that were in fact used only for the second
3490 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3491 attention to any local part prefix or suffix that was matched by the
3492 router that accepted the address. It now behaves in the same way as it
3493 does for delivery: the affixes are removed from the local part unless
3494 rcpt_include_affixes is set on the transport.
3496 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3497 timeout during the DATA phase of an incoming message.
3499 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3500 applied a patch sanctioned by the Sieve maintainer.
3502 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3503 the uid or gid is negative. A case of a negative gid caused this to be
3504 noticed. The fix allows for either to be negative.
3506 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3507 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3510 PH/36 The change PH/12 above was broken. Fixed it.
3512 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3513 the grounds that routing the same address twice would always produce the
3514 same answer. This might have been true once, but it is certainly no
3515 longer true now. Routing a child address may depend on the previous
3516 routing that produced that child. Some complicated redirection strategies
3517 went wrong when messages had multiple recipients, and made Exim's
3518 behaviour dependent on the order in which the addresses were given.
3520 I have moved the duplicate checking until after the routing is complete.
3521 Exim scans the addresses that are assigned to local and remote
3522 transports, and removes any duplicates. This means that more work will be
3523 done, as duplicates will always all be routed, but duplicates are
3524 presumably rare, so I don't expect this is of any significance.
3526 For deliveries to pipes, files, and autoreplies, the duplicate checking
3527 still happens during the routing process, since they are not going to be
3530 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3531 It corrects a timeout issue with spamd. This is Ian's comment: "The
3532 background is that sometimes spamd either never reads data from a
3533 connection it has accepted, or it never writes response data. The exiscan
3534 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3535 blindly assumes that writes won't block so it may never time out."
3537 PH/39 Allow G after quota size as well as K and M.
3539 PH/40 The value set for $authenticated_id in an authenticator may not contain
3540 binary zeroes or newlines because the value is written to log lines and
3541 to spool files. There was no check on this. Now the value is run through
3542 the string_printing() function so that such characters are converted to
3543 printable escape sequences.
3545 PH/41 $message_linecount is a new variable that contains the total number of
3546 lines in the message. Compare $body_linecount, which is the count for the
3549 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3550 bounce and delay warning messages, except in certain special cases, which
3553 (a) An SMTP error message from a remote host;
3554 (b) A message specified in a :fail: redirection;
3555 (c) A message specified in a "fail" command in a system filter;
3556 (d) A message specified in a FAIL return from the queryprogram router;
3557 (e) A message specified by the cannot_route_message router option.
3559 In these cases only, Exim does include the error details in bounce and
3560 warning messages. There are also a few cases where bland messages such
3561 as "unrouteable address" or "local delivery error" are given.
3563 PH/43 $value is now also set for the "else" part of a ${run expansion.
3565 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3566 being worked on, but at least Exim now implements the latest version to
3569 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3570 process to complete was treated as a delivery failure, a timeout while
3571 writing the message to the pipe was logged, but erroneously treated as a
3572 successful delivery. Such timeouts include transport filter timeouts. For
3573 consistency with the overall process timeout, these timeouts are now
3574 treated as errors, giving rise to delivery failures by default. However,
3575 there is now a new Boolean option for the pipe transport called
3576 timeout_defer, which, if set TRUE, converts the failures into defers for
3577 both kinds of timeout. A transport filter timeout is now identified in
3580 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3581 systems where "make" and "gmake" are different, calling "gmake" at top
3582 level broke things. I've arranged for the value of $(MAKE) to be passed
3583 from the Makefile to this script so that it can call the same version of
3587 A note about Exim versions 4.44 and 4.50
3588 ----------------------------------------
3590 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3591 changes of various kinds. As a consequence, a big documentation update was
3592 needed. This delayed the release for rather longer than seemed good, especially
3593 in the light of a couple of (minor) security issues. Therefore, the changes
3594 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3595 release. So 4.44 and 4.50 are in effect two different branches that both start
3598 I have left the 4.50 change log unchanged; it contains all the changes since
3599 4.43. The change log for 4.44 is below; many of its items are identical to
3600 those for 4.50. This seems to be the most sensible way to preserve the
3601 historical information.
3607 1. Minor wording change to the doc/README.SIEVE file.
3609 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3610 computation of the current number of files was incorrect.
3612 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3613 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3616 4. Give more explanation in the error message when the command for a transport
3617 filter fails to execute.
3619 5. There are several places where Exim runs a non-Exim command in a
3620 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3621 was being done only for the command run by the queryprogram router. It is
3622 now done for all such subprocesses. The other cases are: ${run, transport
3623 filters, and the commands run by the lmtp and pipe transports.
3625 6. Added CONFIGURE_GROUP build-time option.
3627 7. Some older OS have a limit of 256 on the maximum number of file
3628 descriptors. Exim was using setrlimit() to set 1000 as a large value
3629 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3630 systems. I've change it so that if it can't get 1000, it tries for 256.
3632 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3633 was an oversight, and furthermore, ever since the addition of extra
3634 controls (e.g. 4.43/32), the checks on when to allow different forms of
3635 "control" were broken. There should now be diagnostics for all cases when a
3636 control that does not make sense is encountered.
3638 9. Added the /retain_sender option to "control=submission".
3640 10. $recipients is now available in the predata ACL (oversight).
3642 11. Tidy the search cache before the fork to do a delivery from a message
3643 received from the command line. Otherwise the child will trigger a lookup
3644 failure and thereby defer the delivery if it tries to use (for example) a
3645 cached ldap connection that the parent has called unbind on.
3647 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3648 of $address_data from the recipient verification was clobbered by the
3649 sender verification.
3651 13. The value of address_data from a sender verification is now available in
3652 $sender_address_data in subsequent conditions in the ACL statement.
3654 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3656 15. Added a new option "connect=<time>" to callout options, to set a different
3659 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3660 was its contents. (It was OK if the option was not defined at all.)
3662 17. A "Completed" log line is now written for messages that are removed from
3663 the spool by the -Mrm option.
3665 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3666 information about exactly what failed.
3668 19. Added -dd to debug only the daemon process.
3670 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3671 handles timeouts, both on the server side and network timeouts. Renamed the
3672 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3674 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3675 It is now set to "smtps".
3677 22. $host_address is now set to the target address during the checking of
3678 ignore_target_hosts.
3680 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3681 being passed; this would have caused $sender_host_name to have been used if
3682 matching the list had actually called for a host name (not very likely,
3683 since this list is usually IP addresses). A host name is now passed as
3686 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3687 code that specifies a non-restarting handler (typically sigaction() in
3688 modern systems) in an attempt to fix a rare and obscure crash bug.
3690 25. Narrowed the window for a race in the daemon that could cause it to ignore
3691 SIGCHLD signals. This is not a major problem, because they are used only to
3692 wake it up if nothing else does.
3694 26. A malformed maildirsize file could cause Exim to calculate negative values
3695 for the mailbox size or file count. Odd effects could occur as a result.
3696 The maildirsize information is now recalculated if the size or filecount
3699 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3700 support for a long time. Removed HAVE_SYS_VFS_H.
3702 28. Installed the latest version of exipick from John Jetmore.
3704 29. In an address list, if the pattern was not a regular expression, an empty
3705 subject address (from a bounce message) matched only if the pattern was an
3706 empty string. Non-empty patterns were not even tested. This was the wrong
3707 because it is perfectly reasonable to use an empty address as part of a
3708 database query. An empty address is now tested by patterns that are
3709 lookups. However, all the other forms of pattern expect the subject to
3710 contain a local part and a domain, and therefore, for them, an empty
3711 address still always fails if the pattern is not itself empty.
3713 30. Exim went into a mad DNS loop when attempting to do a callout where the
3714 host was specified on an smtp transport, and looking it up yielded more
3715 than one IP address.
3717 31. Re-factored the code for checking spool and log partition space into a
3718 function that finds that data and another that does the check. The former
3719 is then used to implement four new variables: $spool_space, $log_space,
3720 $spool_inodes, and $log_inodes.
3722 32. The RFC2047 encoding function was originally intended for short strings
3723 such as real names; it was not keeping to the 75-character limit for
3724 encoded words that the RFC imposes. It now respects the limit, and
3725 generates multiple encoded words if necessary. To be on the safe side, I
3726 have increased the buffer size for the ${rfc2047: expansion operator from
3729 33. It is now permitted to omit both strings after an "if" condition; if the
3730 condition is true, the result is "true". As before, when the second string
3731 is omitted, a false condition yields an empty string. This makes it less
3732 cumbersome to write custom ACL and router conditions.
3734 34. Failure to deliver a bounce message always caused it to be frozen, even if
3735 there was an errors_to setting on the router. The errors_to setting is now
3738 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3739 canonical form (fully expanded) before being placed in
3740 $sender_host_address.
3742 36. The table in the code that translates DNS record types into text (T_A to
3743 "A" for instance) was missing entries for NS and CNAME. It is just possible
3744 that this could have caused confusion if both these types were looked up
3745 for the same domain, because the text type is used as part of Exim's
3746 per-process caching. But the chance of anyone hitting this buglet seems
3749 37. The dnsdb lookup has been extended in a number of ways.
3751 (1) There is a new type, "zns", which walks up the domain tree until it
3752 finds some nameserver records. It should be used with care.
3754 (2) There is a new type, "mxh", which is like "mx" except that it returns
3755 just the host names, not the priorities.
3757 (3) It is now possible to give a list of domains (or IP addresses) to be
3758 looked up. The behaviour when one of the lookups defers can be
3759 controlled by a keyword.
3761 (4) It is now possible to specify the separator character for use when
3762 multiple records are returned.
3764 38. The dnslists ACL condition has been extended: it is now possible to supply
3765 a list of IP addresses and/or domains to be looked up in a particular DNS
3768 39. Added log_selector=+queue_time_overall.
3770 40. When running the queue in the test harness, wait just a tad after forking a
3771 delivery process, to get repeatability of debugging output.
3773 41. Include certificate and key file names in error message when GnuTLS fails
3774 to set them up, because the GnuTLS error message doesn't include the name
3775 of the failing file when there is a problem reading it.
3777 42. Allow both -bf and -bF in the same test run.
3779 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3781 44. The "Exiscan patch" is now merged into the mainline Exim source.
3783 45. Sometimes the final signoff response after QUIT could fail to get
3784 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3785 < 0 before doing a fflush(). This bug looks as though it goes back to the
3786 introduction of TLS in release 3.20, but "sometimes" must have been rare
3787 because the tests only now provoked it.
3789 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3790 (this can affect the format of dates).
3792 47. exim_tidydb, when checking for the continued existence of a message for
3793 which it has found a message-specific retry record, was not finding
3794 messages that were in split spool directories. Consequently, it was
3795 deleting retry records that should have stayed in existence.
3797 48. Steve fixed some bugs in eximstats.
3799 49. The SPA authentication driver was not abandoning authentication and moving
3800 on to the next authenticator when an expansion was forced to fail,
3801 contradicting the general specification for all authenticators. Instead it
3802 was generating a temporary error. It now behaves as specified.
3804 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3805 (the order specifies the preference for clients). The order is now AES256,
3806 AES128, 3DES, ARCFOUR128.
3808 51. Small patch to Sieve code - explicitly set From: when generating an
3811 52. Exim crashed if a remote delivery caused a very long error message to be
3812 recorded - for instance if somebody sent an entire SpamAssassin report back
3813 as a large number of 550 error lines. This bug was coincidentally fixed by
3814 increasing the size of one of Exim's internal buffers (big_buffer) that
3815 happened as part of the Exiscan merge. However, to be on the safe side, I
3816 have made the code more robust (and fixed the comments that describe what
3819 53. Now that there can be additional text after "Completed" in log lines (if
3820 the queue_time_overall log selector is set), a one-byte patch to exigrep
3821 was needed to allow it to recognize "Completed" as not the last thing in
3824 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3825 patch that reportedly fixes this has been added. I am not expert enough to
3826 create a test for it. This is what the patch creator wrote:
3828 "I found a little strange behaviour of ldap code when working with
3829 Windows 2003 AD Domain, where users was placed in more than one
3830 Organization Units. When I tried to give exim partial DN, the exit code
3831 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3832 But simultaneously result of request was absolutely normal ldap result,
3833 so I produce this patch..."
3835 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3836 REFERENCE, so I have modified the code to exclude the patch when that macro
3839 55. Some experimental protocols are using DNS PTR records for new purposes. The
3840 keys for these records are domain names, not reversed IP addresses. The
3841 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3842 leaves it alone. Component reversal etc. now happens only for IP addresses.
3845 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3847 57. Double the size of the debug message buffer (to 2048) so that more of very
3848 long debug lines gets shown.
3850 58. The exicyclog utility now does better if the number of log files to keep
3851 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3853 59. Two changes related to the smtp_active_hostname option:
3855 (1) $smtp_active_hostname is now available as a variable.
3856 (2) The default for smtp_banner uses $smtp_active_hostname instead
3857 of $primary_hostname.
3859 60. The host_aton() function is supposed to be passed a string that is known
3860 to be a valid IP address. However, in the case of IPv6 addresses, it was
3861 not checking this. This is a hostage to fortune. Exim now panics and dies
3862 if the condition is not met. A case was found where this could be provoked
3863 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3864 components; fortuitously, this particular loophole had already been fixed
3865 by change 4.50/55 above.
3867 If there are any other similar loopholes, the new check in host_aton()
3868 itself should stop them being exploited. The report I received stated that
3869 data on the command line could provoke the exploit when Exim was running as
3870 exim, but did not say which command line option was involved. All I could
3871 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3872 running as the user.
3875 61. There was a buffer overflow vulnerability in the SPA authentication code
3876 (which came originally from the Samba project). I have added a test to the
3877 spa_base64_to_bits() function which I hope fixes it.
3880 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3881 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3883 63. The daemon start-up calls getloadavg() while still root for those OS that
3884 need the first call to be done as root, but it missed one case: when
3885 deliver_queue_load_max is set with deliver_drop_privilege. This is
3886 necessary for the benefit of the queue runner, because there is no re-exec
3887 when deliver_drop_privilege is set.
3889 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3890 This has been fixed.
3892 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3893 was in use, was not putting the data itself into the right store pool;
3894 consequently, it could be overwritten for a subsequent message in the same
3895 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3898 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3900 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3901 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3902 the function were treating the return as a boolean value, which happened to
3903 work because 0=false and not-0=true, but is not correct code.
3905 68. The host_aton() function was not handling scoped IPv6 addresses (those
3906 with, for example, "%eth0" on the end) correctly.
3908 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3909 negated items (that is, ~something) in unsigned ints. Some compilers
3910 apparently mutter when there is no cast.
3912 70. If an address verification called from an ACL failed, and did not produce a
3913 user-specific message (i.e. there was only a "system" message), nothing was
3914 put in $acl_verify_message. In this situation, it now puts the system
3917 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3918 session; change 4.31/43 added the unwanted input to the log line - except
3919 that it did not do this in the start of session case. It now does.
3921 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3922 This is wrong and can cause the other end to generate a synchronization
3923 error if it is another Exim or anything else that does the synchronization
3924 check. A QUIT command is no longer sent after a timeout.
3926 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3927 during host lookups.
3929 74. The defer_ok option of callout verification was not working if it was used
3930 when verifying addresses in header lines, that is, for this case:
3932 verify = header_sender/callout=defer_ok
3934 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3935 those file descriptors could be used for SMTP connections. If anything
3936 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3937 could be sent to the SMTP client, causing chaos. The daemon now opens
3938 stdin, stdout, and stderr to /dev/null when it puts itself into the
3941 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3942 log by default. The user can override this with suitable Perl magic.
3944 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3945 the log message when discard triggers, was not working for the DATA ACL or
3946 for the non-SMTP ACL.
3948 78. Error message wording change in sieve.c.
3950 79. If smtp_accept_max_per_host was set, the number of connections could be
3951 restricted to fewer than expected, because the daemon was trying to set up
3952 a new connection before checking whether the processes handling previous
3953 connections had finished. The check for completed processes is now done
3954 earlier. On busy systems, this bug wouldn't be noticed because something
3955 else would have woken the daemon, and it would have reaped the completed
3958 80. If a message was submitted locally by a user whose login name contained one
3959 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3960 It caused a spool format error. I have fixed the spool reading code. A
3961 related problem was that the "from" clause in the Received: line became
3962 illegal because of the space(s). It is now covered by ${quote_local_part.
3964 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3967 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3969 83. Patch from Sieve maintainer to fix unterminated string problem in
3970 "vacation" handling.
3972 84. Some minor changes to the Linux configuration files to help with other
3973 OS variants using glibc.
3975 85. One more patch for Sieve to update vacation handling to latest spec.
3978 ----------------------------------------------------
3979 See the note above about the 4.44 and 4.50 releases.
3980 ----------------------------------------------------
3986 1. Change 4.43/35 introduced a bug that caused file counts to be
3987 incorrectly computed when quota_filecount was set in an appendfile
3990 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3991 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3994 3. Give more explanation in the error message when the command for a transport
3995 filter fails to execute.
3997 4. There are several places where Exim runs a non-Exim command in a
3998 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3999 was being done only for the command run by the queryprogram router. It is
4000 now done for all such subprocesses. The other cases are: ${run, transport
4001 filters, and the commands run by the lmtp and pipe transports.
4003 5. Some older OS have a limit of 256 on the maximum number of file
4004 descriptors. Exim was using setrlimit() to set 1000 as a large value
4005 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4006 systems. I've change it so that if it can't get 1000, it tries for 256.
4008 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4009 was an oversight, and furthermore, ever since the addition of extra
4010 controls (e.g. 4.43/32), the checks on when to allow different forms of
4011 "control" were broken. There should now be diagnostics for all cases when a
4012 control that does not make sense is encountered.
4014 7. $recipients is now available in the predata ACL (oversight).
4016 8. Tidy the search cache before the fork to do a delivery from a message
4017 received from the command line. Otherwise the child will trigger a lookup
4018 failure and thereby defer the delivery if it tries to use (for example) a
4019 cached ldap connection that the parent has called unbind on.
4021 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4022 of $address_data from the recipient verification was clobbered by the
4023 sender verification.
4025 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4026 was its contents. (It was OK if the option was not defined at all.)
4028 11. A "Completed" log line is now written for messages that are removed from
4029 the spool by the -Mrm option.
4031 12. $host_address is now set to the target address during the checking of
4032 ignore_target_hosts.
4034 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4035 being passed; this would have caused $sender_host_name to have been used if
4036 matching the list had actually called for a host name (not very likely,
4037 since this list is usually IP addresses). A host name is now passed as
4040 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4041 code that specifies a non-restarting handler (typically sigaction() in
4042 modern systems) in an attempt to fix a rare and obscure crash bug.
4044 15. Narrowed the window for a race in the daemon that could cause it to ignore
4045 SIGCHLD signals. This is not a major problem, because they are used only to
4046 wake it up if nothing else does.
4048 16. A malformed maildirsize file could cause Exim to calculate negative values
4049 for the mailbox size or file count. Odd effects could occur as a result.
4050 The maildirsize information is now recalculated if the size or filecount
4053 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4054 support for a long time. Removed HAVE_SYS_VFS_H.
4056 18. Updated exipick to current release from John Jetmore.
4058 19. Allow an empty sender to be matched against a lookup in an address list.
4059 Previously the only cases considered were a regular expression, or an
4062 20. Exim went into a mad DNS lookup loop when doing a callout where the
4063 host was specified on the transport, if the DNS lookup yielded more than
4066 21. The RFC2047 encoding function was originally intended for short strings
4067 such as real names; it was not keeping to the 75-character limit for
4068 encoded words that the RFC imposes. It now respects the limit, and
4069 generates multiple encoded words if necessary. To be on the safe side, I
4070 have increased the buffer size for the ${rfc2047: expansion operator from
4073 22. Failure to deliver a bounce message always caused it to be frozen, even if
4074 there was an errors_to setting on the router. The errors_to setting is now
4077 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4078 canonical form (fully expanded) before being placed in
4079 $sender_host_address.
4081 24. Updated eximstats to version 1.33
4083 25. Include certificate and key file names in error message when GnuTLS fails
4084 to set them up, because the GnuTLS error message doesn't include the name
4085 of the failing file when there is a problem reading it.
4087 26. Expand error message when OpenSSL has problems setting up cert/key files.
4090 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4091 (this can affect the format of dates).
4093 28. exim_tidydb, when checking for the continued existence of a message for
4094 which it has found a message-specific retry record, was not finding
4095 messages that were in split spool directories. Consequently, it was
4096 deleting retry records that should have stayed in existence.
4098 29. eximstats updated to version 1.35
4099 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4100 1.35 - bugfix such that pie charts by volume are generated correctly
4102 30. The SPA authentication driver was not abandoning authentication and moving
4103 on to the next authenticator when an expansion was forced to fail,
4104 contradicting the general specification for all authenticators. Instead it
4105 was generating a temporary error. It now behaves as specified.
4107 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4108 (the order specifies the preference for clients). The order is now AES256,
4109 AES128, 3DES, ARCFOUR128.
4111 31. Small patch to Sieve code - explicitly set From: when generating an
4114 32. Exim crashed if a remote delivery caused a very long error message to be
4115 recorded - for instance if somebody sent an entire SpamAssassin report back
4116 as a large number of 550 error lines. This bug was coincidentally fixed by
4117 increasing the size of one of Exim's internal buffers (big_buffer) that
4118 happened as part of the Exiscan merge. However, to be on the safe side, I
4119 have made the code more robust (and fixed the comments that describe what
4122 33. Some experimental protocols are using DNS PTR records for new purposes. The
4123 keys for these records are domain names, not reversed IP addresses. The
4124 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4125 leaves it alone. Component reversal etc. now happens only for IP addresses.
4128 34. The host_aton() function is supposed to be passed a string that is known
4129 to be a valid IP address. However, in the case of IPv6 addresses, it was
4130 not checking this. This is a hostage to fortune. Exim now panics and dies
4131 if the condition is not met. A case was found where this could be provoked
4132 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4133 components; fortuitously, this particular loophole had already been fixed
4134 by change 4.50/55 or 4.44/33 above.
4136 If there are any other similar loopholes, the new check in host_aton()
4137 itself should stop them being exploited. The report I received stated that
4138 data on the command line could provoke the exploit when Exim was running as
4139 exim, but did not say which command line option was involved. All I could
4140 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4141 running as the user.
4144 35. There was a buffer overflow vulnerability in the SPA authentication code
4145 (which came originally from the Samba project). I have added a test to the
4146 spa_base64_to_bits() function which I hope fixes it.
4149 36. The daemon start-up calls getloadavg() while still root for those OS that
4150 need the first call to be done as root, but it missed one case: when
4151 deliver_queue_load_max is set with deliver_drop_privilege. This is
4152 necessary for the benefit of the queue runner, because there is no re-exec
4153 when deliver_drop_privilege is set.
4155 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4156 was in use, was not putting the data itself into the right store pool;
4157 consequently, it could be overwritten for a subsequent message in the same
4158 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4161 38. Sometimes the final signoff response after QUIT could fail to get
4162 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4163 < 0 before doing a fflush(). This bug looks as though it goes back to the
4164 introduction of TLS in release 3.20, but "sometimes" must have been rare
4165 because the tests only now provoked it.
4171 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4172 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4173 Now it returns an integer. A number of calls were still expecting a T/F
4174 return. Fortuitously, in all cases, the tests worked in OK situations,
4175 which is the norm. However, things would have gone wrong on any write
4176 failures on the smtp file descriptor. This function is used when sending
4177 messages over SMTP and also when doing verify callouts.
4179 2. When Exim is called to do synchronous delivery of a locally submitted
4180 message (the -odf or -odi options), it no longer closes stderr before doing
4183 3. Implemented the mua_wrapper option.
4185 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4187 5. Implemented the functions header_remove(), header_testname(),
4188 header_add_at_position(), and receive_remove_recipient(), and exported them
4191 6. If an ACL "warn" statement specified the addition of headers, Exim already
4192 inserted X-ACL-Warn: at the start if there was no header name. However, it
4193 was not making this test for the second and subsequent header lines if
4194 there were newlines in the string. This meant that an invalid header could
4195 be inserted if Exim was badly configured.
4197 7. Allow an ACL "warn" statement to add header lines at the start or after all
4198 the Received: headers, as well as at the end.
4200 8. Added the rcpt_4xx retry error code.
4202 9. Added postmaster_mailfrom=xxx to callout verification option.
4204 10. Added mailfrom=xxxx to the callout verification option, for verify=
4207 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4208 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4210 12. Inserted some casts to stop certain compilers warning when using pointer
4211 differences as field lengths or precisions in printf-type calls (mostly
4212 affecting debugging statements).
4214 13. Added optional readline() support for -be (dynamically loaded).
4216 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4217 same clock tick as a message's arrival, so that its received time was the
4218 same as the "first fail" time on the retry record, and that message
4219 remained on the queue past the ultimate address timeout, every queue runner
4220 would try a delivery (because it was past the ultimate address timeout) but
4221 after another failure, the ultimate address timeout, which should have then
4222 bounced the address, did not kick in. This was a "< instead of <=" error;
4223 in most cases the first failure would have been in the next clock tick
4224 after the received time, and all would be well.
4226 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4227 being recognized when the domain list was tested by the match_domain
4228 condition in an expansion string.
4230 16. Added the ${str2b64: operator.
4232 17. Exim was always calling setrlimit() to set a large limit for the number of
4233 processes, without checking whether the existing limit was already
4234 adequate. (It did check for the limit on file descriptors.) Furthermore,
4235 errors from getrlimit() and setrlimit() were being ignored. Now they are
4236 logged to the main and panic logs, but Exim does carry on, to try to do its
4237 job under whatever limits there are.
4239 18. Imported PCRE 5.0.
4241 19. Trivial typo in log message " temporarily refused connection" (the leading
4244 20. If the log selector return_path_on_delivery was set and an address was
4245 redirected to /dev/null, the delivery process crashed because it assumed
4246 that a return path would always be set for a "successful" delivery. In this
4247 case, the whole delivery is bypassed as an optimization, and therefore no
4250 21. Internal re-arrangement: the function for sending a challenge and reading
4251 a response while authentication was assuming a zero-terminated challenge
4252 string. It's now changed to take a pointer and a length, to allow for
4253 binary data in such strings.
4255 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4257 23. Exim was not respecting finduser_retries when seeking the login of the
4258 uid under which it was called; it was always trying 10 times. (The default
4259 setting of finduser_retries is zero.) Also, it was sleeping after the final
4260 failure, which is pointless.
4262 24. Implemented tls_on_connect_ports.
4264 25. Implemented acl_smtp_predata.
4266 26. If the domain in control=submission is set empty, Exim assumes that the
4267 authenticated id is a complete email address when it generates From: or
4268 Sender: header lines.
4270 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4271 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4272 chown and chgrp in /bin and hostname in /usr/bin.
4274 28. Exim was keeping the "process log" file open after each use, just as it
4275 does for the main log. This opens the possibility of it remaining open for
4276 long periods when the USR1 signal hits a daemon. Occasional processlog
4277 errors were reported, that could have been caused by this. Anyway, it seems
4278 much more sensible not to leave this file open at all, so that is what now
4281 29. The long-running daemon process does not normally write to the log once it
4282 has entered its main loop, and it closes the log before doing so. This is
4283 so that log files can straightforwardly be renamed and moved. However,
4284 there are a couple of unusual error situations where the daemon does write
4285 log entries, and I had neglected to close the log afterwards.
4287 30. The text of an SMTP error response that was received during a remote
4288 delivery was being truncated at 512 bytes. This is too short for some of
4289 the long messages that one sometimes sees. I've increased the limit to
4292 31. It is now possible to make retry rules that apply only when a message has a
4293 specific sender, in particular, an empty sender.
4295 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4296 it possible to be selective about when SMTP synchronization is enforced.
4298 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4300 32. Implemented hosts_connection_nolog.
4302 33. Added an ACL for QUIT.
4304 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4307 35. Added mailbox_size and mailbox_filecount to appendfile.
4309 36. Added control = no_multiline_responses to ACLs.
4311 37. There was a bug in the logic of the code that waits for the clock to tick
4312 in the case where the clock went backwards by a substantial amount such
4313 that the microsecond fraction of "now" was more than the microsecond
4314 fraction of "then" (but the whole seconds number was less).
4316 38. Added support for the libradius Radius client library this is found on
4317 FreeBSD (previously only the radiusclient library was supported).
4323 1. When certain lookups returned multiple values in the form name=value, the
4324 quoting of the values was not always being done properly. Specifically:
4325 (a) If the value started with a double quote, but contained no whitespace,
4327 (b) If the value contained whitespace other than a space character (i.e.
4328 tabs or newlines or carriage returns) it was not quoted.
4329 This fix has been applied to the mysql and pgsql lookups by writing a
4330 separate quoting function and calling it from the lookup code. The fix
4331 should probably also be applied to nisplus, ibase and oracle lookups, but
4332 since I cannot test any of those, I have not disturbed their existing code.
4334 2. A hit in the callout cache for a specific address caused a log line with no
4335 reason for rejecting RCPT. Now it says "Previous (cached) callout
4336 verification failure".
4338 3. There was an off-by-one bug in the queryprogram router. An over-long
4339 return line was truncated at 256 instead of 255 characters, thereby
4340 overflowing its buffer with the terminating zero. As well as fixing this, I
4341 have increased the buffer size to 1024 (and made a note to document this).
4343 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4344 when Exim is waiting for an SMTP response from a remote server, Exim
4345 restarts its select() call on the socket, thereby resetting its timeout.
4346 This is not a problem when such interrupts are rare. Somebody set up a cron
4347 job to run exiwhat every 2 minutes, which is less than the normal select()
4348 timeout (5 or 10 minutes). This meant that the select() timeout never
4349 kicked in because it was always reset. I have fixed this by comparing the
4350 time when an interrupt arrives with the time at the start of the first call
4351 to select(). If more time than the timeout has elapsed, the interrupt is
4352 treated as a timeout.
4354 5. Some internal re-factoring in preparation for the addition of Sieve
4355 extensions (by MH). In particular, the "personal" test is moved to a
4356 separate function, and given an option for scanning Cc: and Bcc: (which is
4357 not set for Exim filters).
4359 6. When Exim created an email address using the login of the caller as the
4360 local part (e.g. when creating a From: or Sender: header line), it was not
4361 quoting the local part when it contained special characters such as @.
4363 7. Installed new OpenBSD configuration files.
4365 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4366 try to make them clearer.
4368 9. Callout options, other than the timeout value, were being ignored when
4369 verifying sender addresses in header lines. For example, when using
4371 verify = header_sender/callout=no_cache
4373 the cache was (incorrectly) being used.
4375 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4376 only the Cygwin environment.
4378 11. When return_path_on_delivery was set as a log selector, if different remote
4379 addresses in the same message used different return paths and parallel
4380 remote delivery occurred, the wrong values would sometimes be logged.
4381 (Whenever a remote delivery process finished, the return path value from
4382 the most recently started remote delivery process was logged.)
4384 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4385 lines when AUTH and/or TLS are in use. This is the "received protocol"
4386 field. Exim used to use "asmtp" for authenticated SMTP, without any
4387 indication (in the protocol name) for TLS use. Now it follows the RFC and
4388 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4389 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4390 names appear in log lines as well as in Received: header lines.
4392 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4393 extensions, and comparison tests, and to fix some bugs.
4395 14. Changes to the "personal" filter test:
4397 (1) The test was buggy in that it was just doing the equivalent of
4398 "contains" tests on header lines. For example, if a user's address was
4399 anne@some.where, the "personal" test would incorrectly be true for
4401 To: susanne@some.where
4403 This test is now done by extracting each address from the header in turn,
4404 and checking the entire address. Other tests that are part of "personal"
4405 are now done using regular expressions (for example, to check local parts
4406 of addresses in From: header lines).
4408 (2) The list of non-personal local parts in From: addresses has been
4409 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4410 taken from the Sieve specification recommendations.
4412 (3) If the message contains any header line starting with "List-" it is
4413 treated as non-personal.
4415 (4) The test for "circular" in the Subject: header line has been removed
4416 because it now seems ill-conceived.
4418 15. Minor typos in src/EDITME comments corrected.
4420 16. Installed latest exipick from John Jetmore.
4422 17. If headers_add on a router specified a text string that was too long for
4423 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4424 of string_sprintf() is now avoided.
4426 18. $message_body_size was not set (it was always zero) when running the DATA
4427 ACL and the local_scan() function.
4429 19. For the "mail" command in an Exim filter, no default was being set for
4430 the once_repeat time, causing a random time value to be used if "once" was
4431 specified. (If the value happened to be <= 0, no repeat happened.) The
4432 default is now 0s, meaning "never repeat". The "vacation" command was OK
4433 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4434 (I found it when inspecting the code).
4436 20. There is now an overall timeout for performing a callout verification. It
4437 defaults to 4 times the callout timeout, which applies to individual SMTP
4438 commands during the callout. The overall timeout applies when there is more
4439 than one host that can be tried. The timeout is checked before trying the
4440 next host. This prevents very long delays if there are a large number of
4441 hosts and all are timing out (e.g. when the network connections are timing
4442 out). The value of the overall timeout can be changed by specifying an
4443 additional sub-option for "callout", called "maxwait". For example:
4445 verify = sender/callout=5s,maxwait=20s
4447 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4448 to the end before writing, but this should make it even safer).
4450 22. Exim was forgetting that it had advertised PIPELINING for the second and
4451 subsequent messages on an SMTP connection. It was also not resetting its
4452 memory on STARTTLS and an internal HELO.
4454 23. When Exim logs an SMTP synchronization error within a session, it now
4455 records whether PIPELINING has been advertised or not.
4457 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4458 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4459 rather than long int.
4461 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4463 26. Added the never_mail option to autoreply.
4469 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4470 crash if the getsockname() call failed; this can happen if a connection is
4471 closed very soon after it is established. The problem was simply in the
4472 order in which certain operations were done, causing Exim to try to write
4473 to the SMTP stream before it had set up the file descriptor. The bug has
4474 been fixed by making things happen in the correct order.
4480 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4481 before the connection was closed, thus losing the rejection response.
4483 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4484 some early Solaris releases, but causes trouble in current releases where
4485 socklen_t is defined.
4487 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4490 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4493 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4494 that releasing the top of it at the end releases what was used for sub-
4495 expansions (unless the block got too big). However, discard this block if
4496 the first thing is a variable or header, so that we can use its block when
4497 it is dynamic (useful for very large $message_headers, for example).
4499 6. Lookups now cache *every* query, not just the most recent. A new, separate
4500 store pool is used for this. It can be recovered when all lookup caches are
4501 flushed. Lookups now release memory at the end of their result strings.
4502 This has involved some general refactoring of the lookup sources.
4504 7. Some code has been added to the store_xxx() functions to reduce the amount
4505 of flapping under certain conditions.
4507 8. log_incoming_interface used to affect only the <= reception log lines. Now
4508 it causes the local interface and port to be added to several more SMTP log
4509 lines, for example "SMTP connection from", and rejection lines.
4511 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4513 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4515 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4516 could be overwritten at the end of the current message (or the start of a
4517 new message if it was set in a HELO ACL). The value is now preserved for
4518 the duration of the SMTP connection.
4520 12. If a transport had a headers_rewrite setting, and a matching header line
4521 contained an unqualified address, that address was qualified, even if it
4522 did not match any rewriting rules. The underlying bug was that the values
4523 of the flags that permit the existence of unqualified sender and recipient
4524 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4525 non-local messages, and by -bnq for local messages) were not being
4526 preserved with the message after it was received.
4528 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4529 "next input=" as part of the text comprising the host identity instead of
4530 the correct text. The code was using the same buffer for two different
4531 strings. However, depending on which order the printing function evaluated
4532 its arguments, the bug did not always show up. Under Linux, for example, my
4533 test suite worked just fine.
4535 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4536 doesn't work with some older versions of Perl. It has been changed to "my",
4537 which in any case is probably the better facility to use.
4539 15. A really picky compiler found some instances of statements for creating
4540 error messages that either had too many or two few arguments for the format
4543 16. The size of the buffer for calls to the DNS resolver has been increased
4544 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4545 for addresses that have a lot of PTR records. This alleviates a problem; it
4546 does not fully solve it.
4548 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4549 buffer now truncates the list and logs the incident, which is the same
4550 action as happens when Exim is looking up a host name and its aliases.
4551 Previously in this situation something unpredictable would happen;
4552 sometimes it was "internal error: store_reset failed".
4554 18. If a server dropped the connection unexpectedly when an Exim client was
4555 using GnuTLS and trying to read a response, the client delivery process
4556 crashed while trying to generate an error log message.
4558 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4559 string, for example:
4561 warn message = H1: something\nH2: something
4563 the text was added as a single header line from Exim's point of view
4564 though it ended up OK in the delivered message. However, searching for the
4565 second and subsequent header lines using $h_h2: did not work. This has been
4566 fixed. Similarly, if a system filter added multiple headers in this way,
4567 the routers could not see them.
4569 20. Expanded the error message when iplsearch is called with an invalid key to
4570 suggest using net-iplsearch in a host list.
4572 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4573 ACLs are no longer actually imposed (and a message to that effect is
4576 22. If a "gecos" field in a passwd entry contained escaped characters, in
4577 particular, if it contained a \" sequence, Exim got it wrong when building
4578 a From: or a Sender: header from that name. A second bug also caused
4579 incorrect handling when an unquoted " was present following a character
4580 that needed quoting.
4582 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4583 was not being matched caselessly.
4585 24. Arranged for all hyphens in the exim.8 source to be escaped with
4588 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4589 properly. Recipient callout cache records were still being keyed to include
4590 the sender, even when use_sender was set false. This led to far more
4591 callouts that were necessary. The sender is no longer included in the key
4592 when use_sender is false.
4594 26. Added "control = submission" modifier to ACLs.
4596 27. Added the ${base62d: operator to decode base 62 numbers.
4598 28. dnsdb lookups can now access SRV records.
4600 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4601 the configuration file.
4603 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4604 (-v) mode. This makes the output for a verbose queue run more intelligible.
4606 31. Added a use_postmaster feature to recipient callouts.
4608 32. Added the $body_zerocount variable, containing the number of binary zero
4609 bytes in the message body.
4611 33. The time of last modification of the "new" subdirectory is now used as the
4612 "mailbox time last read" when there is a quota error for a maildir
4615 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4617 35. Added +ignore_unknown as a special item in host lists.
4619 36. Code for decoding IPv6 addresses in host lists is now included, even if
4620 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4621 address was recognized as an IP address, but was then not correctly decoded
4622 into binary, causing unexpected and incorrect effects when compared with
4629 1. Very minor rewording of debugging text in manualroute to say "list of
4630 hosts" instead of "hostlist".
4632 2. If verify=header_syntax was set, and a header line with an unqualified
4633 address (no domain) and a large number of spaces between the end of the
4634 name and the colon was received, the reception process suffered a buffer
4635 overflow, and (when I tested it) crashed. This was caused by some obsolete
4636 code that should have been removed. The fix is to remove it!
4638 3. When running in the test harness, delay a bit after writing a bounce
4639 message to get a bit more predictability in the log output.
4641 4. Added a call to search_tidyup() just before forking a reception process. In
4642 theory, someone could use a lookup in the expansion of smtp_accept_max_
4643 per_host which, without the tidyup, could leave open a database connection.
4645 5. Added the variables $recipient_data and $sender_data which get set from a
4646 lookup success in an ACL "recipients" or "senders" condition, or a router
4647 "senders" option, similar to $domain_data and $local_part_data.
4649 6. Moved the writing of debug_print from before to after the "senders" test
4652 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4653 problems for message scanning, either using a data ACL, or using
4654 local_scan() because the Received: header was not generated till after they
4655 were called (in order to set the time as the time of reception completion).
4656 I have revised the way this works. The header is now generated after the
4657 body is received, but before the ACL or local_scan() are called. After they
4658 are run, the timestamp in the header is updated.
4664 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4665 before starting a queue runner without re-exec. This happened only when
4666 deliver_drop_privilege was set or when the Exim user was set to root. The
4667 effect of the bug was that timeouts during subsequent deliveries caused
4668 crashes instead of being properly handled. The handler is now left at its
4669 default (and expected) setting.
4671 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4672 message, again when deliver_drop_privilege is set or Exim is run as root.
4673 The bug described in (1) was not present in this case, but the tidying up
4674 of the other signals was missing. I have made the two cases consistent.
4676 3. The ignore_target_hosts setting on a manualroute router was being ignored
4677 for hosts that were looked up using the /MX notation.
4679 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4682 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4683 operated on the sender address. After changing the $sender_address to <>
4684 for the sender address verify, Exim was re-instated it as the original
4685 (before rewriting) address, but remembering that it had rewritten it, so it
4686 wasn't rewriting it again. This bug also had the effect of breaking the
4687 sender address verification caching when the sender address was rewritten.
4689 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4690 This has been changed so that if the ip literal address matches
4691 ignore_target_hosts, the router declines.
4693 7. Added expansion conditions match_domain, match_address, and match_local_
4694 part (NOT match_host).
4696 8. The placeholder for the Received: header didn't have a length field set.
4698 9. Added code to Exim itself and to exim_lock to test for a specific race
4699 condition that could lead to file corruption when using MBX delivery. The
4700 issue is with the lockfile that is created in /tmp. If this file is removed
4701 after a process has opened it but before that process has acquired a lock,
4702 there is the potential for a second process to recreate the file and also
4703 acquire a lock. This could lead to two Exim processes writing to the file
4704 at the same time. The added code performs the same test as UW imapd; it
4705 checks after acquiring the lock that its file descriptor still refers to
4706 the same named file.
4708 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4709 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4712 11. Added the smtp_active_hostname option. If used, this will typically be made
4713 to depend on the incoming interface address. Because $interface_address is
4714 not set up until the daemon has forked a reception process, error responses
4715 that can happen earlier (such as "too many connections") no longer contain
4718 12. If an expansion in a condition on a "warn" statement fails because a lookup
4719 defers, the "warn" statement is abandoned, and the next ACL statement is
4720 processed. Previously this caused the whole ACL to be aborted.
4722 13. Added the iplsearch lookup type.
4724 14. Added ident_timeout as a log selector.
4726 15. Added tls_certificate_verified as a log selector.
4728 16. Added a global option tls_require_ciphers (compare the smtp transport
4729 option of the same name). This controls incoming TLS connections.
4731 17. I finally figured out how to make tls_require_ciphers do a similar thing
4732 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4733 before starting the TLS session.
4735 18. Tabs are now shown as \t in -bP output.
4737 19. If the log selector return_path_on_delivery was set, Exim crashed when
4738 bouncing a message because it had too many Received: header lines.
4740 20. If two routers both had headers_remove settings, and the first one included
4741 a superfluous trailing colon, the final name in the first list and the
4742 first name in the second list were incorrectly joined into one item (with a
4743 colon in the middle).
4749 1. Added -C and -D options to the exinext utility, mainly to make it easier
4750 to include in the automated testing, but these could be helpful when
4751 multiple configurations are in use.
4753 2. The exinext utility was not formatting the output nicely when there was
4754 an alternate port involved in the retry record key, nor when there was a
4755 message id as well (for retries that were specific to a specific message
4756 and a specific host). It was also confused by IPv6 addresses, because of
4757 the additional colons they contain. I have fixed the IPv4 problem, and
4758 patched it up to do a reasonable job for IPv6.
4760 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4761 delivery, the log line now contains "pipelined" if PIPELINING was used.
4763 4. An SMTP transport process used to panic and die if the bind() call to set
4764 an explicit outgoing interface failed. This has been changed; it is now
4765 treated in the same way as a connect() failure.
4767 5. A reference to $sender_host_name in the part of a conditional expansion
4768 that was being skipped was still causing a DNS lookup. This no longer
4771 6. The def: expansion condition was not recognizing references to header lines
4772 that used bh_ and bheader_.
4774 7. Added the _cache feature to named lists.
4776 8. The code for checking quota_filecount in the appendfile transport was
4777 allowing one more file than it should have been.
4779 9. For compatibility with Sendmail, the command line option
4787 and sets the incoming protocol and host name (for trusted callers). The
4788 host name and its colon can be omitted when only the protocol is to be set.
4789 Note the Exim already has two private options, -pd and -ps, that refer to
4790 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4791 "s", but I don't think that's a major issue.
4793 10. A number of refactoring changes to the code, none of which should affect
4796 (a) The number of logging options was getting close to filling up the
4797 32-bit word that was used as a bit map. I have split them into two classes:
4798 those that are passed in the argument to log_write(), and those that are
4799 only ever tested independently outside of that function. These are now in
4800 separate 32-bit words, so there is plenty of room for expansion again.
4801 There is no change in the user interface or the logging behaviour.
4803 (b) When building, for example, log lines, the code previously used a
4804 macro that called string_cat() twice, in order to add two strings. This is
4805 not really sufficiently general. Furthermore, there was one instance where
4806 it was actually wrong because one of the argument was used twice, and in
4807 one call a function was used. (As it happened, calling the function twice
4808 did not affect the overall behaviour.) The macro has been replaced by a
4809 function that can join an arbitrary number of extra strings onto a growing
4812 (c) The code for expansion conditions now uses a table and a binary chop
4813 instead of a serial search (which was left over from when there were very
4814 few conditions). Also, it now recognizes conditions like "pam" even when
4815 the relevant support is not compiled in: a suitably worded error message is
4816 given if an attempt is made to use such a condition.
4818 11. Added ${time_interval:xxxxx}.
4820 12. A bug was causing one of the ddress fields not to be passed back correctly
4821 from remote delivery subprocesses. The field in question was not being
4822 subsequently used, so this caused to problems in practice.
4824 13. Added new log selectors queue_time and deliver_time.
4826 14. Might have fixed a bug in maildirsizefile handling that threw up
4827 "unexpected character" debug warnings, and recalculated the data
4828 unnecessarily. In any case, I expanded the warning message to give more
4831 15. Added the message "Restricted characters in address" to the statements in
4832 the default ACL that block characters like @ and % in local parts.
4834 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4835 Three changes have been made:
4837 (a) There was a serious bug; a negative response to MAIL caused the whole
4838 recipient domain to be cached as invalid, thereby blocking all messages
4839 to all local parts at the same domain, from all senders. This bug has
4840 been fixed. The domain is no longer cached after a negative response to
4841 MAIL if the sender used is not empty.
4843 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4846 (c) A new callout option, "use_sender" has been added for people who want
4847 the modified behaviour.
4853 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4856 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4857 indeed breaks things for older releases.
4859 3. Added additional logging to the case where there is a problem reading data
4860 from a filter that is running in a subprocess using a pipe, in order to
4861 try to track down a specific problem.
4863 4. Testing facility fudge: when running in the test harness and attempting
4864 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4865 getting "No route to host". Convert this to a timeout.
4867 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4870 6. Some OS don't have socklen_t but use size_t instead. This affects the
4871 fifth argument of getsockopt() amongst other things. This is now
4872 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4873 can be set for individual OS. I have set it for SunOS5, OSF1, and
4874 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4875 some earlier ones do not.
4877 7. Change 4.30/15 was not doing the test caselessly.
4879 8. The standard form for an IPv6 address literal was being rejected by address
4880 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4881 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4882 this, as well as the form without the "IPv6" on the front (but only when
4883 address literals are enabled, of course).
4885 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4887 10. Exim crashed if a message with an empty sender address specified by -f
4888 encountered a router with an errors_to setting. This could be provoked only
4889 by a command such as
4893 where an empty string was supplied; "<>" did not hit this bug.
4895 11. Installed PCRE release 4.5.
4897 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4898 remained set. It is now erased.
4900 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4901 times from message ids (which are base 36 rather than the normal 62).
4903 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4904 were being counted as actual protocol errors, and logged if the log
4905 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4906 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4907 and DATA following a set of rejected RCPTs do not count as protocol errors.
4908 In other words, Exim assumes they were pipelined, though this may not
4909 actually be the case. Of course, in all cases the client gets an
4910 appropriate error code.
4912 15. If a lookup fails in an ACL condition, a message about the failure may
4913 be available; it is used if testing the ACL cannot continue, because most
4914 such messages specify what the cause of the deferral is. However, some
4915 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4916 that caused an old message to be retained and used if a later statement
4917 caused a defer, replacing the real cause of the deferral.
4919 16. If an IP address had so many PTR records that the DNS lookup buffer
4920 was not large enough to hold them, Exim could crash while trying to process
4921 the truncated data. It now detects and logs this case.
4923 17. Further to 4.21/58, another change has been made: if (and only if) the
4924 first line of a message (the first header line) ends with CRLF, a bare LF
4925 in a subsequent header line has a space inserted after it, so as not to
4926 terminate the header.
4928 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4929 unnecessarily, used atoi() instead of strtol(), and didn't check the
4930 termination when getting file sizes from file names by regex.
4932 19. Completely re-implemented the support for maildirsize files, in the light
4933 of a number of problems with the previous contributed implementation
4934 (4.30/29). In particular:
4936 . If the quota is zero, the maildirsize file is maintained, but no quota is
4939 . If the maildir directory does not exist, it is created before any attempt
4940 to write a maildirsize file.
4942 . The quota value in the file is just a cache; if the quota is changed in
4943 the transport, the new value overrides.
4945 . A regular expression is available for excluding directories from the
4948 20. The autoreply transport checks the characters in options that define the
4949 message's headers; it allows continued headers, but it was checking with
4950 isspace() after an embedded newline instead of explicitly looking for a
4953 21. If all the "regular" hosts to which an address was routed had passed their
4954 expiry times, and had not reached their retry times, the address was
4955 bounced, even if fallback hosts were defined. Now Exim should go on to try
4958 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4959 equivalent code in the SMTP transport. Some hosts send humungous responses
4960 to HELO/EHLO, more than 1024 it seems.
4962 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4963 gives compiler warnings in some environments. I've now done it "properly",
4966 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4967 (because of problems with the built-in one) was declared to return uschar *
4968 instead of char *, causing compiler failure.
4970 25. Fixed a file descriptor leak when processing alias/forward files.
4972 26. Fixed a minor format string issue in dbfn.c.
4974 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4976 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4977 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4980 29. When a sender address is verified, it is cached, to save repeating the test
4981 when there is more than one recipient in a message. However, when the
4982 verification involves a callout, it is possible for different callout
4983 options to be set for different recipients. It is too complicated to keep
4984 track of this in the cache, so now Exim always runs a verification when a
4985 callout is required, relying on the callout cache for the optimization.
4986 The overhead is duplication of the address routing, but this should not be
4989 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4990 to be verified with callout=postmaster, and the main callout worked but the
4991 postmaster check failed, the verification correctly failed. However, if a
4992 subsequent RCPT command asked for sender verification *without* the
4993 postmaster check, incorrect caching caused this verification also to fail,
4996 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4997 it was not caching the DNS options (qualify_single, search_parents) that
4998 were used when the lookup failed. A subsequent lookup with different
4999 options therefore always gave the same answer, though there were cases
5000 where it should not have. (Example: a "domains = !$mx_any" option on a
5001 dnslookup router: the "domains" option is always processed without any
5002 widening, but the router might have qualify_single set.) Now Exim uses the
5003 cached value only when the same options are set.
5005 32. Added John Jetmore's "exipick" utility to the distribution.
5007 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5008 than a timeout (e.g. a certificate is required, and is not provided), an
5009 Exim server now closes the connection immediately. Previously it waited for
5010 the client to close - but if the client is SSL, it seems that they each
5011 wait for each other, leading to a delay before one of them times out.
5013 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5014 maintained 0.8.x compatibility because I don't think many are using it, and
5015 it is clearly obsolete.
5017 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5020 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5021 client certificate was expired. A simple patch fixes this, though I don't
5022 understand the full logic of why the verify callback is called multiple
5025 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5026 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5027 which causes problems with some clients (such as the Certicom SSL Plus
5028 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5029 disables the coutermeasure allowing Eudora to connect."
5031 38. Exim was not checking that a write() to a log file succeeded. This could
5032 lead to Bad Things if a log got too big, in particular if it hit a file
5033 size limit. Exim now panics and dies if it cannot write to a log file, just
5034 as it does if it cannot open a log file.
5036 39. Modified OS/Makefile-Linux so that it now contains
5038 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5040 The two -D definitions ensure that Exim is compiled with large file
5041 support, which makes it possible to handle log files that are bigger than
5044 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5045 instance) a domain was checked against a named list that involved a lookup,
5046 causing $domain_data to be set, then another domain was checked against the
5047 same list, then the first domain was re-checked, the value of $domain_data
5048 after the final check could be wrong. In particular, if the second check
5049 failed, it could be set empty. This bug probably also applied to
5052 41. The strip_trailing_dot option was not being applied to the address given
5053 with the -f command-line option.
5055 42. The code for reading a message's header from the spool was incrementing
5056 $received_count, but never initializing it. This meant that the value was
5057 incorrect (doubled) while delivering a message in the same process in which
5058 it was received. In the most common configuration of Exim, this never
5059 happens - a fresh exec is done - but it can happen when
5060 deliver_drop_privilege is set.
5062 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5063 it now includes up to 150 characters of the unexpected data in the log
5066 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5067 and building data strings. The size of both of these buffers was 10 000
5068 bytes - far larger than anybody would *ever* want, thought I. Needless to
5069 say, somebody hit the limit. I have increased the maximum line length to
5070 20 000 and the maximum data length of concatenated lines to 100 000. I have
5071 also fixed two bugs, because there was no checking on these buffers. Tsk,
5072 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5073 buffer is too small.
5075 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5076 lsearch lookups. Now it does.
5078 46. When parsing a route_list item in a manualroute router, a fixed-length
5079 buffer was used for the list of hosts. I made this 1024 bytes long,
5080 thinking that nobody would ever have a list of hosts that long. Wrong.
5081 Somebody had a whole pile of complicated expansion conditions, and the
5082 string was silently truncated, leading to an expansion error. It turns out
5083 that it is easier to change to an unlimited length (owing to other changes
5084 that have happened since this code was originally written) than to build
5085 structure for giving a limitation error. The length of the item that
5086 expands into the list of hosts is now unlimited.
5088 47. The lsearch lookup could not handle data where the length of text line was
5089 more than 4095 characters. Such lines were truncated, leading to shortened
5090 data being returned. It should now handle lines of any length.
5092 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5093 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5096 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5097 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5098 no longer happens in comment lines. A list of replacements is now placed
5099 at the head of all of the source files, except those whose only change is
5100 to replace PERL_COMMAND in the very first #! line.
5102 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5103 messages on the queue, with a bottom-up merge sort, using code contributed
5104 by Michael Haardt. This should make operations like -bp somewhat faster on
5105 large queues. It won't affect queue runners, except when queue_run_in_order
5108 51. Installed eximstats 1.31 in the distribution.
5110 52. Added support for SRV lookups to the dnslookup router.
5112 53. If an ACL referred to $message_body or $message_body_end, the value was not
5113 reset for any messages that followed in the same SMTP session.
5115 54. The store-handling optimization for building very long strings was not
5116 differentiating between the different store pools. I don't think this
5117 actually made any difference in practice, but I've tidied it.
5119 55. While running the routers to verify a sender address, $sender_address
5120 was still set to the sender address. This is wrong, because when routing to
5121 send a bounce to the sender, it would be empty. Therefore, I have changed
5122 it so that, while verifying a sender address, $sender_address is set to <>.
5123 (There is no change to what happens when verifying a recipient address.)
5125 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5126 target A or AAAA records (if not already returned) without resetting the
5127 qualify_single or search_parents options of the DNS resolver. These are
5128 inappropriate in this case because the targets of MX and SRV records must
5129 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5130 target that, when qualified, matched something in the local domain. These
5131 two options are now turned off when doing these lookups.
5133 57. It seems that at least some releases of Reiserfs (which does not have the
5134 concept of a fixed number of inodes) returns zero and not -1 for the
5135 number of available inodes. This interacted badly with check_spool_inodes,
5136 which assumed that -1 was the "no such thing" setting. What I have done is
5137 to check that the total number of inodes is greater than zero before doing
5138 the test of how many are available.
5140 58. When a "warn" ACL statement has a log_message modifier, the message is
5141 remembered, and not repeated. This is to avoid a lot of repetition when a
5142 message has many recipients that cause the same warning to be written.
5143 Howewer, Exim was preserving the list of already written lines for an
5144 entire SMTP session, which doesn't seem right. The memory is now reset if a
5145 new message is started.
5147 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5148 debugging output unless log_rewrite was also set.
5150 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5151 of (int)(handle) when we know that handle contains (void *)(-1).
5153 61. The Exim daemon panic-logs an error return when it closes the incoming
5154 connection. However "connection reset by peer" seems to be common, and
5155 isn't really an error worthy of noting specially, so that particular error
5158 62. When Exim is trying to find all the local interfaces, it used to panic and
5159 die if the ioctl to get the interface flags failed. However, it seems that
5160 on at least one OS (Solaris 9) it is possible to have an interface that is
5161 included in the list of interfaces, but for which you get a failure error
5162 for this call. This happens when the interface is not "plumbed" into a
5163 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5164 failure of the "get flags" call assumes that the interface is down.
5166 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5167 makes life easier for people who are doing arithmetic on fields extracted
5168 from dates, where you often get leading zeros that should not be
5169 interpreted as octal.
5171 64. Added qualify_domain to the redirect router, to override the global
5174 65. If a pathologically long header line contained very many addresses (the
5175 report of this problem mentioned 10 000) and each of them was rewritten,
5176 Exim could use up a very large amount of memory. (It kept on making new
5177 copies of the header line as it rewrote, and never released the old ones.)
5178 At the expense of a bit more processing, the header rewriting function has
5179 been changed so that it no longer eats memory in this way.
5181 66. The generation of the Received: header has been moved from the time that a
5182 message starts to be received, to the time that it finishes. The timestamp
5183 in the Received: header should now be very close to that of the <= log
5184 line. There are two side-effects of this change:
5186 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5187 logged header lines no longer include the local Received: line, because
5188 it has not yet been created. The same applies to a copy of the message
5189 that is returned to a non-SMTP sender when a message is rejected.
5191 (b) When a filter file is tested using -bf, no additional Received: header
5192 is added to the test message. After some thought, I decided that this
5195 This change does not affect the value of $received_for. It is still set
5196 after address rewriting, but before local_scan() is called.
5198 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5200 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5201 gave an unhelpful panic error message, and a defer error. I have managed to
5202 change this behaviour so that it now rejects any supplied certificate,
5203 which seems right, as the list of acceptable certificates is empty.
5205 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5206 gave an unhelpful defer error. I have not managed to make this reject any
5207 supplied certificates, but the error message it gives is "no certificate
5208 supplied", which is not helpful.
5210 70. exigrep's output now also includes lines that are not associated with any
5211 message, but which match the given pattern. Implemented by a patch from
5212 Martin Sluka, which also tidied up the Perl a bit.
5214 71. Recipient callout verification, like sender verification, was using <> in
5215 the MAIL FROM command. This isn't really the right thing, since the actual
5216 sender may affect whether the remote host accepts the recipient or not. I
5217 have changed it to use the actual sender in the callout; this means that
5218 the cache record is now keyed on a recipient/sender pair, not just the
5219 recipient address. There doesn't seem to be a real danger of callout loops,
5220 since a callout by the remote host to check the sender would use <>.
5221 [SEE ABOVE: changed after hitting problems.]
5223 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5224 temporary errors. However, in the case of such a code being given after
5225 the end of a data transmission (i.e. after ".") Exim was failing to write
5226 a retry record for the message. (Yes, there was some broken host that was
5227 actually sending 8xx at this point.)
5229 73. An unknown lookup type in a host list could cause Exim to panic-die when
5230 the list was checked. (An example that provoked this was putting <; in the
5231 middle of a list instead of at the start.) If this happened during a DATA
5232 ACL check, a -D file could be left lying around. This kind of configuration
5233 error no longer causes Exim to die; instead it causes a defer errror. The
5234 incident is still logged to the main and panic logs.
5236 74. Buglet left over from Exim 3 conversion. The message "too many messages
5237 in one connection" was written to the rejectlog but not the mainlog, except
5238 when address rewriting (yes!) was being logged.
5240 75. Added write_rejectlog option.
5242 76. When a system filter was run not as root (that is, when system_filter_user
5243 was set), the values of the $n variables were not being returned to the
5244 main process; thus, they were not subsequently available in the $sn
5247 77. Added +return_path_on_delivery log selector.
5249 78. A connection timeout was being treated differently from recipients deferred
5250 when testing hosts_max_try with a message that was older than the host's
5251 retry timeout. (The host should not be counted, thus allowing all hosts to
5252 be tried at least once before bouncing.) This may have been the cause of an
5253 occasionally reported bug whereby a message would remain on the queue
5254 longer than the retry timeout, but would be bounced if a delivery was
5255 forced. I say "may" because I never totally pinned down the problem;
5256 setting up timeout/retry tests is difficult. See also the next item.
5258 79. The ultimate address timeout was not being applied to errors that involved
5259 a combination of host plus message (for example, a timeout on a MAIL
5260 command). When an address resolved to a number of possible hosts, and they
5261 were not all tried for each delivery (e.g. because of hosts_max_try), a
5262 message could remain on the queue longer than the retry timeout.
5264 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5267 81. Fixed an obscure SMTP outgoing bug which required at least the following
5268 conditions: (a) there was another message waiting for the same server;
5269 (b) the server returned 5xx to all RCPT commands in the first message so
5270 that the message was not completed; (c) the server dropped the connection
5271 or gave a negative response to the RSET that Exim sends to abort the
5272 transaction. The observed case was a dropped connection after DATA that had
5273 been sent in pipelining mode. That is, the server had advertised PIPELINING
5274 but was not implementing it correctly. The effect of the bug was incorrect
5275 behaviour, such as trying another host, and this could lead to a crash.
5281 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5282 and daemon.c were passed as pointers to ints; they should have been
5283 pointers to socklen_t variables (which are typically unsigned ints).
5285 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5288 3. Fixed a really odd bug that affected only the testing scheme; patching a
5289 certain fixed string in the binary changed the value of another string that
5290 happened to be identical to the end of the original first string.
5292 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5293 name", it returns that address as the IP address. On some operating
5294 systems (e.g. Solaris), it also passes back the IP address string as the
5295 "host name". However, on others (e.g. Linux), it passes back an empty
5296 string. Exim wasn't checking for this, and was changing the host name to an
5297 empty string, assuming it had been canonicized.
5299 5. Although rare, it is permitted to have more than one PTR record for a given
5300 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5301 all the names associated with an address, because they do in Solaris.
5302 However, it seems that they do not in Linux for data that comes from the
5303 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5304 I found this out when I moved to a new Linux workstation and tried to run
5305 the Exim test suite.
5307 To get round this problem I have changed the code so that it now does its
5308 own call to the DNS to look up PTR records when searching for a host name.
5309 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5310 addresses that are only in /etc/hosts are still found.
5312 This behaviour is, however, controlled by an option called host_lookup_
5313 order, which defaults to "bydns:byaddr". If people want to use the other
5314 order, or indeed, just use one or the other means of lookup, they can
5315 specify it in this variable.
5317 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5318 some operating systems, this comes back from gethostbyaddr() as an empty
5319 string, and this is what Exim used to test for. However, it seems that in
5320 other systems, "." is yielded. Exim now tests for this case too.
5322 7. The values of check_spool_space and check_log_space are now held internally
5323 as a number of kilobytes instead of an absolute number of bytes. If a
5324 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5325 kilobyte. This means that much larger values can be stored.
5327 8. Exim monitor: an attempt to get the action menu when not actually pointing
5328 at a message produces an empty menu entitled "No message selected". This
5329 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5330 no entries in it ("Shell widget menu has zero width and/or height"). So I
5331 have added a single, blank menu entry in this case.
5333 9. Added ${quote_local_part.
5335 10. MIME decoding is now applied to the contents of Subject: header lines when
5338 11. Now that a reference to $sender_host_address automatically causes a reverse
5339 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5340 host lookup before query-style lookups in lists that might use this
5341 variable. This has therefore been abolished, and the "net-" prefix is no
5342 longer necessary for query-style lookups.
5344 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5345 have been a typo for LFLAGS, so it has been changed.
5347 13. The install script calls Exim with "-C /dev/null" in order to find the
5348 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5349 to be output. Howeve, since Exim outputs its version number before the
5350 error, it didn't break the script. It just looked ugly. I fixed this by
5351 always allowing "-C /dev/null" if the caller is root.
5353 14. Ignore overlarge ACL variable number when reading spool file - insurance
5354 against a later release with more variables having written the file.
5356 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5357 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5358 this, as well as the form without the "IPv6" on the front.
5360 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5361 OS/Makefile-Darwin file.
5363 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5364 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5366 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5367 variable. However, some OS use ints for inodes. I've added cast to long int
5368 to get rid of the compiler warning.
5370 19. I had forgotten to lock out "/../" in configuration file names when
5371 ALT_CONFIG_PREFIX was set.
5373 20. Routers used for verification do not need to specify transports. However,
5374 if such a router generated a host list, and callout was configured, Exim
5375 crashed, because it could not find a port number from the (non-existent)
5376 transport. It now assumes port 25 in this circumstance.
5378 21. Added the -t option to exigrep.
5380 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5381 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5382 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5383 them is set, LOOKUP_LSEARCH is forced.
5385 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5387 24. Added sender and host information to the "rejected by local_scan()" log
5388 line; previously there was no indication of these.
5390 25. Added .include_if_exists.
5392 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5393 Linux. It turns out that not all file systems support this. Apparently some
5394 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5395 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5396 not supported on the file descriptor, is now ignored when Exim is trying to
5397 sync a directory. This applies only to Linux.
5399 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5401 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5402 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5405 29. Added support for maildirsize files from supplied patch (modified a bit).
5407 30. The use of :fail: followed by an empty string could lead Exim to respond to
5408 sender verification failures with (e.g.):
5410 550 Verification failed for <xxx>
5411 550 Sender verify failed
5413 where the first response line was missing the '-' that indicates it is not
5414 the final line of the response.
5416 31. The loop for finding the name of the user that called Exim had a hardwired
5417 limit of 10; it now uses the value of finduser_retries, which is used for
5418 all other user lookups.
5420 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5423 33. Exim was neglecting to zero errno before one call of strtol() when
5424 expanding a string and expecting an integer value. On some systems this
5425 resulted in spurious "integer overflow" errors. Also, it was casting the
5426 result into an int without checking.
5428 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5429 did not work. The code looks as if it has *never* worked, though it appears
5430 to have been documented since at least releast 1.62. I have made it work.
5432 35. The "timeout_DNS" error in retry rules, also documented since at least
5433 1.62, also never worked. As it isn't clear exactly what this means, and
5434 clearly it isn't a major issue, I have abolished the feature by treating it
5435 as "timeout", and writing a warning to the main and panic logs.
5437 36. The display of retry rules for -brt wasn't always showing the error code
5440 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5441 timeout_connect_A, timeout_connect_MX.
5443 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5444 to the empty sender.
5446 39. The daemon was not analysing the content of -oX till after it had closed
5447 stderr and disconnected from the controlling terminal. This meant that any
5448 syntax errors were only noted on the panic log, and the return code from
5449 the command was 0. By re-arranging the code a little, I've made the
5450 decoding happen first, so such errors now appear on stderr, and the return
5451 code is 1. However, the actual setting up of the sockets still happens in
5452 the disconnected process, so errors there are still only recorded on the
5455 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5456 connections (as happens on some IP stacks) was logged at start up time as
5457 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5458 it from "IPv6 and IPv4", which means that two separate sockets are being
5461 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5462 says whether AF_INET or AF_INET6 was passed as an argument.
5464 42. Exiwhat output was messed up when time zones were included in log
5467 43. Exiwhat now gives more information about the daemon's listening ports,
5468 and whether -tls-on-connect was used.
5470 44. The "port" option of the smtp transport is now expanded.
5472 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5473 silently ignored. Now an error message is written to the main and panic
5476 46. There's a new ACL modifier called "logwrite" which writes to a log file
5477 as soon as it is encountered.
5479 47. Added $local_user_uid and $local_user_gid at routing time.
5481 48. Exim crashed when trying to verify a sender address that was being
5484 49. Exim was recognizing only a space character after ".include". It now also
5485 recognizes a tab character.
5487 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5488 extracting the relevant information from the specification. The man page no
5489 longer contains scrambled data for the -d option, and I've added a section
5490 at the front about calling Exim under different names.
5492 51. Added "extra_headers" argument to the "mail" command in filter files.
5494 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5497 53. Installed eximstats 1.29.
5499 54. Added transport_filter_timeout as a generic transport option.
5501 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5502 Cc: header lines. This was required by RFC 822, but it not required by RFC
5505 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5506 incoming messages that did not have them. Now it does so only if the
5507 message originates locally, that is, if there is no associated remote host
5508 address. When Resent- header lines are present, this applies to the Resent-
5509 lines rather than the non-Resent- lines.
5511 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5512 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5514 58. Messages for configuration errors now include the name of the main
5515 configuration file - useful now that there may be more than one file in a
5516 list (.included file names were always shown).
5518 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5519 for those rare installations that do not start the daemon as root or run it
5520 setuid root. I've cut out the call to initgroups() if the daemon is not
5523 60. The Exim user and group can now be bound into the binary as text strings
5524 that are looked up at the start of Exim's processing.
5526 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5528 62. Added $mailstore_basename variable.
5530 63. Installed patch to sieve.c from Michael Haardt.
5532 64. When Exim failed to open the panic log after failing to open the main log,
5533 the original message it was trying to log was written to stderr and debug
5534 output, but if they were not available (the usual case in production), it
5535 was lost. Now it is written to syslog before the two lines that record the
5536 failures to open the logs.
5538 65. Users' Exim filters run in subprocesses under the user's uid. It is
5539 possible for a "deliver" command or an alias in a "personal" command to
5540 provoke an address rewrite. If logging of address rewriting is configured,
5541 this fails because the process is not running as root or exim. There may be
5542 a better way of dealing with this, but for the moment (because 4.30 needs
5543 to be released), I have disabled address rewrite logging when running a
5544 filter in a non-root, non-exim process.
5550 1. The buildconfig auxiliary program wasn't quoting the value set for
5551 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5552 not defined. This bug was masked in 4.22 by the effect that was fixed in
5555 2. Some messages that were rejected after a message id was allocated were
5556 shown as "incomplete" by exigrep. It no longer does this for messages that
5557 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5559 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5560 have allow_domain_literals set, the ID did not get logged in the <= line.
5561 Domain literals are now always recognized in Message-ID: header lines.
5563 4. The first argument for a ${extract expansion item is the key name or field
5564 number. Leading and trailing spaces in this item were not being ignored,
5565 causing some misleading effects.
5567 5. When deliver_drop_privilege was set, single queue runner processes started
5568 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5569 same command in the process it spins off) were not dropping privilege.
5571 6. When the daemon running as "exim" started a queue runner, it always
5572 re-executed Exim in the spun-off process. This is a waste of effort when
5573 deliver_drop_privilege is set. The new process now just calls the
5574 queue-runner function directly.
5580 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5583 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5584 ignored. Though the use of -oP was forcing the writing of a pid file, it
5585 was always written to the default place.
5587 3. If the message "no IP address found for host xxxx" is generated during
5588 incoming verification, it is now followed by identification of the incoming
5589 connection (so you can more easily find what provoked it).
5591 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5593 5. Added some features to "harden" Exim a bit more against certain attacks:
5595 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5596 be put in Local/Makefile. This is like the never_users runtime option,
5597 but it cannot be overridden. The default setting is "root".
5599 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5600 prefix string with which any file named in a -C command line option
5603 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5604 is retained for -C and -D only if the caller of Exim is root. Without
5605 it, the exim user may also use -C and -D and retain privilege.
5607 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5608 command line option is disabled.
5610 6. Macro names set by the -D option must start with an upper case letter, just
5611 like macro names defined in the configuration file.
5613 7. Added "dereference=" facility to LDAP.
5615 8. Two instances of the typo "uknown" in the source files are fixed.
5617 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5618 the Configure-Makefile script screwed up while processing it.
5620 10. Incorporated PCRE 4.4.
5622 11. The SMTP synchronization check was not operating right at the start of an
5623 SMTP session. For example, it could not catch a HELO sent before the client
5624 waited for the greeting. There is now a check for outstanding input at the
5625 point when the greeting is written. Because of the duplex, asynchronous
5626 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5627 way, but not yet received, when the check is performed.
5629 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5630 on TCP/IP sockets, because this apparently causes some broken clients to
5633 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5634 unchanged) from the Cygwin maintainer.
5636 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5637 was supported instead of testing for mbx. Effectively a typo.
5639 15. The spa authenticator server code was not checking that the input it
5640 received was valid base64.
5642 16. The debug output line for the "set" modifier in ACLs was not showing the
5643 name of the variable that was being set.
5645 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5647 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5648 lookup on its own. Something else was needed to trigger the lookup. For
5649 example, a match in host_lookup or the need for a host name in a host list.
5650 Now, if $sender_host_name is referenced and the host name has not yet been
5651 looked up, a lookup is performed. If the lookup fails, the variable remains
5652 empty, and $host_lookup_failed is set to "1".
5654 19. Added "eqi" as a case-independent comparison operator.
5656 20. The saslauthd authentication condition could segfault if neither service
5657 nor realm was specified.
5659 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5660 error message that was logged was misleading, and incoming SMTP
5661 connections were dropped. The message is now more accurate, and temporary
5662 errors are given to SMTP connections.
5664 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5665 (except RSET) with a 421 error, until QUIT is received. However, it was
5666 failing to send a response to QUIT.
5668 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5669 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5670 ACL is now run after the helo_try_verify_hosts code.
5672 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5673 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5674 case-independent, so other case variants are also recognized). Apparently
5675 some systems use these upper case variants.
5677 25. If more than two messages were waiting for the same host, and a transport
5678 filter was specified for the transport, Exim sent two messages over the
5679 same TCP/IP connection, and then failed with "socket operation on non-
5680 socket" when it tried to send the third.
5682 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5684 27. The extern definition of crypt16() in expand.c was not being excluded when
5685 the OS had its own crypt16() function.
5687 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5688 as a preferred synonym for return_size_limit, both as an option and as an
5691 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5693 30. Changed the default configuration ACL to relax the local part checking rule
5694 for addresses that are not in any local domains. For these addresses,
5695 slashes and pipe symbols are allowed within local parts, but the sequence
5696 /../ is explicitly forbidden.
5698 31. SPA server authentication was not clearing the challenge buffer before
5701 32. log_message in a "warn" ACL statement was writing to the reject log as
5702 well as to the main log, which contradicts the documentation and doesn't
5703 seem right (because no rejection is happening). So I have stopped it.
5705 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5706 However, I am unable to do any testing of this.
5708 34. Fixed an infelicity in the appendfile transport. When checking directories
5709 for a mailbox, to see if any needed to be created, it was accidentally
5710 using path names with one or more superfluous leading slashes; tracing
5711 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5713 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5714 contents are added to the log line that is written for every discarded
5715 recipient. (Previously a log_message setting was ignored.)
5717 36. The ${quote: operator now quotes the string if it is empty.
5719 37. The install script runs exim in order to find its version number. If for
5720 some reason other than non-existence or emptiness, which it checks, it
5721 could not run './exim', it was installing it with an empty version number,
5722 i.e. as "exim-". This error state is now caught, and the installation is
5725 38. An argument was missing from the function that creates an error message
5726 when Exim fails to connect to the socket for saslauthd authentication.
5727 This could cause Exim to crash, or give a corrupted message.
5729 39. Added isip, isip4, and isip6 to ${if conditions.
5731 40. The ACL variables $acl_xx are now saved with the message, and can be
5732 accessed later in routers, transports, and filters.
5734 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5735 strings in the file are not string-expanded.
5737 42. If a MAIL command specified a SIZE value that was too large to fit into an
5738 int variable, the check against message_size_limit failed. Such values are
5739 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5740 day this will have to be increased, but I don't think I want to be around
5741 when emails are that large.
5748 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5749 iconv() is not standard in FreeBSD.
5751 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5752 IPv6 enabled. The observed symptom was a segmentation fault on return from
5753 the function os_common_find_running_interfaces() in src/os.c.
5755 3. In the check_special_case() function in daemon.c I had used "errno" as an
5756 argument name, which causes warnings on some systems. This was basically a
5757 typo, since it was named "eno" in the comments!
5759 4. The code that waits for the clock to tick (at a resolution of some fraction
5760 of a second) so as to ensure message-id uniqueness was always waiting for
5761 at least one whole tick, when it could have waited for less. [This is
5762 almost certainly not relevant at current processor speeds, where it is
5763 unlikely to ever wait at all. But we try to future-proof.]
5765 5. The function that sleeps for a time interval that includes fractions of a
5766 second contained a race. It did not block SIGALRM between setting the
5767 timer, and suspending (a couple of lines later). If the interval was short
5768 and the sigsuspend() was delayed until after it had expired, the suspension
5769 never ended. On busy systems this could lead to processes getting stuck for
5772 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5773 process, before it forks any delivery processes. The open lookup caching
5774 mechanism meant that the open file or database connection was passed into
5775 the delivery process. The problem was that delivery processes always tidy
5776 up cached lookup data. This could cause a problem for the next delivery
5777 process started by the queue runner, because the external queue runner
5778 process does not know about the closure. So the next delivery process
5779 still has data in the lookup cache. In the case of a file lookup, there was
5780 no problem because closing a file descriptor in a subprocess doesn't affect
5781 the parent. However, if the lookup was caching a connection to a database,
5782 the connection was closed, and the second delivery process was likely to
5783 see errors such as "PGSQL: query failed: server closed the connection
5784 unexpectedly". The problem has been fixed by closing all cached lookups
5785 in a queue runner before running a delivery process.
5787 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5788 seem to have the "const" qualifier which it has on other OS. I've
5791 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5792 *of the same type* (client or server) with the same public name that an
5793 error should be diagnosed.
5795 9. When Exim looked up a host name for an IP address, but failed to find the
5796 original IP address when looking up the host name (a safety check), it
5797 output the message "<ip address> does not match any IP for NULL", which was
5798 confusing, to say the least. The bug was that the host name should have
5799 appeared instead of "NULL".
5801 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5802 user that is built into the binary, and the -C or -D options is used, root
5803 privilege is dropped before the configuration file is read. In addition,
5804 logging is switched to stderr instead of the normal log files. If the
5805 configuration then re-defines the Exim user, the unprivileged environment
5806 is probably not what is expected, so Exim logs a panic warning message (but
5809 However, if deliver_drop_privilege is set, the unprivileged state may well
5810 be exactly what is intended, so the warning has been cut out in that case,
5811 and Exim is allowed to try to write to its normal log files.
5817 1. smtp_return_error_details was not giving details for temporary sender
5818 or receiver verification errors.
5820 2. Diagnose a configuration error if two authenticators have the same public
5823 3. Exim used not to create the message log file for a message until the first
5824 delivery attempt. This could be confusing when incoming messages were held
5825 for policy or load reasons. The message log file is now created at the time
5826 the message is received, and an initial "Received" line is written to it.
5828 4. The automatically generated man page for command line options had a minor
5829 bug that caused no ill effects; however, a more serious problem was that
5830 the procedure for building the man page automatically didn't always
5831 operate. Consequently, release 4.20 contains an out-of-date version. This
5832 shouldn't happen again.
5834 5. When building Exim with embedded Perl support, the script that builds the
5835 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5836 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5838 6. The freeze_tell option was not being used for messages that were frozen on
5839 arrival, either by an ACL or by local_scan().
5841 7. Added the smtp_incomplete_transaction log selector.
5843 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5844 was accepting AUTH without a new EHLO.
5846 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5847 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5850 10. Logging of TCP/IP connections (when configured) now happens in the main
5851 daemon process instead of the child process, so that the TCP/IP connection
5852 count is more accurate (but it can never be perfect).
5854 11. The use of "drop" in a nested ACL was not being handled correctly in the
5855 outer ACL. Now, if condition failure induced by the nested "drop" causes
5856 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5857 or "require"), the connection is dropped.
5859 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5860 that yield "discard" can now be used with an "accept" or a "discard" verb,
5861 but an error is generated for any others (because I can't see a useful way
5862 to define what should happen).
5864 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5865 are now processed in the same way as lines in the Exim configuration file.
5866 In particular, continuation lines are supported.
5868 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5870 15. Added -ti meaning -t -i.
5872 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5873 domains, and warn by logging if others are found.
5875 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5876 returned from GIFCONF when Exim is trying to find the list of interfaces on
5877 a host. The code in os.c has been modified to copy each ifreq to an aligned
5878 structure in all cases.
5880 Also, in some cases, the returned ifreq's were being copied to a 'struct
5881 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5882 means the last couple of bytes of an IPv6 address could be chopped if the
5883 ifreq contained only a normal sockaddr (14 bytes storage).
5885 18. Named domain lists were not supported in the hosts_treat_as_local option.
5886 An entry such as +xxxx was not recognized, and was treated as a literal
5889 19. Ensure that header lines added by a DATA ACL are included in the reject log
5890 if the ACL subsequently rejects the message.
5892 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5893 MD5 (which is deprecated).
5895 21. When testing a filter file using -bf, Exim was writing a message when it
5896 took the sender from a "From " line in the message, but it was not doing so
5897 when it took $return_path from a Return-Path: header line. It now does.
5899 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5900 with a valid header line field name (a series of printing characters
5901 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5903 23. Changed "disc" in the source to "disk" to conform to the documentation and
5904 the book and for uniformity.
5906 24. Ignore Sendmail's -Ooption=value command line item.
5908 25. When execve() failed while trying to run a command in a pipe transport,
5909 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5910 could be confused with a return value of 69 from the command itself. This
5911 has been changed to 127, the value the shell returns if it is asked to run
5912 a non-existent command. The wording for the related log line suggests a
5913 non-existent command as the problem.
5915 26. If received_header_text expands to an empty string, do not add a Received:
5916 header line to the message. (Well, it adds a token one on the spool, but
5917 marks it "old" so that it doesn't get used or transmitted.)
5919 27. Installed eximstats 1.28 (addition of -nt option).
5921 28. There was no check for failure on the call to getsockname() in the daemon
5922 code. This can fail if there is a shortage of resources on the system, with
5923 ENOMEM, for example. A temporary error is now given on failure.
5925 29. Contrary to the C standard, it seems that in some environments, the
5926 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5927 program. Exim now does this explicitly; it affects the formatting of
5928 timestamps using strftime().
5930 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5931 complaints. I've now initialized all the variables, to avoid this.
5933 32. Header lines added by a system filter were not being "seen" during
5934 transport-time rewrites.
5936 33. The info_callback() function passed to OpenSSL is set up with type void
5937 (*)(SSL *, int, int), as described somewhere. However, when calling the
5938 function (actually a macro) that sets it up, the type void(*)() is
5939 expected. I've put in a cast to prevent warnings from picky compilers.
5941 34. If a DNS black list lookup found a CNAME record, but there were no A
5942 records associated with the domain it pointed at, Exim crashed.
5944 35. If a DNS black list lookup returned more than one A record, Exim ignored
5945 all but the first. It now scans all returned addresses if a particular IP
5946 value is being sought. In this situation, the contents of the
5947 $dnslist_value variable are a list of all the addresses, separated by a
5950 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5951 to accept a host name and all its aliases if the forward lookup for any of
5952 them yielded the IP address of the incoming connection. Now it accepts only
5953 those names whose forward lookup yields the correct IP address. Any other
5954 names are discarded. This closes a loophole whereby a rogue DNS
5955 administrator could create reverse DNS records to break through a
5956 wildcarded host restriction in an ACL.
5958 37. If a user filter or a system filter that ran in a subprocess used any of
5959 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5960 the wrong values were passed to the pipe command ($thisaddress had the
5961 value of $0, $0 had the value of $1, etc). This bug was introduced by
5962 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5964 38. Improved the line breaking for long SMTP error messages from ACLs.
5965 Previously, if there was no break point between 40 and 75 characters, Exim
5966 left the rest of the message alone. Two changes have been made: (a) I've
5967 reduced the minimum length to 35 characters; (b) if it can't find a break
5968 point between 35 and 75 characters, it looks ahead and uses the first one
5969 that it finds. This may give the occasional overlong line, but at least the
5970 remaining text gets split now.
5972 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5973 file descriptors might be low, and that setting 1000 would always raise it.
5974 It turns out that in some environments, the limit is already over 1000 and
5975 that lowering it causes trouble. So now Exim takes care not to decrease it.
5977 40. When delivering a message, the value of $return_path is set to $sender_
5978 address at the start of routing (routers may change the value). By an
5979 oversight, this default was not being set up when an address was tested by
5980 -bt or -bv, which affected the outcome if any router or filter referred to
5983 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5984 log only when "message" or "log_message" are set. However, if one of the
5985 conditions was an address verification, or a call to a nested ACL, the
5986 messages generated by the underlying test were being passed through. This
5987 no longer happens. The underlying message is available in $acl_verify_
5988 message for both "message" and "log_message" expansions, so it can be
5989 passed through if needed.
5991 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5992 new expansion $bh_ to give the encoded byte string without charset
5993 translation. Translation happens only if iconv() is available; HAVE_ICONV
5994 indicates this at build time. HEADERS_CHARSET gives the charset to
5995 translate to; headers_charset can change it in the configuration, and
5996 "headers charset" can change it in an individual filter file.
5998 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5999 that creates RFC 2047 encoded "words" labels them as that charset instead
6000 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6001 expansion operator; (ii) when Exim creates a From: line for a local
6002 message; (iii) when a header line is rewritten to include a "phrase" part.
6004 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6005 buggy, causing it to skip the first lines of messages whose message ID
6006 ended in 'D'. This would not have bitten before Exim release 4.14, because
6007 message IDs were unlikely to end in 'D' before then. The effect was to have
6008 incorrect size information for certain domains.
6010 45. #include "config.h" was missing at the start of the crypt16.c module. This
6011 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6014 46. If there was a timeout during a "random" callout check, Exim treated it as
6015 a failure of the random address, and carried on sending RSET and the real
6016 address. If the delay was just some slowness somewhere, the response to the
6017 original RCPT would be taken as a response to RSET and so on, causing
6018 mayhem of various kinds.
6020 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6021 when I implemented it. It didn't allow for the fact that some option values
6022 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6023 the right test for positive values.
6025 48. Domain names in DNS records are case-independent. Exim always looks them up
6026 in lower case. Some resolvers return domain names in exactly the case they
6027 appear in the zone file, that is, they may contain uppercase letters. Not
6028 all resolvers do this - some return always lower case. Exim was treating a
6029 change of case by a resolver as a change of domain, similar to a widening
6030 of a domain abbreviation. This triggered its re-routing code and so it was
6031 trying to route what was effectively the same domain again. This normally
6032 caused routing to fail (because the router wouldn't handle the domain
6033 twice). Now Exim checks for this case specially, and just changes the
6034 casing of the domain that it ultimately uses when it transmits the message
6037 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6040 50. If a filter generated a file delivery with a non-absolute name (possible if
6041 no home directory exists for the router), the forbid_file option was not
6044 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6045 the existing equality matching.
6047 52. Exim was using ints instead of ino_t variables in some places where it was
6048 dealing with inode numbers.
6050 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6051 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6052 called TMPDIR, and if it finds it is different, it changes its value.
6054 54. The smtp_printf() function is now made available to local_scan() so
6055 additional output lines can be written before returning. There is also an
6056 smtp_fflush() function to enable the detection of a dropped connection.
6057 The variables smtp_input and smtp_batched_input are exported to
6060 55. Changed the default runtime configuration: the message "Unknown user"
6061 has been removed from the ACL, and instead placed on the localuser router,
6062 using the cannot_route_message feature. This means that any verification
6063 failures that generate their own messages won't get overridden. Similarly,
6064 the "Unrouteable address" message that was in the ACL for unverifiable
6065 relay addresses has also been removed.
6067 56. Added hosts_avoid_esmtp to the smtp transport.
6069 57. The exicyclog script was not checking for the esoteric option
6070 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6071 will work only if exicyclog is run under the appropriate euid.
6073 58. Following a discussion on the list, the rules by which Exim recognises line
6074 endings on incoming messages have been changed. The -dropcr and drop_cr
6075 options are now no-ops, retained only for backwards compatibility. The
6076 following line terminators are recognized: LF CRLF CR. However, special
6077 processing applies to CR:
6079 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6080 nor a local message in the state where . is a terminator.
6082 (ii) If a bare CR is encountered in a header line, an extra space is added
6083 after the line terminator so as not to end the header. The reasoning
6084 behind this is that bare CRs in header lines are most likely either
6085 to be mistakes, or people trying to play silly games.
6087 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6088 was being incorrectly given as 18 bytes larger than it should have been.
6089 This is a VOB (very old bug).
6091 60. This may never have affected anything current, but just in case it has:
6092 When the local host is found other than at the start of a list of hosts,
6093 the local host, those with the same MX, and any that follow, are discarded.
6094 When the list in question was part of a longer list of hosts, the following
6095 hosts (not currently being processed) were also being discarded. This no
6096 longer happens. I'm not sure if this situation could ever has previously
6099 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6102 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6103 header. This is something that is recommended in a new Internet Draft, and
6104 is something that is documented as being done by Sendmail. There are two
6105 possible values. For messages generated by the autoreply transport, Exim
6108 Auto-Submitted: auto-replied
6110 whereas for all other generated messages (e.g. bounces) it adds
6112 Auto-Submitted: auto-generated
6114 63. The "personal" condition in filters now includes a test for the
6115 Auto-Submitted: header. If it contains the string "auto-" the message it
6116 not considered personal.
6118 64. Added rcpt_include_affixes as a generic transport option.
6120 65. Added queue_only_override (default true).
6122 66. Added the syslog_duplication option.
6124 67. If what should have been the first header line of a message consisted of
6125 a space followed by a colon, Exim was mis-interpreting it as a header line.
6126 It isn't of course - it is syntactically invalid and should therefore be
6127 treated as the start of the message body. The misbehaviour could have
6128 caused a number of strange effects, including loss of data in subsequent
6129 header lines, and spool format errors.
6131 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6132 client host had authenticated. This control can now be exercised by an ACL
6133 for more flexibility.
6135 69. By default, callouts do not happen when testing with -bh. There is now a
6136 variant, -bhc, which does actually run the callout code, including
6137 consulting and updating the callout cache.
6139 70. Added support for saslauthd authentication, courtesy of Alexander
6142 71. If statvfs() failed on the spool or log directories while checking their
6143 size for availability, Exim confusingly gave the error "space shortage".
6144 Furthermore, in debugging mode it crashed with a floating point exception.
6145 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6146 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6147 serious problem, Exim now writes to the main and panic logs when this
6148 happens, with details of the failure. It then refuses to accept the
6149 incoming message, giving the message "spool directory problem" or "log
6150 directory problem" with a 421 code for SMTP messages.
6152 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6153 0, 1, and 2 exist, because some OS complain for execs without them (see
6154 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6155 descriptors. However, the code omitted to check that the open succeeded,
6156 causing mysterious errors if for some reason the permissions on /dev/null
6157 got screwed. Now Exim writes a message to the main and panic logs, and
6158 bombs out if it can't open /dev/null.
6160 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6161 interact so that it is all more flexible. It is supposed to remain
6162 backwards compatible. Also added extra_local_interfaces.
6164 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6165 to bomb out with an assertion failure - to the client this appears as a
6166 connection drop. This problem occurs in the part of the code that was taken
6167 from the Samba project. Fortunately, the assertion is in a very simple
6168 function, so I have fixed this by reproducing the function inline in the
6169 one place where it is called, and arranging for authentication to fail
6170 instead of killing the process with assert().
6172 75. The SPA client code was not working when the server requested OEM rather
6173 than Unicode encoding.
6175 76. Added code to make require_files with a specific uid setting more usable in
6176 the case where statting the file as root fails - usually a non-root-mounted
6177 NFS file system. When this happens and the failure is EACCES, Exim now
6178 forks a subprocess and does the per-uid checking as the relevant uid.
6180 77. Added process_log_path.
6182 78. If log_file_path was not explicitly set, a setting of check_log_space or
6183 check_log_inodes was ignored.
6185 79. If a space check for the spool or log partitions fails, the incident is now
6186 logged. Of course, in the latter case the data may get lost...
6188 80. Added the %p formatting code to string_format() so that it can be used to
6189 print addresses in debug_print(). Adjusted all the address printing in the
6190 debugging in store.c to use %p rather than %d.
6192 81. There was a concern that a line of code in smtp_in.c could overflow a
6193 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6194 initially expressed, the concern was not well-founded, because trailing
6195 spaces are removed early. However, if the trailing spaces were followed by
6196 a NULL, they did not get removed, so the overflow was possible. Two fixes
6199 (a) I re-wrote the offending code in a cleaner fashion.
6200 (b) If an incoming SMTP command contains a NULL character, it is rejected
6203 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6204 runs initgroups(), so that if the Exim user is in any additional groups,
6205 they will be used during message reception.
6211 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob