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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
60 TL/01 When running the test suite, the README says that variables such as
61 no_msglog_check are global and can be placed anywhere in a specific
62 test's script, however it was observed that placement needed to be near
63 the beginning for it to behave that way. Changed the runtest perl
64 script to read through the entire script once to detect and set these
65 variables, reset to the beginning of the script, and then run through
66 the script parsing/test process like normal.
68 TL/02 The BSD's have an arc4random API. One of the functions to induce
69 adding randomness was arc4random_stir(), but it has been removed in
70 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
71 function when detected.
73 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
74 cause callback expansion.
76 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
77 syntax errors in an expansion can be treated as a string instead of
78 logging or causing an error, due to the internal use of bool_lax
79 instead of bool when processing it.
81 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
82 server certificates when making smtp deliveries.
84 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
86 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
88 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
90 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
91 Merged patch from Sebastian Wiedenroth.
93 JH/05 Fix results-pipe from transport process. Several recipients, combined
94 with certificate use, exposed issues where response data items split
95 over buffer boundaries were not parsed properly. This eventually
96 resulted in duplicates being sent. This issue only became common enough
97 to notice due to the introduction of conection certificate information,
98 the item size being so much larger. Found and fixed by Wolfgang Breyha.
100 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
101 size buffer was used, resulting in syntax errors when an expansion
104 JH/07 Add support for directories of certificates when compiled with a GnuTLS
105 version 3.3.6 or later.
107 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
108 is EXPERIMENTAL_EVENT, the main-configuration and transport options
109 both become "event_action", the variables become $event_name, $event_data
110 and $event_defer_errno. There is a new variable $verify_mode, usable in
111 routers, transports and related events. The tls:cert event is now also
112 raised for inbound connections, if the main configuration event_action
115 TL/06 In test suite, disable OCSP for old versions of openssl which contained
116 early OCSP support, but no stapling (appears to be less than 1.0.0).
118 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
119 server certificate names available under the smtp transport option
120 "tls_verify_cert_hostname" now do not permit multi-component wildcard
123 JH/10 Time-related extraction expansions from certificates now use the main
124 option "timezone" setting for output formatting, and are consistent
125 between OpenSSL and GnuTLS compilations. Bug 1541.
127 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
128 encoded parameter in the incoming message. Bug 1558.
130 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
131 include certificate info, eximon was claiming there were spoolfile
134 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
136 JH/14 Log delivery-related information more consistently, using the sequence
137 "H=<name> [<ip>]" wherever possible.
139 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
140 are problematic for Debian distribution, omit them from the release
143 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
145 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
148 JH/17 Fix a null-indirection in certextract expansions when a nondefault
149 output list separator was used.
154 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
155 checkers that were complaining about end of non-void function with no
158 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
159 This was a regression intruduced in 4.83 by another bugfix.
161 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
163 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
164 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
170 TF/01 Correctly close the server side of TLS when forking for delivery.
172 When a message was received over SMTP with TLS, Exim failed to clear up
173 the incoming connection properly after forking off the child process to
174 deliver the message. In some situations the subsequent outgoing
175 delivery connection happened to have the same fd number as the incoming
176 connection previously had. Exim would try to use TLS and fail, logging
177 a "Bad file descriptor" error.
179 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
180 utilities have not been installed.
182 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
183 temporary space as the ACL may create new global variables.
185 TL/01 LDAP support uses per connection or global context settings, depending
186 upon the detected version of the libraries at build time.
188 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
189 to extract and use the src ip:port in logging and expansions as if it
190 were a direct connection from the outside internet. PPv2 support was
191 updated based on HAProxy spec change in May 2014.
193 JH/02 Add ${listextract {number}{list}{success}{fail}}.
195 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
196 Properly escape header and check for NULL return.
198 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
201 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
203 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
204 characters in header names, implemented as a verify condition.
205 Contributed by Michael Fischer v. Mollard.
207 TL/05 Rename SPF condition results err_perm and err_temp to standardized
208 results permerror and temperror. Previous values are deprecated but
209 still accepted. In a future release, err_perm and err_temp will be
210 completely removed, which will be a backward incompatibility if the
211 ACL tests for either of these two old results. Patch contributed by
212 user bes-internal on the mailing list.
214 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
216 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
217 selectors, in both main and reject logs.
219 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
222 JH/07 Add malware type "sock" for talking to simple daemon.
224 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
226 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
227 routers/transports under cutthrough routing.
229 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
230 numbers. Touch up "bool" conditional to keep the same definition.
232 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
234 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
236 JH/12 Expand items in router/transport headers_add or headers_remove lists
237 individually rather than the list as a whole. Bug 1452.
239 Required for reasonable handling of multiple headers_ options when
240 they may be empty; requires that headers_remove items with embedded
241 colons must have them doubled (or the list-separator changed).
243 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
244 view the policy declared in the DMARC record. Currently, $dmarc_status
245 is a combined value of both the record presence and the result of the
248 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
250 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
251 dnslookup router and the smtp transport (applying to the forward
254 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
255 of ldap servers used for a specific lookup. Patch provided by Heiko
258 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
259 New variable $lookup_dnssec_authenticated for observability.
261 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
262 Patch submitted by Lars Timman.
264 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
266 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
267 Requires trusted mode and valid format message id, aborts otherwise.
268 Patch contributed by Heiko Schlichting.
270 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
271 certextract with support for various fields. Bug 1358.
273 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
274 is requested by default, modifiable by smtp transport option
277 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
278 operate on certificate variables to give certificate fingerprints
279 Also new ${sha256:cert_variable}.
281 JH/23 The PRDR feature is moved from being Experimental into the mainline.
283 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
284 Christian Aistleitner.
286 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
288 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
289 file. Patch from Wolfgang Breyha.
291 JH/25 Expand the coverage of the delivery $host and $host_address to
292 client authenticators run in verify callout. Bug 1476.
294 JH/26 Port service names are now accepted for tls_on_connect_ports, to
295 align with daemon_smtp_ports. Bug 72.
297 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
298 support and error reporting did not work properly.
300 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
301 and is readable. Patch from Andrew Colin Kissa.
303 TL/14 Enhance documentation of ${run expansion and how it parses the
304 commandline after expansion, particularly in the case when an
305 unquoted variable expansion results in an empty value.
307 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
309 PP/02 Fix internal collision of T_APL on systems which support RFC3123
310 by renaming away from it. Addresses GH issue 15, reported by
313 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
315 TL/15 SECURITY: prevent double expansion in math comparison functions
316 (can expand unsanitized data). Not remotely exploitable.
323 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
325 PP/02 Make -n do something, by making it not do something.
326 When combined with -bP, the name of an option is not output.
328 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
331 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
332 $sender_host_name and config options to manage this, and basic check
335 PP/05 DSCP support for outbound connections and control modifier for inbound.
337 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
338 (Only plugin which currently uses this is kerberos4, which nobody should
339 be using, but we should make it available and other future plugins might
340 conceivably use it, even though it would break NAT; stuff *should* be
341 using channel bindings instead).
343 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
344 name; added for Sendmail compatibility; requires admin caller.
345 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
346 just ignore it); requires trusted caller.
347 Also parse but ignore: -Ac -Am -X<logfile>
350 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
352 TL/02 Add +smtp_confirmation as a default logging option.
354 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
355 Patch by Magnus Holmgren from 2007-02-20.
357 TL/04 Bugzilla 1281 - Spec typo.
358 Bugzilla 1283 - Spec typo.
359 Bugzilla 1290 - Spec grammar fixes.
361 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
363 TL/06 Add Experimental DMARC support using libopendmarc libraries.
365 TL/07 Fix an out of order global option causing a segfault. Reported to dev
366 mailing list by by Dmitry Isaikin.
368 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
370 JH/02 Support "G" suffix to numbers in ${if comparisons.
372 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
374 NM/01 Bugzilla 1197 - Spec typo
375 Bugzilla 1196 - Spec examples corrections
377 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
379 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
380 gnutls_enable_pkcs11, but renamed to more accurately indicate its
383 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
384 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
386 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
387 "acl {{name}{arg}...}", and optional args on acl condition
390 JH/05 Permit multiple router/transport headers_add/remove lines.
392 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
394 JH/07 Avoid using a waiting database for a single-message-only transport.
395 Performance patch from Paul Fisher. Bugzilla 1262.
397 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
400 JH/09 Add $headers_added variable, with content from use of ACL modifier
401 add_header (but not yet added to the message). Bugzilla 199.
403 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
404 Pulled from Bugzilla 817 by Wolfgang Breyha.
406 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
408 (nb: this is the same fix as in Exim 4.80.1)
410 JH/11 Add A= logging on delivery lines, and a client_set_id option on
413 JH/12 Add optional authenticated_sender logging to A= and a log_selector
416 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
418 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
419 advertise SMTP AUTH mechanism to us, instead of a generic
420 protocol violation error. Also, make Exim more robust to bad
421 data from the Dovecot auth socket.
423 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
425 When a queue runner is handling a message, Exim first routes the
426 recipient addresses, during which it prunes them based on the retry
427 hints database. After that it attempts to deliver the message to
428 any remaining recipients. It then updates the hints database using
431 So if a recipient address works intermittently, it can get repeatedly
432 deferred at routing time. The retry hints record remains fresh so the
433 address never reaches the final cutoff time.
435 This is a fairly common occurrence when a user is bumping up against
436 their storage quota. Exim had some logic in its local delivery code
437 to deal with this. However it did not apply to per-recipient defers
438 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
440 This change adds a proper retry rule check during routing so that the
441 final cutoff time is checked against the message's age. We only do
442 this check if there is an address retry record and there is not a
443 domain retry record; this implies that previous attempts to handle
444 the address had the retry_use_local_parts option turned on. We use
445 this as an approximation for the destination being like a local
446 delivery, as in LMTP.
448 I suspect this new check makes the old local delivery cutoff check
449 redundant, but I have not verified this so I left the code in place.
451 TF/02 Correct gecos expansion when From: is a prefix of the username.
453 Test 0254 submits a message to Exim with the header
457 When I ran the test suite under the user fanf2, Exim expanded
458 the header to contain my full name, whereas it should have added
459 a Resent-Sender: header. It erroneously treats any prefix of the
460 username as equal to the username.
462 This change corrects that bug.
464 GF/01 DCC debug and logging tidyup
465 Error conditions log to paniclog rather than rejectlog.
466 Debug lines prefixed by "DCC: " to remove any ambiguity.
468 TF/03 Avoid unnecessary rebuilds of lookup-related code.
470 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
471 Bug spotted by Jeremy Harris; was flawed since initial commit.
472 Would have resulted in OCSP responses post-SNI triggering an Exim
473 NULL dereference and crash.
475 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
477 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
478 Bug detection, analysis and fix by Samuel Thibault.
479 Bugzilla 1331, Debian bug #698092.
481 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
483 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
484 Server implementation by Todd Lyons, client by JH.
485 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
486 config variable "prdr_enable" controls whether the server
487 advertises the facility. If the client requests PRDR a new
488 acl_data_smtp_prdr ACL is called once for each recipient, after
489 the body content is received and before the acl_smtp_data ACL.
490 The client is controlled by bolth of: a hosts_try_prdr option
491 on the smtp transport, and the server advertisement.
492 Default client logging of deliveries and rejections involving
493 PRDR are flagged with the string "PRDR".
495 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
496 fclose(). Diagnosis by Todd Lyons.
498 PP/17 Update configure.default to handle IPv6 localhost better.
499 Patch by Alain Williams (plus minor tweaks).
502 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
503 This is now consistent with GnuTLS, and is now documented: the
504 previous undocumented portable approach to treating the option as
505 unset was to force an expansion failure. That still works, and
506 an empty string is now equivalent.
508 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
509 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
510 not performing validation itself.
512 PP/20 Added force_command boolean option to pipe transport.
513 Patch from Nick Koston, of cPanel Inc.
515 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
518 TF/04 Added udpsend ACL modifer and hexquote expansion operator
520 PP/21 Fix eximon continuous updating with timestamped log-files.
521 Broken in a format-string cleanup in 4.80, missed when I repaired the
522 other false fix of the same issue.
523 Report and fix from Heiko Schlichting.
526 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
527 Report from Prashanth Katuri.
529 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
530 It's SecureTransport, so affects any MacOS clients which use the
531 system-integrated TLS libraries, including email clients.
533 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
534 using a MIME ACL for non-SMTP local injection.
535 Report and assistance in diagnosis by Warren Baker.
537 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
539 JH/16 Fix comparisons for 64b. Bugzilla 1385.
541 TL/09 Add expansion variable $authenticated_fail_id to keep track of
542 last id that failed so it may be referenced in subsequent ACL's.
544 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
547 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
548 ldap library initialization, allowing self-signed CA's to be
549 used. Also properly sets require_cert option later in code by
550 using NULL (global ldap config) instead of ldap handle (per
551 session). Bug diagnosis and testing by alxgomz.
553 TL/12 Enhanced documentation in the ratelimit.pl script provided in
554 the src/util/ subdirectory.
556 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
557 renamed to Transport Post Delivery Action by Jeremy Harris, as
560 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
561 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
562 redis_servers = needs to be configured which will be used by the redis
563 lookup. Patch from Warren Baker, of The Packet Hub.
565 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
567 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
568 hostname or reverse DNS when processing a host list. Used suggestions
569 from multiple comments on this bug.
571 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
573 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
574 Missed a few lines, added it to make the runtest require no keyboard
577 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
578 contains upper case chars. Make router use caseful_local_part.
580 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
581 support when GnuTLS has been built with p11-kit.
587 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
589 This, or similar/improved, will also be change PP/11 of 4.82.
595 PP/01 Handle short writes when writing local log-files.
596 In practice, only affects FreeBSD (8 onwards).
597 Bugzilla 1053, with thanks to Dmitry Isaikin.
599 NM/01 Bugzilla 949 - Documentation tweak
601 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
604 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
606 PP/02 Implemented gsasl authenticator.
608 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
610 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
611 `pkg-config foo` for cflags/libs.
613 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
614 with rest of GSASL and with heimdal_gssapi.
616 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
617 `pkg-config foo` for cflags/libs for the TLS implementation.
619 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
620 properties get this fed in as external SSF. A number of robustness
621 and debugging improvements to the cyrus_sasl authenticator.
623 PP/08 cyrus_sasl server now expands the server_realm option.
625 PP/09 Bugzilla 1214 - Log authentication information in reject log.
626 Patch by Jeremy Harris.
628 PP/10 Added dbmjz lookup type.
630 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
632 PP/12 MAIL args handles TAB as well as SP, for better interop with
633 non-compliant senders.
634 Analysis and variant patch by Todd Lyons.
636 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
637 Bug report from Lars Müller <lars@samba.org> (via SUSE),
638 Patch from Dirk Mueller <dmueller@suse.com>
640 PP/13 tls_peerdn now print-escaped for spool files.
641 Observed some $tls_peerdn in wild which contained \n, which resulted
642 in spool file corruption.
644 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
645 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
646 or write after TLS renegotiation, which otherwise led to messages
649 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
650 as a tracking header (ie: a signed header comes before the signature).
651 Patch from Wolfgang Breyha.
653 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
654 comma-sep list; embedded commas doubled.
656 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
658 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
660 Report and patch from Dmitry Banschikov.
662 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
663 Removed SSL_clear() after SSL_new() which led to protocol negotiation
664 failures. We appear to now support TLS1.1+ with Exim.
666 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
667 lets Exim select keys and certificates based upon TLS SNI from client.
668 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
669 before an outbound SMTP session. New log_selector, +tls_sni.
671 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
672 NULL dereference. Report and patch from Alun Jones.
674 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
675 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
676 Not seeing resolver debug output on NetBSD, but suspect this is a
677 resolver implementation change.
679 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
680 Left warnings. Added "eximon gdb" invocation mode.
682 PP/21 Defaulting "accept_8bitmime" to true, not false.
684 PP/22 Added -bw for inetd wait mode support.
686 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
687 locate the relevant includes and libraries. Made this the default.
689 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
690 Bugzilla 1246, report and most of solution from Tomasz Kusy.
692 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
693 This may cause build issues on older platforms.
695 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
696 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
697 gnutls_require_mac & gnutls_require_protocols (no longer supported).
698 Added SNI support via GnuTLS too.
699 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
701 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
703 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
704 Applied second patch from Janne, implementing suggestion to default
705 multiple-strings-in-record handling to match SPF spec.
707 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
709 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
710 read-only, out of scope).
711 Patch from Wolfgang Breyha, report from Stuart Northfield.
713 PP/29 Fix three issues highlighted by clang analyser static analysis.
714 Only crash-plausible issue would require the Cambridge-specific
715 iplookup router and a misconfiguration.
716 Report from Marcin Mirosław.
718 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
720 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
721 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
722 As part of this, removing so much warning spew let me fix some minor
723 real issues in debug logging.
725 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
726 assignment on my part. Fixed.
728 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
729 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
730 Janne Snabb (who went above and beyond: thank you).
732 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
733 string otherwise requires a connection and a bunch more work and it's
734 relatively easy to get wrong. Should also expose TLS library linkage
737 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
738 64-bit ${eval} (JH/03).
740 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
741 GNU libc to support some of the 64-bit stuff, should not lead to
742 conflicts. Defined before os.h is pulled in, so if a given platform
743 needs to override this, it can.
745 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
746 protection layer was required, which is not implemented.
747 Bugzilla 1254, patch from Wolfgang Breyha.
749 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
750 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
751 tls_dhparam take prime identifiers. Also unbreak combination of
752 OpenSSL+DH_params+TLSSNI.
754 PP/39 Disable SSLv2 by default in OpenSSL support.
760 PP/01 Solaris build fix for Oracle's LDAP libraries.
761 Bugzilla 1109, patch from Stephen Usher.
763 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
765 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
768 TF/02 Fix a couple more cases where we did not log the error message
769 when unlink() failed. See also change 4.74-TF/03.
771 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
772 lock up or crash if it happened to be inside a call to libc when it
773 got a SIGUSR1 from exiwhat.
775 The SIGUSR1 handler appends the current process status to the process
776 log which is later printed by exiwhat. It used to use the general
777 purpose logging code to do this, but several functions it calls are
778 not safe for signals.
780 The new output code in the SIGUSR1 handler is specific to the process
781 log, and simple enough that it's easy to inspect for signal safety.
782 Removing some special cases also simplifies the general logging code.
783 Removing the spurious timestamps from the process log simplifies
786 TF/04 Improved ratelimit ACL condition.
788 The /noupdate option has been deprecated in favour of /readonly which
789 has clearer semantics. The /leaky, /strict, and /readonly update modes
790 are mutually exclusive. The update mode is no longer included in the
791 database key; it just determines when the database is updated. (This
792 means that when you upgrde Exim will forget old rate measurements.)
794 Exim now checks that the per_* options are used with an update mode that
795 makes sense for the current ACL. For example, when Exim is processing a
796 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
797 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
798 must specify per_mail/readonly. If you omit the update mode it defaults to
799 /leaky where that makes sense (as before) or /readonly where required.
801 The /noupdate option is now undocumented but still supported for
802 backwards compatibility. It is equivalent to /readonly except that in
803 ACLs where /readonly is required you may specify /leaky/noupdate or
804 /strict/noupdate which are treated the same as /readonly.
806 A useful new feature is the /count= option. This is a generalization
807 of the per_byte option, so that you can measure the throughput of other
808 aggregate values. For example, the per_byte option is now equivalent
809 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
811 The per_rcpt option has been generalized using the /count= mechanism
812 (though it's more complicated than the per_byte equivalence). When it is
813 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
814 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
815 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
816 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
817 Note that using per_rcpt with a non-readonly update mode in more than
818 one ACL will cause the recipients to be double-counted. (The per_mail
819 and per_byte options don't have this problem.)
821 The handling of very low rates has changed slightly. If the computed rate
822 is less than the event's count (usually one) then this event is the first
823 after a long gap. In this case the rate is set to the same as this event's
824 count, so that the first message of a spam run is counted properly.
826 The major new feature is a mechanism for counting the rate of unique
827 events. The new per_addr option counts the number of different
828 recipients that someone has sent messages to in the last time period. It
829 behaves like per_rcpt if all the recipient addresses are different, but
830 duplicate recipient addresses do not increase the measured rate. Like
831 the /count= option this is a general mechanism, so the per_addr option
832 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
833 example, measure the rate that a client uses different sender addresses
834 with the options per_mail/unique=$sender_address. There are further
835 details in the main documentation.
837 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
839 TF/06 Removed a few PCRE remnants.
841 TF/07 Automatically extract Exim's version number from tags in the git
842 repository when doing development or release builds.
844 PP/02 Raise smtp_cmd_buffer_size to 16kB.
845 Bugzilla 879. Patch from Paul Fisher.
847 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
848 Heavily based on revision 40f9a89a from Simon Arlott's tree.
851 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
853 PP/05 Variable $av_failed, true if the AV scanner deferred.
854 Bugzilla 1078. Patch from John Horne.
856 PP/06 Stop make process more reliably on build failure.
857 Bugzilla 1087. Patch from Heiko Schlittermann.
859 PP/07 Make maildir_use_size_file an _expandable_ boolean.
860 Bugzilla 1089. Patch from Heiko Schlittermann.
862 PP/08 Handle ${run} returning more data than OS pipe buffer size.
863 Bugzilla 1131. Patch from Holger Weiß.
865 PP/09 Handle IPv6 addresses with SPF.
866 Bugzilla 860. Patch from Wolfgang Breyha.
868 PP/10 GnuTLS: support TLS 1.2 & 1.1.
870 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
873 PP/11 match_* no longer expand right-hand-side by default.
874 New compile-time build option, EXPAND_LISTMATCH_RHS.
875 New expansion conditions, "inlist", "inlisti".
877 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
879 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
881 PP/14 fix log_write() format string regression from TF/03.
882 Bugzilla 1152. Patch from Dmitry Isaikin.
888 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
890 PP/02 Harmonised TLS library version reporting; only show if debugging.
891 Layout now matches that introduced for other libraries in 4.74 PP/03.
893 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
895 PP/04 New "dns_use_edns0" global option.
897 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
900 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
901 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
903 TK/01 Updated PolarSSL code to 0.14.2.
904 Bugzilla 1097. Patch from Andreas Metzler.
906 PP/07 Catch divide-by-zero in ${eval:...}.
909 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
912 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
913 format-string attack -- SECURITY: remote arbitrary code execution.
915 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
916 time unintentionally subject to list matching rules, letting the header
917 cause arbitrary Exim lookups (of items which can occur in lists, *not*
918 arbitrary string expansion). This allowed for information disclosure.
920 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
921 INT_MIN/-1 -- value coerced to INT_MAX.
927 NM/01 Workround for PCRE version dependency in version reporting
930 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
931 This fixes portability to compilers other than gcc, notably
932 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
934 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
935 makefiles for portability to HP-UX and POSIX correctness.
937 PP/01 Permit LOOKUP_foo enabling on the make command-line.
938 Also via indented variable definition in the Makefile.
939 (Debugging by Oliver Heesakkers).
941 PP/02 Restore caching of spamd results with expanded spamd_address.
942 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
944 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
945 Improves build reliability. Fix from: Frank Elsner
947 NM/02 Fix wide character breakage in the rfc2047 coding
948 Fixes bug 1064. Patch from Andrey N. Oktyabrski
950 NM/03 Allow underscore in dnslist lookups
951 Fixes bug 1026. Patch from Graeme Fowler
953 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
954 Code patches from Adam Ciarcinski of NetBSD.
956 NM/04 Fixed exiqgrep to cope with mailq missing size issue
959 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
960 is logged, to avoid truncation. Patch from John Horne.
962 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
963 Patch from Jakob Hirsch.
965 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
966 SQL string expansion failure details.
967 Patch from Andrey Oktyabrski.
969 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
970 Patch from Simon Arlott.
972 PP/09 New lookups functionality failed to compile on old gcc which rejects
973 extern declarations in function scope.
974 Patch from Oliver Fleischmann
976 PP/10 Use sig_atomic_t for flags set from signal handlers.
977 Check getgroups() return and improve debugging.
978 Fixed developed for diagnosis in bug 927 (which turned out to be
981 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
982 Patch from Mark Zealey.
984 PP/12 Bugzilla 1056: Improved spamd server selection.
985 Patch from Mark Zealey.
987 PP/13 Bugzilla 1086: Deal with maildir quota file races.
988 Based on patch from Heiko Schlittermann.
990 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
991 Patch from Uwe Doering, sign-off by Michael Haardt.
993 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
994 variable declaration deep within a block. Bug and patch from
997 PP/15 lookups-Makefile IRIX compatibilty coercion.
999 PP/16 Make DISABLE_DKIM build knob functional.
1001 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1002 Patch by Simon Arlott
1004 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1005 variable argument macros. Our copy now differs from upstream.
1011 TF/01 Failure to get a lock on a hints database can have serious
1012 consequences so log it to the panic log.
1014 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1015 controlled using the smtp_confirmation log selector.
1017 TF/03 Include the error message when we fail to unlink a spool file.
1019 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1020 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1021 for maintaining out-of-tree patches for some time.
1023 PP/01 Bugzilla 139: Documentation and portability issues.
1024 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1025 Handle per-OS dynamic-module compilation flags.
1027 PP/02 Let /dev/null have normal permissions.
1028 The 4.73 fixes were a little too stringent and complained about the
1029 permissions on /dev/null. Exempt it from some checks.
1030 Reported by Andreas M. Kirchwitz.
1032 PP/03 Report version information for many libraries, including
1033 Exim version information for dynamically loaded libraries. Created
1034 version.h, now support a version extension string for distributors
1035 who patch heavily. Dynamic module ABI change.
1037 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1038 privilege escalation vulnerability whereby the Exim run-time user
1039 can cause root to append content of the attacker's choosing to
1042 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1045 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1046 If dropping privileges for untrusted macros, we disabled normal logging
1047 on the basis that it would fail; for the Exim run-time user, this is not
1048 the case, and it resulted in successful deliveries going unlogged.
1049 Fixed. Reported by Andreas Metzler.
1055 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1056 only prepend for the Resent-* case. Fixes regression introduced in
1057 Exim 4.70 by NM/22 for Bugzilla 607.
1059 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1060 increasing numbers of administrators be bitten by this.
1062 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1064 PP/03 Bugzilla 994: added openssl_options main configuration option.
1066 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1068 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1070 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1072 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1073 without a peer certificate, leading to a segfault because of an
1074 assumption that peers always have certificates. Be a little more
1075 paranoid. Problem reported by Martin Tscholak.
1077 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1078 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1079 NB: ClamAV planning to remove STREAM in "middle of 2010".
1080 CL also introduces -bmalware, various -d+acl logging additions and
1081 more caution in buffer sizes.
1083 PP/09 Implemented reverse_ip expansion operator.
1085 PP/10 Bugzilla 937: provide a "debug" ACL control.
1087 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1089 PP/12 Bugzilla 973: Implement --version.
1091 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1093 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1095 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1097 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1098 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1099 ignore trailing whitespace.
1101 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1103 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1106 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1107 Notification from Dr Andrew Aitchison.
1109 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1110 ExtendedDetectionInfo response format.
1111 Notification from John Horne.
1113 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1116 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1117 XSL and documented dependency on system catalogs, with examples of how
1120 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1123 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1124 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1125 configuration file which is writeable by the Exim user or group.
1127 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1128 of configuration files to cover files specified with the -C option if
1129 they are going to be used with root privileges, not just the default
1132 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1133 option (effectively making it always true).
1135 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1136 files to be used while preserving root privileges.
1138 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1139 that rogue child processes cannot use them.
1141 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1142 run-time user, instead of root.
1144 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1145 Exim run-time user without dropping privileges.
1147 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1148 result string, instead of calling string_vformat() twice with the same
1151 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1152 for other users. Others should always drop root privileges if they use
1153 -C on the command line, even for a whitelisted configure file.
1155 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1157 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1163 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1164 $data_path, and $header_path variables; fixed documentation bugs and
1167 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1168 exipick to access non-standard spools, including the "frozen" queue
1171 NM/01 Bugzilla 965: Support mysql stored procedures.
1172 Patch from Alain Williams
1174 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1176 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1177 Patch from Andreas Metzler
1179 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1180 Patch from Kirill Miazine
1182 NM/05 Bugzilla 671: Added umask to procmail example.
1184 JJ/03 installed exipick 20100323.0, fixing doc bug
1186 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1187 directory. Notification and patch from Dan Rosenberg.
1189 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1191 TK/02 Improve log output when DKIM signing operation fails.
1193 MH/01 Treat the transport option dkim_domain as a colon separated
1194 list, not as a single string, and sign the message with each element,
1195 omitting multiple occurences of the same signer.
1197 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1198 Bugzilla 985, 986. Patch by Simon Arlott
1200 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1201 Patch by Simon Arlott
1203 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1204 MBX locking. Notification from Dan Rosenberg.
1210 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1212 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1214 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1216 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1218 NM/04 Bugzilla 915: Flag broken perl installation during build.
1224 TK/01 Added patch by Johannes Berg that expands the main option
1225 "spamd_address" if it starts with a dollar sign.
1227 TK/02 Write list of recipients to X-Envelope-Sender header when building
1228 the mbox-format spool file for content scanning (suggested by Jakob
1231 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1232 (http://www.dcc-servers.net/) support via dccifd. Activated by
1233 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1235 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1236 by Mark Daniel Reidel <mr@df.eu>.
1238 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1239 When building exim an external PCRE library is now needed -
1240 PCRE is a system library on the majority of modern systems.
1241 See entry on PCRE_LIBS in EDITME file.
1243 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1244 conversation. Added nologin parameter to request.
1245 Patch contributed by Kirill Miazine.
1247 TF/01 Do not log submission mode rewrites if they do not change the address.
1249 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1251 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1252 log files in place. Contributed by Roberto Lima.
1254 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1256 TF/03 Bugzilla 615: When checking the local_parts router precondition
1257 after a local_part_suffix or local_part_prefix option, Exim now
1258 does not use the address's named list lookup cache, since this
1259 contains cached lookups for the whole local part.
1261 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1262 Robert Millan. Documentation is in experimental-spec.txt.
1264 TF/04 Bugzilla 668: Fix parallel build (make -j).
1266 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1268 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1269 Patch provided by Jan Srzednicki.
1271 TF/05 Leading white space used to be stripped from $spam_report which
1272 wrecked the formatting. Now it is preserved.
1274 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1275 that they are available at delivery time.
1277 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1279 TF/08 TLS error reporting now respects the incoming_interface and
1280 incoming_port log selectors.
1282 TF/09 Produce a more useful error message if an SMTP transport's hosts
1283 setting expands to an empty string.
1285 NM/06 Bugzilla 744: EXPN did not work under TLS.
1286 Patch provided by Phil Pennock.
1288 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1289 Patch provided by Richard Godbee.
1291 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1292 acl_smtp_notquit, added index entry.
1294 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1295 Patch provided by Eugene Bujak.
1297 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1298 accept(). Patch provided by Maxim Dounin.
1300 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1301 Patch provided by Phil Pennock.
1303 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1305 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1306 Patch provided by Brad "anomie" Jorsch.
1308 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1309 Patch provided by Dean Brooks.
1311 TK/05 Add native DKIM support (does not depend on external libraries).
1313 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1314 Patch provided by Graeme Fowler.
1316 NM/16 Bugzilla 851: Documentation example syntax fix.
1318 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1320 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1323 NM/19 Bugzilla 745: TLS version reporting.
1324 Patch provided by Phil Pennock.
1326 NM/20 Bugzilla 167: bool: condition support.
1327 Patch provided by Phil Pennock.
1329 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1330 clients. Patch provided by Phil Pennock.
1332 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1333 Patch provided by Brad "anomie" Jorsch.
1335 NM/23 Bugzilla 687: Fix misparses in eximstats.
1336 Patch provided by Heiko Schlittermann.
1338 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1339 Patch provided by Heiko Schlittermann.
1341 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1342 plus update to original patch.
1344 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1346 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1347 Patch provided by David Brownlee.
1349 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1351 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1353 NM/30 Bugzilla 888: TLS documentation bugfixes.
1355 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1357 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1358 Unlike the original bugzilla I have changed all shell scripts in src tree.
1360 NM/33 Bugzilla 898: Transport filter timeout fix.
1361 Patch by Todd Rinaldo.
1363 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1364 Patch by Serge Demonchaux.
1366 NM/35 Bugzilla 39: Base64 decode bug fixes.
1367 Patch by Jakob Hirsch.
1369 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1371 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1373 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1375 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1381 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1382 ALT-N's libdkim that I have put here:
1383 http://duncanthrax.net/exim-experimental/
1385 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1386 were called 'true' and it seems that C99 defines that as a reserved
1387 keyword to be used with 'bool' variable types. That means you could
1388 not include C99-style headers which use bools without triggering
1389 build errors in sieve.c.
1391 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1392 as mailq or other aliases. Changed the --help handling significantly
1393 to do whats expected. exim_usage() emits usage/help information.
1395 SC/01 Added the -bylocaldomain option to eximstats.
1397 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1399 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1401 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1407 PH/01 Another patch from the Sieve maintainer.
1409 PH/02 When an IPv6 address is converted to a string for single-key lookup
1410 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1411 dots are used instead of colons so that keys in lsearch files need not
1412 contain colons. This was done some time before quoting was made available
1413 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1414 (notated using the quote facility) so as to distinguish them from IPv4
1415 keys. This meant that lookups for IP addresses in host lists did not work
1416 for iplsearch lookups.
1418 This has been fixed by arranging for IPv6 addresses to be expressed with
1419 colons if the lookup type is iplsearch. This is not incompatible, because
1420 previously such lookups could never work.
1422 The situation is now rather anomolous, since one *can* have colons in
1423 ordinary lsearch keys. However, making the change in all cases is
1424 incompatible and would probably break a number of configurations.
1426 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1429 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1430 conversion specification without a maximum field width, thereby enabling
1431 a rogue spamd server to cause a buffer overflow. While nobody in their
1432 right mind would setup Exim to query an untrusted spamd server, an
1433 attacker that gains access to a server running spamd could potentially
1434 exploit this vulnerability to run arbitrary code as the Exim user.
1436 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1437 $primary_hostname instead of what libspf2 thinks the hosts name is.
1439 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1440 a directory entry by the name of the lookup key. Previously, if a
1441 symlink pointed to a non-existing file or a file in a directory that
1442 Exim lacked permissions to read, a lookup for a key matching that
1443 symlink would fail. Now it is enough that a matching directory entry
1444 exists, symlink or not. (Bugzilla 503.)
1446 PH/03 The body_linecount and body_zerocount variables are now exported in the
1449 PH/04 Added the $dnslist_matched variable.
1451 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1452 This means they are set thereafter only if the connection becomes
1455 PH/06 Added the client_condition to authenticators so that some can be skipped
1456 by clients under certain conditions.
1458 PH/07 The error message for a badly-placed control=no_multiline_responses left
1459 "_responses" off the end of the name.
1461 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1463 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1464 (without spaces) instead of just copying the configuration text.
1466 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1468 PH/11 Added $max_received_linelength.
1470 PH/12 Added +ignore_defer and +include_defer to host lists.
1472 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1473 way in which PCRE > 7.0 is built.
1475 PH/14 Implemented queue_only_load_latch.
1477 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1478 MAIL command. The effect was to mangle the value on 64-bit systems.
1480 PH/16 Another patch from the Sieve maintainer.
1482 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1484 PH/18 If a system quota error occurred while trying to create the file for
1485 a maildir delivery, the message "Mailbox is full" was not appended to the
1486 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1487 only to a quota excession during the actual writing of the file.
1489 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1490 characters?) which causes problems in log lines. The DN values are now
1491 passed through string_printing() before being added to log lines.
1493 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1494 and InterBase are left for another time.)
1496 PH/21 Added message_body_newlines option.
1498 PH/22 Guard against possible overflow in moan_check_errorcopy().
1500 PH/23 POSIX allows open() to be a macro; guard against that.
1502 PH/24 If the recipient of an error message contained an @ in the local part
1503 (suitably quoted, of course), incorrect values were put in $domain and
1504 $local_part during the evaluation of errors_copy.
1510 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1511 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1514 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1515 issue a MAIL command.
1517 PH/02 In an ACL statement such as
1519 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1521 if a client was not listed at all, or was listed with a value other than
1522 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1523 the condition was not true (as it should be), so access was not denied.
1524 The bug was that the ! inversion was incorrectly passed on to the second
1525 item. This has been fixed.
1527 PH/03 Added additional dnslists conditions == and =& which are different from
1528 = and & when the dns lookup returns more than one IP address.
1530 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1531 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1533 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1534 FSYNC, which compiles an option called disable_fsync that allows for
1535 bypassing fsync(). The documentation is heavily laced with warnings.
1537 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1539 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1540 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1541 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1542 including adding "make clean"; (3) Added -fPIC when compiling the test
1543 dynamically loaded module, to get rid of a warning.
1545 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1546 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1547 = 0s. The bug is otherwise harmless.
1549 PH/07 There was a bug in the dovecot authenticator such that the value of
1550 $auth1 could be overwritten, and so not correctly preserved, after a
1551 successful authentication. This usually meant that the value preserved by
1552 the server_setid option was incorrect.
1554 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1556 PH/09 Installed PCRE release 7.0.
1558 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1559 run for batched SMTP input. It is now run at the start of every message
1560 in the batch. While fixing this I discovered that the process information
1561 (output by running exiwhat) was not always getting set for -bs and -bS
1562 input. This is fixed, and it now also says "batched" for BSMTP.
1564 PH/11 Added control=no_pipelining.
1566 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1567 patch, slightly modified), and move the expansion of helo_data till after
1568 the connection is made in the smtp transport (so it can use these
1571 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1573 PH/14 Added log_selector = +pid.
1575 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1577 PH/16 Add ${if forany and ${if forall.
1579 PH/17 Added dsn_from option to vary the From: line in DSNs.
1581 PH/18 Flush SMTP output before performing a callout, unless control =
1582 no_callout_flush is set.
1584 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1585 was true (the default) a successful delivery failed to delete the retry
1586 item, thus causing premature timeout of the address. The bug is now
1589 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1591 PH/21 Long custom messages for fakedefer and fakereject are now split up
1592 into multiline reponses in the same way that messages for "deny" and
1593 other ACL rejections are.
1595 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1596 with slight modification.
1598 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1599 draft, changing the syntax and factoring some duplicate code".
1601 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1602 for deliveries of the second and subsequent messages over the same SMTP
1605 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1606 ${reduce, with only minor "tidies".
1608 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1610 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1611 expansion side effects.
1613 PH/27 When a message times out after an over-quota error from an Exim-imposed
1614 quota, the bounce message says "mailbox is full". This message was not
1615 being given when it was a system quota that was exceeded. It now should
1618 MH/03 Made $recipients available in local_scan(). local_scan() already has
1619 better access to the recipient list through recipients_list[], but
1620 $recipients can be useful in postmaster-provided expansion strings.
1622 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1623 in the case of a MAIL command with additional options following the
1624 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1625 were accidentally chopped off.
1627 PH/29 SMTP synchronization checks are implemented when a command is read -
1628 there is a check that no more input is waiting when there shouldn't be
1629 any. However, for some commands, a delay in an ACL can mean that it is
1630 some time before the response is written. In this time, more input might
1631 arrive, invalidly. So now there are extra checks after an ACL has run for
1632 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1633 pipelining has not been advertised.
1635 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1637 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1638 returned in the SMTP response when smtp_return_error_details was false.
1639 This has been fixed.
1641 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1642 instead of the C I/O that was originally supplied, because problems were
1643 reported on Solaris.
1645 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1646 Exim which did not show up earlier: it was assuming that a call to
1647 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1648 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1649 was a macro that became an assignment, so it seemed to work. This has
1650 changed to a proper function call with a void return, hence the compile
1651 error. Exim's code has been fixed.
1653 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1656 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1658 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1660 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1661 because it turns out to be too big or there is a timeout) there is no
1662 "Completed" line in the log. When some messages of this type were
1663 selected by exigrep, they were listed as "not completed". Others were
1664 picked up by some special patterns. I have improved the selection
1665 criteria to be more general.
1667 PH/38 The host_find_failed option in the manualroute router can now be set
1668 to "ignore", to completely ignore a host whose IP address cannot be
1669 found. If all hosts are ignored, the behaviour is controlled by the new
1670 host_all_ignored option.
1672 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1673 homing or because of multiple MX records with /mx) generated more than
1674 one IP address, and the following item turned out to be the local host,
1675 all the secondary addresses of the first item were incorrectly removed
1676 from the list, along with the local host and any following hosts (which
1677 is what is supposed to happen).
1679 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1680 whoever called Exim into the -H file. In the case of the daemon it was
1681 behaving confusingly. When first started, it used values for whoever
1682 started the daemon, but after a SIGHUP it used the Exim user (because it
1683 calls itself on a restart). I have changed the code so that it now always
1686 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1687 message are rejected with the same error (e.g. no authentication or bad
1688 sender address), and a DATA command is nevertheless sent (as can happen
1689 with PIPELINING or a stupid MUA), the error message that was given to the
1690 RCPT commands is included in the rejection of the DATA command. This is
1691 intended to be helpful for MUAs that show only the final error to their
1694 PH/42 Another patch from the Sieve maintainer.
1696 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1697 Eximstats - Fixed some broken HTML links and added missing column headers
1699 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1700 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1702 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1704 PH/43 Yet another patch from the Sieve maintainer.
1706 PH/44 I found a way to check for a TCP/IP connection going away before sending
1707 the response to the final '.' that terminates a message, but only in the
1708 case where the client has not sent further data following the '.'
1709 (unfortunately, this is allowed). However, in many cases there won't be
1710 any further data because there won't be any more messages to send. A call
1711 to select() can be used: if it shows that the input is "ready", there is
1712 either input waiting, or the socket has been closed. An attempt to read
1713 the next input character can distinguish the two cases. Previously, Exim
1714 would have sent an OK response which the client would never have see.
1715 This could lead to message repetition. This fix should cure that, at
1716 least in a lot of common cases.
1718 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1719 advertised in response to EHLO.
1725 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1726 fixed by 4.65/MH/01 (is this a record?) are fixed:
1728 (i) An empty string was always treated as zero by the numeric comparison
1729 operators. This behaviour has been restored.
1731 (ii) It is documented that the numeric comparison operators always treat
1732 their arguments as decimal numbers. This was broken in that numbers
1733 starting with 0 were being interpreted as octal.
1735 While fixing these problems I realized that there was another issue that
1736 hadn't been noticed. Values of message_size_limit (both the global option
1737 and the transport option) were treated as octal if they started with 0.
1738 The documentation was vague. These values are now always treated as
1739 decimal, and I will make that clear in the documentation.
1745 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1746 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1749 MH/01 Don't check that the operands of numeric comparison operators are
1750 integers when their expansion is in "skipping" mode (fixes bug
1751 introduced by 4.64-PH/07).
1753 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1754 child addresses, Exim now panics and dies. Previously, because the count
1755 is held in a short int, deliveries were likely to be lost. As such a
1756 large number of recipients for a single message is ridiculous
1757 (performance will be very, very poor), I have chosen to impose a limit
1758 rather than extend the field.
1764 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1765 leftover -K file (the existence of which was triggered by #402).
1766 While we were at it, introduced process PID as part of the -K
1767 filename. This should rule out race conditions when creating
1770 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1771 processing considerably. Previous code took too long for large mails,
1772 triggering a timeout which in turn triggers #401.
1774 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1775 in the DK code in transports.c. sendfile() is not really portable,
1776 hence the _LINUX specificness.
1778 TF/01 In the add_headers option to the mail command in an Exim filter,
1779 there was a bug that Exim would claim a syntax error in any
1780 header after the first one which had an odd number of characters
1783 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1784 callout verification, Exim cached a "reject" for the entire domain. This
1785 is correct for most verifications, but it is not correct for a recipient
1786 verification with use_sender or use_postmaster set, because in that case
1787 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1788 case of MAIL FROM:<> rejection from other early rejections (e.g.
1789 rejection of HELO). When verifying a recipient using a non-null MAIL
1790 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1791 Whatever the result of the callout, the value of the domain cache is
1792 left unchanged (for any other kind of callout, getting as far as trying
1793 RCPT means that the domain itself is ok).
1795 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1798 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1799 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1800 session, the daemon ignores ECONNECT errors and logs others; it now
1801 ignores EPIPE as well.
1803 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1804 (quoted-printable decoding).
1806 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1807 later the small subsequent patch to fix an introduced bug.
1809 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1811 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1813 PH/08 An error is now given if message_size_limit is specified negative.
1815 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1816 to be given (somewhat) arbitrary names.
1818 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1821 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1822 miscellaneous code fixes
1824 PH/10 Added the log_reject_target ACL modifier to specify where to log
1827 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1828 hostname. This is wrong, because it relates to the incoming message (and
1829 probably the interface on which it is arriving) and not to the outgoing
1830 callout (which could be using a different interface). This has been
1831 changed to use the value of the helo_data option from the smtp transport
1832 instead - this is what is used when a message is actually being sent. If
1833 there is no remote transport (possible with a router that sets up host
1834 addresses), $smtp_active_hostname is used.
1836 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1837 tweaks were necessary in order to get it to work (see also 21 below):
1838 (a) The code assumed that strncpy() returns a negative number on buffer
1839 overflow, which isn't the case. Replaced with Exim's string_format()
1841 (b) There were several signed/unsigned issues. I just did the minimum
1842 hacking in of casts. There is scope for a larger refactoring.
1843 (c) The code used strcasecmp() which is not a standard C function.
1844 Replaced with Exim's strcmpic() function.
1845 (d) The code set only $1; it now sets $auth1 as well.
1846 (e) A simple test gave the error "authentication client didn't specify
1847 service in request". It would seem that Dovecot has changed its
1848 interface. Fortunately there's a specification; I followed it and
1849 changed what the client sends and it appears to be working now.
1851 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1854 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1855 address A is aliased to B and C, where B exists and C does not. Without
1856 -v the output is "A verified" because verification stops after a
1857 successful redirection if more than one address is generated. However,
1858 with -v the child addresses are also verified. Exim was outputting "A
1859 failed to verify" and then showing the successful verification for C,
1860 with its parentage. It now outputs "B failed to verify", showing B's
1861 parentage before showing the successful verification of C.
1863 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1864 look up a TXT record in a specific list after matching in a combined
1867 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1868 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1869 they consult the DNS. I had assumed they would set it the way they
1870 wanted; and indeed my experiments on Linux seem to show that in some
1871 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1872 To be on the safe side, however, I have now made the interface to
1873 host_find_byname() similar to host_find_bydns(), with an argument
1874 containing the DNS resolver options. The host_find_byname() function now
1875 sets these options at its start, just as host_find_bydns() does. The smtp
1876 transport options dns_qualify_single and dns_search_parents are passed to
1877 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1878 of host_find_byname() use the default settings of RES_DEFNAMES
1879 (qualify_single) but not RES_DNSRCH (search_parents).
1881 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1882 spool_read_header() do less string testing, by means of a preliminary
1883 switch on the second character of optional "-foo" lines. (This is
1884 overdue, caused by the large number of possibilities that now exist.
1885 Originally there were few.) While I was there, I also converted the
1886 str(n)cmp tests so they don't re-test the leading "-" and the first
1887 character, in the hope this might squeeze out yet more improvement.
1889 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1890 flag allowing group syntax was set by the header_syntax check but not
1891 turned off, possible causing trouble later; (2) The flag was not being
1892 set at all for the header_verify test, causing "group"-style headers to
1893 be rejected. I have now set it in this case, and also caused header_
1894 verify to ignore an empty address taken from a group. While doing this, I
1895 came across some other cases where the code for allowing group syntax
1896 while scanning a header line wasn't quite right (mostly, not resetting
1897 the flag correctly in the right place). These bugs could have caused
1898 trouble for malformed header lines. I hope it is now all correct.
1900 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1901 with the "reply" argument non-NULL. The code, however (which originally
1902 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1903 but it didn't always do it. This confused somebody who was copying the
1904 code for some other use. I have removed all the tests.
1906 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1907 feature that was used to support insecure browsers during the U.S. crypto
1908 embargo. It requires special client support, and Exim is probably the
1909 only MTA that supported it -- and would never use it because real RSA is
1910 always available. This code has been removed, because it had the bad
1911 effect of slowing Exim down by computing (never used) parameters for the
1912 RSA_EXPORT functionality.
1914 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1915 authenticator to fail if there's a tab character in the incoming data
1916 (there should never be unless someone is messing about, as it's supposed
1917 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1918 if the connection is using TLS or if the remote IP is the same as the
1919 local IP, and the "valid-client-cert option" if a client certificate has
1922 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1923 authenticators. This can be used for authorization after authentication
1924 succeeds. (In the case of plaintext, it servers for both authentication
1927 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1928 if any retry times were supplied.
1930 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1931 connection, where there is no client IP address to check. In this
1932 situation, the verify now always succeeds.
1934 PH/25 Applied John Jetmore's -Mset patch.
1936 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1938 PH/27 In a string expansion for a processed (not raw) header when multiple
1939 headers of the same name were present, leading whitespace was being
1940 removed from all of them, but trailing whitespace was being removed only
1941 from the last one. Now trailing whitespace is removed from each header
1942 before concatenation. Completely empty headers in a concatenation (as
1943 before) are ignored.
1945 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1946 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1948 PH/29 [Removed. This was a change that I later backed out, and forgot to
1949 correct the ChangeLog entry (that I had efficiently created) before
1950 committing the later change.]
1952 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1953 address errors (4xx response to RCPT) over the same connection as other
1954 messages routed to the same hosts. Such deliveries are always "forced",
1955 so retry times are not inspected. This resulted in far too many retries
1956 for the affected addresses. The effect occurred only when there were more
1957 hosts than the hosts_max_try setting in the smtp transport when it had
1958 the 4xx errors. Those hosts that it had tried were not added to the list
1959 of hosts for which the message was waiting, so if all were tried, there
1960 was no problem. Two fixes have been applied:
1962 (i) If there are any address or message errors in an SMTP delivery, none
1963 of the hosts (tried or untried) are now added to the list of hosts
1964 for which the message is waiting, so the message should not be a
1965 candidate for sending over the same connection that was used for a
1966 successful delivery of some other message. This seems entirely
1967 reasonable: after all the message is NOT "waiting for some host".
1968 This is so "obvious" that I'm not sure why it wasn't done
1969 previously. Hope I haven't missed anything, but it can't do any
1970 harm, as the worst effect is to miss an optimization.
1972 (ii) If, despite (i), such a delivery is accidentally attempted, the
1973 routing retry time is respected, so at least it doesn't keep
1974 hammering the server.
1976 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1977 in ${readsocket because some servers need this prod.
1979 PH/32 Added some extra debug output when updating a wait-xxx database.
1981 PH/33 The hint "could be header name not terminated by colon", which has been
1982 given for certain expansion errors for a long time, was not being given
1983 for the ${if def:h_colon_omitted{... case.
1985 PH/34 The spec says: "With one important exception, whenever a domain list is
1986 being scanned, $domain contains the subject domain." There was at least
1987 one case where this was not true.
1989 PH/35 The error "getsockname() failed: connection reset by peer" was being
1990 written to the panic log as well as the main log, but it isn't really
1991 panic-worthy as it just means the connection died rather early on. I have
1992 removed the panic log writing for the ECONNRESET error when getsockname()
1995 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1996 runs only) independently of the message's sender address. This meant
1997 that, if the 4xx error was in fact related to the sender, a different
1998 message to the same recipient with a different sender could confuse
1999 things. In particualar, this can happen when sending to a greylisting
2000 server, but other circumstances could also provoke similar problems.
2001 I have changed the default so that the retry time for these errors is now
2002 based a combination of the sender and recipient addresses. This change
2003 can be overridden by setting address_retry_include_sender=false in the
2006 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2007 remote server are returned as part of bounce messages. This was not
2008 happening for LMTP over a pipe (the lmtp transport), but now it is the
2009 same for both kinds of LMTP.
2011 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2012 in header lines that were in fact CNAMEs. This is no longer the case.
2014 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2015 and queue runs started by the daemon processed all messages. This has
2016 been fixed so that -R and -S can now usefully be given with -q<time>.
2018 PH/40 Import PCRE release 6.7 (fixes some bugs).
2020 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2022 PH/42 Give an error if -q is specified more than once.
2024 PH/43 Renamed the variables $interface_address and $interface_port as
2025 $received_ip_address and $received_port, to make it clear that these
2026 values apply to message reception, and not to the outgoing interface when
2027 a message is delivered. (The old names remain recognized, of course.)
2029 PH/44 There was no timeout on the connect() call when using a Unix domain
2030 socket in the ${readsocket expansion. There now is.
2032 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2033 be meaningful with "accept".
2035 SC/01 Eximstats V1.43
2036 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2038 SC/02 Eximstats V1.44
2039 Use a glob alias rather than an array ref in the generated
2040 parser. This improves both readability and performance.
2042 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2043 Collect SpamAssassin and rejection statistics.
2044 Don't display local sender or destination tables unless
2045 there is data to show.
2046 Added average volumes into the top table text output.
2048 SC/04 Eximstats V1.46
2049 Collect data on the number of addresses (recipients)
2050 as well as the number of messages.
2052 SC/05 Eximstats V1.47
2053 Added 'Message too big' to the list of mail rejection
2054 reasons (thanks to Marco Gaiarin).
2056 SC/06 Eximstats V1.48
2057 Mainlog lines which have GMT offsets and are too short to
2058 have a flag are now skipped.
2060 SC/07 Eximstats V1.49 (Alain Williams)
2061 Added the -emptyok flag.
2063 SC/08 Eximstats V1.50
2064 Fixes for obtaining the IP address from reject messages.
2066 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2067 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2068 whitesspace changes from 4.64-PH/27
2070 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2073 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2074 are found, allow negative numbers in numeric criteria)
2076 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2078 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2079 to match changes made in 4.64-PH/43
2081 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2083 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2084 "domain" is actually an IP address, and force a failure. This locks out
2085 those revolvers/nameservers that support "A-for-A" lookups, in
2086 contravention of the specifications.
2088 PH/48 When a host name was looked up from an IP address, and the subsequent
2089 forward lookup of the name timed out, the host name was left in
2090 $sender_host_name, contrary to the specification.
2092 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2093 restricted to single-key lookups, Exim was not diagnosing an error if
2094 * or *@ was used with a query-style lookup.
2096 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2098 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2099 long ago, but noone interested enough thought of it. Let's just say that
2100 the "1.1" means that there are some new functions that weren't there at
2101 some point in the past.
2103 PH/51 Error processing for expansion failure of helo_data from an smtp
2104 transport during callout processing was broken.
2106 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2107 tested/used via the -bh/-bhc/-bs options.
2109 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2110 bug, fixed in subsequent PCRE releases).
2112 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2113 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2115 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2121 SC/01 Use a glob alias rather than an array ref in eximstats generated
2122 parser. This improves both readability and performance.
2124 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2125 Don't display local sender or destination tables in eximstats unless
2126 there is data to show.
2127 Added average volumes into the eximstats top table text output.
2129 SC/03 Collect data on the number of addresses (recipients) as well
2130 as the number of messages in eximstats.
2132 TF/01 Correct an error in the documentation for the redirect router. Exim
2133 does (usually) call initgroups() when daemonizing.
2135 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2136 with consistent privilege compared to when running as a daemon.
2138 TF/03 Note in the spec that $authenticated_id is not set for local
2139 submissions from trusted users.
2141 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2142 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2144 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2145 by adding some example configuration directives to the default
2146 configuration file. A little bit of work is required to uncomment the
2147 directives and define how usernames and passwords are checked, but
2148 there is now a framework to start from.
2150 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2151 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2152 without this. I don't know how relevant this is to other LDAP libraries.
2154 PH/02 Add the verb name to the "unknown ACL verb" error.
2156 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2158 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2160 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2161 directory not expanded when it should be if an expanded home directory
2162 was set for the address (which is overridden by the transport).
2164 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2167 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2168 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2169 because it is too late at that time, and has no effect.
2171 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2172 security issue with \' (bugzilla #107). I could not use the
2173 PQescapeStringConn() function, because it needs a PGconn value as one of
2176 PH/08 When testing addresses using -bt, indicate those final addresses that
2177 are duplicates that would not cause an additional delivery. At least one
2178 person was confused, thinking that -bt output corresponded to deliveries.
2179 (Suppressing duplicates isn't a good idea as you lose the information
2180 about possibly different redirections that led to the duplicates.)
2182 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2183 systems where poll() doesn't work, in particular OS X.
2185 PH/10 Added more information to debugging output for retry time not reached.
2187 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2188 operations in malware.c.
2190 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2193 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2194 syslog_duplication set false, log lines that would normally be written
2195 both the the main log and to the reject log were not written to syslog at
2198 PH/14 In the default configuration, change the use of "message" in ACL warn
2199 statements to "add_header".
2201 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2202 not followed by a command (e.g. "seen endif").
2204 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2205 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2208 PH/17 Added extra conditions to the default value of delay_warning_condition
2212 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2213 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2214 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2217 The Auto-Submitted: and various List- headers are standardised, whereas I
2218 don't think Precedence: ever was.
2220 PH/18 Refactored debugging code in route_finduser() to show more information,
2221 in particular, the error code if getpwnam() issues one.
2223 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2224 This is apparently needed in addition to the PH/07 change above to avoid
2225 any possible encoding problems.
2227 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2228 but not after initializing Perl.
2230 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2231 output them only if debugging. By default they are written stderr,
2232 apparently, which is not desirable.
2234 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2237 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2240 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2242 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2243 authenticated or an ident call has been made. Suppress the default
2244 values for $authenticated_id and $authenticated_sender (but permit -oMai
2245 and -oMas) when testing with -bh.
2247 PH/24 Re-jigged the order of the tests in the default configuration so that the
2248 tests for valid domains and recipients precede the DNS black list and CSA
2249 tests, on the grounds that those ones are more expensive.
2251 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2252 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2253 This bug exists in every version of Exim that I still have, right back to
2256 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2257 However, an attempt to turn on case-sensitivity in a regex key by
2258 including (?-i) didn't work because the subject string was already
2259 lowercased, and the effects were non-intuitive. It turns out that a
2260 one-line patch can be used to allow (?-i) to work as expected.
2266 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2267 other effects) broke the use of negated acl sub-conditions.
2269 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2272 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2273 "Deny" causes Exim to reject the incoming connection with a 554 error.
2274 Unfortunately, if there is a major crisis, such as a disk failure,
2275 tcp-wrappers gives "deny", whereas what one would like would be some
2276 kind of temporary error. A kludge has been added to help with this.
2277 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2278 554 error is used if errno is still zero or contains ENOENT (which occurs
2279 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2282 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2284 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2285 errors. Otherwise a message that provokes a temporary error (when other
2286 messages do not) can cause a whole host to time out.
2288 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2289 addresses were routed directly to files or pipes from a redirect router.
2290 File deliveries just didn't batch; pipe deliveries might have suffered
2293 PH/06 A failure to get a lock for a hints database would erroneously always say
2294 "Failed to get write lock", even when it was really a read lock.
2296 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2297 of 0600. This has been changed to use the value of the lockfile_mode
2298 option (which defaults to 0600).
2300 PH/08 Applied small patch from the Sieve maintainer.
2302 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2303 folder from quota calculations, a direct delivery into this folder messed
2304 up the contents of the maildirsize file. This was because the regex was
2305 used only to exclude .Trash (or whatever) when the size of the mailbox
2306 was calculated. There was no check that a delivery was happening into an
2307 excluded directory. This bug has been fixed by ignoring all quota
2308 processing for deliveries into excluded directories.
2310 PH/10 Added the maildirfolder_create_regex option to appendfile.
2316 PH/01 The code for finding all the local interface addresses on a FreeBSD
2317 system running IPv6 was broken. This may well have applied to all BSD
2318 systems, as well as to others that have similar system calls. The broken
2319 code found IPv4 interfaces correctly, but gave incorrect values for the
2320 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2321 that it would not match correctly against @[] and not recognize the IPv6
2324 PH/02 The ipliteral router was not recognizing addresses of the form user@
2325 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2327 PH/03 Added disable_ipv6.
2329 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2330 decoded form, because it is most often used to construct To: headers
2331 lines in autoreplies, and the decoded form may well be syntactically
2332 invalid. However, $reply_address has leading white space removed, and all
2333 newlines turned into spaces so that the autoreply transport does not
2336 PH/05 If group was specified without a user on a router, and no group or user
2337 was specified on a transport, the group from the router was ignored.
2339 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2340 for visible compile-time settings that can be used to change these
2341 numbers, for those that want even more. Backwards compatibility with old
2342 spool files has been maintained. However, going back to a previous Exim
2343 release will lost any variables that are in spool files.
2345 PH/07 Two small changes when running in the test harness: increase delay when
2346 passing a TCP/IP connection to a new process, in case the original
2347 process has to generate a bounce, and remove special handling of
2348 127.0.0.2 (sic), which is no longer necessary.
2350 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2351 be the same on different OS.
2353 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2356 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2357 whether --show-vars was specified or not
2359 JJ/02 exipick: Added support for new ACL variable spool format introduced
2362 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2363 syntactically invalid From: or Reply-to: line, and a filter used this to
2364 generate an autoreply, and therefore failed to obtain an address for the
2365 autoreply, Exim could try to deliver to a non-existent relative file
2366 name, causing unrelated and misleading errors. What now happens is that
2367 it logs this as a hard delivery error, but does not attempt to create a
2370 PH/11 The exinext utility has a -C option for testing purposes, but although
2371 the given file was scanned by exinext itself; it wasn't being passed on
2372 when Exim was called.
2374 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2375 an end-of-file indication when reading a command response.
2377 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2378 compiled. In many other places in Exim, IPv6 addresses are always
2379 recognized, so I have changed this. It also means that IPv4 domain
2380 literals of the form [IPV4:n.n.n.n] are now always recognized.
2382 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2383 used if the router is not running as root, for example, when verifying at
2384 ACL time, or when using -bh. The debugging output from this situation was
2385 non-existent - all you got was a failure to exec. I have made two
2388 (a) Failures to set uid/gid, the current directory, or a process leader
2389 in a subprocess such as that created by queryprogram now generate
2390 suitable debugging ouput when -d is set.
2392 (b) The queryprogram router detects when it is not running as root,
2393 outputs suitable debugging information if -d is set, and then runs
2394 the subprocess without attempting to change uid/gid.
2396 PH/15 Minor change to Makefile for building test_host (undocumented testing
2399 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2400 additional section of a DNS packet that returns MX or SRV records.
2401 Instead, it always explicitly searches for A/AAAA records. This avoids
2402 major problems that occur when a DNS server includes only records of one
2403 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2404 fixed another bug: if SRV records were looked up and the corresponding
2405 address records were *not* found in the additional section, the port
2406 values from the SRV records were lost.
2408 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2409 using the correct key (the original address) when searching the retry
2410 rules in order to find which one to use for generating the retry hint.
2412 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2413 adding the default one. Similarly, if it contains a Reply-To: header, the
2414 errors_reply_to option, if set, is not used.
2416 PH/19 When calculating a retry time, Exim used to measure the "time since
2417 failure" by looking at the "first failed" field in the retry record. Now
2418 it does not use this if it is later than than the arrival time of the
2419 message. Instead it uses the arrival time. This makes for better
2420 behaviour in cases where some deliveries succeed, thus re-setting the
2421 "first failed" field. An example is a quota failure for a huge message
2422 when small messages continue to be delivered. Without this change, the
2423 "time since failure" will always be short, possible causing more frequent
2424 delivery attempts for the huge message than are intended.
2425 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2427 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2428 $1, $2, $3) because the numerical variables can be reset during some
2429 expansion items (e.g. "match"), thereby losing the authentication data.
2431 PH/21 Make -bV show the size of off_t variables so that the test suite can
2432 decide whether to run tests for quotas > 2G.
2434 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2435 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2436 filecount value is greater than 2G or if a quota value is greater than 2G
2437 on a system where the size of off_t is not greater than 4, a panic error
2440 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2441 never match. The debug and -bh output now contains an explicit error
2442 message indicating a malformed IPv4 address or mask.
2444 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2445 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2446 PH/23 above applies.
2448 PH/25 Do not write to syslog when running in the test harness. The only
2449 occasion when this arises is a failure to open the main or panic logs
2450 (for which there is an explicit test).
2452 PH/26 Added the /no_tell option to "control=freeze".
2454 PH/27 If a host name lookup failed very early in a connection, for example, if
2455 the IP address matched host_lookup and the reverse lookup yielded a name
2456 that did not have a forward lookup, an error message of the form "no IP
2457 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2458 could be logged. Now it outputs the IP address instead of "NULL".
2460 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2461 allows the sender and the authenticated sender to be set when
2462 submitting a message from within Exim. Since child_open_exim() is
2463 documented for local_scan(), the new function should be too.
2465 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2466 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2467 results in an empty string is now treated as unset.
2469 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2471 PH/31 Added sender_verify_fail logging option.
2473 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2474 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2475 tidied the source and removed it altogether.
2477 PH/33 When a queue run was abandoned because the load average was too high, a
2478 log line was always written; now it is written only if the queue_run log
2479 selector is set. In addition, the log line for abandonment now contains
2480 information about the queue run such as the pid. This is always present
2481 in "start" and "stop" lines but was omitted from the "abandon" line.
2483 PH/34 Omit spaces between a header name and the colon in the error message that
2484 is given when verify = headers_syntax fails (if there are lots of them,
2485 the message gets confusing).
2487 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2488 names, as there are now some PTR records that contain slashes. This check
2489 is only to protect against broken name servers that fall over on strange
2490 characters, so the fact that it applies to all lookups doesn't matter.
2492 PH/36 Now that the new test suite is complete, we can remove some of the
2493 special code in Exim that was needed for the old test suite. For example,
2494 sorting DNS records because real resolvers return them in an arbitrary
2495 order. The new test suite's fake resolver always returns records in the
2498 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2499 bounces) except when queue_only is set, to avoid logging races between
2500 the different processes.
2502 PH/38 Panic-die if .include specifies a non-absolute path.
2504 PH/39 A tweak to the "H" retry rule from its user.
2506 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2507 a label. They prevented compilation on older perls.
2509 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2510 a warning to be raised on newish perls.
2512 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2513 on queue. Changes to match documented behaviour of showing count of
2514 messages matching specified criteria.
2516 PH/40 Changed the default ident timeout from 30s to 5s.
2518 PH/41 Added support for the use of login_cap features, on those BSD systems
2519 that have them, for controlling the resources used by pipe deliveries.
2521 PH/42 The content-scanning code uses fopen() to create files in which to put
2522 message data. Previously it was not paying any attention to the mode of
2523 the files. Exim runs with umask(0) because the rest of the code creates
2524 files with open(), and sets the required mode explicitly. Thus, these
2525 files were ending up world-writeable. This was not a big issue, because,
2526 being within the spool directory, they were not world-accessible. I have
2527 created a function called modefopen, which takes an additional mode
2528 argument. It sets umask(777), creates the file, chmods it to the required
2529 mode, then resets the umask. All the relevant calls to fopen() in the
2530 content scanning code have been changed to use this function.
2532 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2533 to 24 hours. This avoids potential overflow problems when processing G
2534 and H retry rules. I suspect nobody ever tinkers with this value.
2536 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2538 PH/45 When the plaintext authenticator is running as a client, the server's
2539 challenges are checked to ensure they are valid base64 strings. By
2540 default, the authentication attempt is cancelled if an invalid string is
2541 received. Setting client_ignore_invalid_base64 true ignores these errors.
2542 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2543 they are received. Thus, the responses can be made to depend on the
2544 challenges. If an invalid string is ignored, an empty string is placed in
2547 PH/46 Messages that are created by the autoreply transport now contains a
2548 References: header, in accordance with RFCs 2822 and 3834.
2550 PH/47 Added authenticated_sender_force to the smtp transport.
2552 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2554 PH/49 Installed latest patch from the Sieve maintainer.
2556 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2557 was also set, the appendfile transport was unnecessarily scanning a
2558 directory of message files (e.g. for maildir delivery) to find the count
2559 of files (along with the size), even though it did not need this
2560 information. It now does the scan only if it needs to find either the
2561 size of the count of files.
2563 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2565 PH/52 Two bugs concerned with error handling when the smtp transport is
2568 (i) Exim was not creating retry information for temporary errors given
2569 for individual recipients after the DATA command when the smtp transport
2570 was used in LMTP mode. This meant that they could be retried too
2571 frequently, and not timed out correctly.
2573 (ii) Exim was setting the flag that allows error details to be returned
2574 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2575 recipients that were returned after the DATA command.
2577 PH/53 This is related to PH/52, but is more general: for any failing address,
2578 when detailed error information was permitted to be returned to the
2579 sender, but the error was temporary, then after the final timeout, only
2580 "retry timeout exceeded" was returned. Now it returns the full error as
2581 well as "retry timeout exceeded".
2583 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2584 do this, and (what is worse) MTAs that accept it.
2586 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2587 will now be deprecated.
2589 PH/56 New os.c-cygwin from the Cygwin maintainer.
2591 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2592 formats (previously only available in exim formats via -bpr, -bpru,
2593 and -bpra. Now also available in native and exiqgrep formats)
2595 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2596 with very large, slow to parse queues
2598 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2600 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2602 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2603 responses to authentication challenges, though it was showing the
2604 challenges; (ii) I've removed the CR characters from the debug output for
2607 PH/58 Allow for the insertion of a newline as well as a space when a string
2608 is turned into more than one encoded-word during RFC 2047 encoding. The
2609 Sieve code now uses this.
2611 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2612 data_4xx, lost_connection, tls_required.
2614 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2615 message was being sent as an SMTP response.
2617 PH/61 Add -l and -k options to exicyclog.
2619 PH/62 When verifying, if an address was redirected to one new address, so that
2620 verification continued, and the new address failed or deferred after
2621 having set something in $address_data, the value of $address_data was not
2622 passed back to the ACL. This was different to the case when no
2623 redirection occurred. The value is now passed back in both cases.
2625 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2626 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2627 use login_cap.h, so on its own it isn't the distinguishing feature. The
2628 new name refers directly to the setclassresources() function.
2630 PH/65 Added configuration files for NetBSD3.
2632 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2634 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2635 is preferred over IPv4.
2637 PH/68 The bounce_return_message and bounce_return_body options were not being
2638 honoured for bounces generated during the reception of non-SMTP messages.
2639 In particular, this applied to messages rejected by the ACL. This bug has
2640 been fixed. However, if bounce_return_message is true and bounce_return_
2641 body is false, the headers that are returned for a non-SMTP message
2642 include only those that have been read before the error was detected.
2643 (In the case of an ACL rejection, they have all been read.)
2645 PH/69 The HTML version of the specification is now built in a directory called
2646 spec_html instead of spec.html, because the latter looks like a path with
2647 a MIME-type, and this confuses some software.
2649 PH/70 Catch two compiler warnings in sieve.c.
2651 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2652 function verify_get_ident() calls ip_connect() to connect a socket, but
2653 if the "connect()" function timed out, ip_connect() used to close the
2654 socket. However, verify_get_ident() also closes the socket later, and in
2655 between Exim writes to the log, which may get opened at this point. When
2656 the socket was closed in ip_connect(), the log could get the same file
2657 descriptor number as the socket. This naturally causes chaos. The fix is
2658 not to close the socket in ip_connect(); the socket should be closed by
2659 the function that creates it. There was only one place in the code where
2660 this was missing, in the iplookup router, which I don't think anybody now
2661 uses, but I've fixed it anyway.
2663 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2664 well as to direct DNS lookups. Otherwise the handling of names in host
2665 lists is inconsistent and therefore confusing.
2671 PH/01 Two changes to the default runtime configuration:
2673 (1) Move the checks for relay_from_hosts and authenticated clients from
2674 after to before the (commented out) DNS black list checks.
2676 (2) Add control=submission to the relay_from_hosts and authenticated
2677 clients checks, on the grounds that messages accepted by these
2678 statements are most likely to be submissions.
2680 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2682 (1) Generate an error if the third argument for the ${prvs expansion is
2685 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2688 (3) Reset the variables that are obtained from the first argument of
2689 ${prvscheck and used in the second argument before leaving the code,
2690 because their memory is reclaimed, so using them afterwards may do
2693 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2694 one (it's much easier than Tom thought :-).
2696 (5) Because of (4), we can now allow for the use of $prvscheck_result
2697 inside the third argument.
2699 PH/03 For some reason, the default setting of PATH when running a command from
2700 a pipe transport was just "/usr/bin". I have changed it to
2703 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2704 anything to be listed in the output from -bV.
2706 PH/05 When a filter generated an autoreply, the entire To: header line was
2707 quoted in the delivery log line, like this:
2709 => >A.N.Other <ano@some.domain> <original@ddress> ...
2711 This has been changed so that it extracts the operative address. There
2712 may be more than one such address. If so, they are comma-separated, like
2715 => >ano@some.domain,ona@other.domain <original@ddress> ...
2717 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2718 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2719 not being looked up in the rDNS to get a host name, Exim was showing the
2720 IP address twice in Received: lines, even though the IP addresses were
2721 identical. For example:
2723 Received: from [1.2.3.4] (helo=[1.2.3.4])
2725 However, if the real host name was known, it was omitting the HELO data
2726 if it matched the actual IP address. This has been tidied up so that it
2727 doesn't show the same IP address twice.
2729 PH/07 When both +timestamp and +memory debugging was on, the value given by
2730 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2731 called by the debug printing, thereby overwriting the timestamp buffer.
2732 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2734 PH/08 When the original message was included in an autoreply transport, it
2735 always said "this is a copy of the message, including all the headers",
2736 even if body_only or headers_only was set. It now gives an appropriate
2739 PH/09 Applied a patch from the Sieve maintainer which:
2741 o fixes some comments
2742 o adds the (disabled) notify extension core
2743 o adds some debug output for the result of if/elsif tests
2744 o points to the current vacation draft in the documentation
2745 and documents the missing references header update
2749 o fixes a bug in processing the envelope test (when testing
2750 multiple envelope elements, the last element determinted the
2753 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2754 Electronic Mail") by including:
2756 Auto-submitted: auto-generated
2758 in the messages that it generates (bounce messages and others, such as
2759 warnings). In the case of bounce messages for non-SMTP mesages, there was
2760 also a typo: it was using "Auto_submitted" (underscore instead of
2761 hyphen). Since every message generated by Exim is necessarily in response
2762 to another message, thes have all been changed to:
2764 Auto-Submitted: auto-replied
2766 in accordance with these statements in the RFC:
2768 The auto-replied keyword:
2770 - SHOULD be used on messages sent in direct response to another
2771 message by an automatic process,
2773 - MUST NOT be used on manually-generated messages,
2775 - MAY be used on Delivery Status Notifications (DSNs) and Message
2776 Disposition Notifications (MDNs),
2778 - MUST NOT be used on messages generated by automatic or periodic
2779 processes, except for messages which are automatic responses to
2782 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2783 to the default Received: header definition.
2785 PH/12 Added log selector acl_warn_skipped (default on).
2787 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2788 variables because (a) they are in the wrong storage pool and (b) even if
2789 they were copied, it wouldn't work properly because of the caching.
2791 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2792 checking when decoding. Apparently there are clients that generate
2793 overlong encoded strings. Why am I not surprised?
2795 PH/15 If the first argument of "${if match_address" was not empty, but did not
2796 contain an "@" character, Exim crashed. Now it writes a panic log message
2797 and treats the condition as false.
2799 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2801 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2802 extension "envelope-auth". The code is finished and in agreement with
2803 other implementations, but there is no documentation so far and in fact,
2804 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2805 not changing the active code.
2807 Print executed "if" and "elsif" statements when debugging is used. This
2808 helps a great deal to understand what a filter does.
2810 Document more things not specified clearly in RFC3028. I had all this
2811 sorted out, when out of a sudden new issues came to my mind. Oops."
2813 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2816 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2817 canonical form (as documented). However, after a host name lookup from
2818 the IP address, check_host() was doing a simple string comparison with
2819 addresses acquired from the DNS when checking that the found name did
2820 have the original IP as one of its addresses. Since any found IPv6
2821 addresses are likely to be in abbreviated form, the comparison could
2822 fail. Luckily, there already exists a function for doing the comparison
2823 by converting both addresses to binary, so now that is used instead of
2824 the text comparison.
2826 PH/20 There was another similar case to PH/19, when a complete host name was
2827 given in a host list; looking up its IP address could give an abbreviated
2828 form, whereas the current host's name might or might not be abbreviated.
2829 The same fix has been applied.
2835 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2836 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2839 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2840 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2842 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2844 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2845 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2846 Also, for "Auto-Submitted", treat anything other than "no" as
2847 non-personal, in accordance with RFC 3834. (Previously it treated
2848 anything starting "auto-" as non-personal.)
2850 TF/01 The control=submission/name=... option had a problem with syntax
2851 errors if the name included a slash character. The /name= option
2852 now slurps the rest of the string, so it can include any characters
2853 but it must come last in the list of options (after /sender_retain
2856 PH/05 Some modifications to the interface to the fake nameserver for the new
2864 TK/01 Added the "success_on_redirect" address verification option. See
2865 NewStuff for rationale and an example.
2867 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2869 PH/02 Patch to exigrep to allow it to work on syslog lines.
2871 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2872 fread() to skip over the body file's header line, because in Cygwin the
2873 header line is locked and is inaccessible.
2875 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2876 co-exist for some time) to make it clear that it is the Exim ID that is
2877 referenced, not the Message-ID: header line.
2879 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2880 string_format() function, because snprintf() does not exist on all
2883 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2884 ${stat: expansion item.
2886 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2887 protocol synchronization error", to keep the pedants happy.
2889 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2890 well as for IRIX systems, when gcc is being used. See the host.c source
2893 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2895 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2898 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2899 transport and to the smtp transport in LMTP mode.
2901 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2903 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2904 run a filter in a subprocess. This could lead to confusion in subsequent
2905 lookups in the parent process. There should also be a search_tidyup() at
2906 the end of the subprocess.
2908 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2909 only if the host matched helo_try_verify_hosts, which caused the
2910 verification to occur when the EHLO/HELO command was issued. The ACL just
2911 tested the remembered result. Now, if a previous verification attempt has
2912 not happened, "verify = helo" does it there and then.
2914 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2916 TK/03 Fix log output including CR from clamd.
2918 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2919 exist provoked a memory error which could cause a segfault.
2921 PH/15 Installed PCRE 6.2
2923 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2925 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2926 of the problem. Specifically, suggested +O2 rather than +O1 for the
2929 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2931 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2932 record, the port was not correctly logged when the outgoing_port log
2933 selector was set (it logged the transort's default port).
2935 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2936 fallback_hosts, and "hosts" in the smtp transport.
2938 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2939 host errors such as "Connection refused".
2941 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2942 authentication with radiusclient 0.4.9:
2944 - Error returned from rc_read_config was caught wrongly
2945 - Username/password not passed on to radius server due to wrong length.
2947 The presumption is that some radiusclient API changes for 4.51/PH/17
2948 were not taken care of correctly. The code is still untested by me (my
2949 Linux distribution still has 0.3.2 of radiusclient), but it was
2950 contributed by a Radius user.
2952 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2953 expanding the "port" option of the smtp transport.
2955 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2956 while reading a MIME header. Thanks to Tom Hughes for a patch.
2958 PH/24 Include config.h inside local_scan.h so that configuration settings are
2961 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2962 that in an ACL for RCPT (for example), you can examine exactly what was
2965 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2966 commands, but it was not correctly comparing the address with the actual
2967 client host address. Thus, it would show the EHLO address in Received:
2968 header lines when this was not necessary.
2970 PH/27 Added the % operator to ${eval:}.
2972 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2973 it should be ignoring failures (because with -C, for example, it has lost
2974 privilege). It wasn't ignoring creation failures other than "already
2977 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2980 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2981 because an input line was too long, either on its own, or by virtue of
2982 too many continuations, the temporary file was not being removed, and the
2983 return code was incorrect.
2985 PH/31 Missing "BOOL" in function definition in filtertest.c.
2987 PH/32 Applied Sieve patches from the maintainer.
2989 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2991 PH/33 Added "verify = not_blind".
2993 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2994 Local/Makefile (with some defaults set). These are used in built scripts
2995 such as exicyclog, but they have never been used in the exim_install
2996 script (though there are many overriding facilities there). I have
2997 arranged that the exim_install script now takes note of these two
3000 PH/35 Installed configuration files for Dragonfly.
3002 PH/36 When a locally submitted message by a trusted user did not contain a
3003 From: header, and the sender address was obtained from -f or from an SMTP
3004 MAIL command, and the trusted user did not use -F to supply a sender
3005 name, $originator_name was incorrectly used when constructing a From:
3006 header. Furthermore, $originator_name was used for submission mode
3007 messages from external hosts without From: headers in a similar way,
3008 which is clearly wrong.
3010 PH/37 Added control=suppress_local_fixups.
3012 PH/38 When log_selector = +received_sender was set, and the addition of the
3013 sender made the log line's construction buffer exactly full, or one byte
3014 less than full, an overflow happened when the terminating "\n" was
3017 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3018 when the result of a list match is failure because a DNS lookup failed.
3020 PH/40 RM_COMMAND is now used in the building process.
3022 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3023 the "build-* directories that it finds.
3025 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3026 address in a domain literal was a prefix of an interface address.
3028 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3029 when verifying a sender address, unless rewrite_headers is false.
3031 PH/44 Wrote a long comment about why errors_to addresses are verified as
3032 recipients, not senders.
3034 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3035 the ratelimit ACL was added.
3037 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3039 PH/46 Added extra information about PostgreSQL errors to the error string.
3041 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3042 avoiding the need to install special zones in a real server. This is
3043 backwards compatible; if it can't find the fake resolver, it drops back.
3044 Thus, both old and new test suites can be run.
3046 TF/02 Added util/ratelimit.pl
3048 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3049 clock is set back in time.
3051 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3052 Candler <B.Candler@pobox.com>.
3054 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3055 OK for addresses that are the result of redirections.
3057 PH/48 A number of further additions for the benefit of the new test suite,
3058 including a fake gethostbyname() that interfaces to the fake DNS resolver
3061 TF/06 The fix for widen_domains has also been applied to qualify_single and
3062 search_parents which are the other dnslookup options that can cause
3065 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3068 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3070 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3071 service" if it encountered an MX record with an empty target hostname.
3072 The message is now "an MX or SRV record indicated no SMTP service".
3074 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3075 if the DNS of the sending site is misconfigured. This is quite a
3076 common situation. This change restores the behaviour of treating a
3077 helo verification defer as a failure.
3079 PH/51 If self=fail was set on a router, the bounce message did not include the
3080 actual error message.
3086 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3088 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3089 command itself ended in error, the underlying message about the transport
3090 filter timeout was being overwritten with the pipe command error. Now the
3091 underlying error message should be appended to the second error message.
3093 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3095 PH/02 Reduce the amount of output that "make" produces by default. Full output
3096 can still be requested.
3098 PH/03 The warning log line about a condition test deferring for a "warn" verb
3099 was being output only once per connection, rather than after each
3100 occurrence (because it was using the same function as for successful
3101 "warn" verbs). This seems wrong, so I have changed it.
3103 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3104 it should not have, which might have caused a crash in the right
3105 circumstances, but probably never did.
3107 PH/04 Installed a modified version of Tony Finch's patch to make submission
3108 mode fix the return path as well as the Sender: header line, and to
3109 add a /name= option so that you can make the user's friendly name appear
3112 TF/03 Added the control = fakedefer ACL modifier.
3114 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3115 Mark Lowes for thorough testing.
3117 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3119 TK/03 Merged latest SRS patch from Miles Wilton.
3121 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3122 with the definition in sysexits.h (which is #included earlier).
3123 Fortunately, Exim does not actually use EX_OK. The code used to try to
3124 preserve the sysexits.h value, by assumimg that macro definitions were
3125 scanned for macro replacements. I have been disabused of this notion,
3126 so now the code just undefines EX_OK before #including unistd.h.
3128 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3129 in the smtp transport. When a block could not be written in a single
3130 write() function, the timeout was being re-applied to each part-write.
3131 This seems wrong - if the receiver was accepting one byte at a time it
3132 would take for ever. The timeout is now adjusted when this happens. It
3133 doesn't have to be particularly precise.
3135 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3136 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3138 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3139 without a domain if the check to <postmaster@domain> fails.
3141 SC/01 Eximstats: added -xls and the ability to specify output files
3142 (patch written by Frank Heydlauf).
3144 SC/02 Eximstats: use FileHandles for outputing results.
3146 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3148 SC/04 Eximstats: fixed display of large numbers with -nvr option
3150 SC/05 Eximstats: fixed merging of reports with empty tables.
3152 SC/06 Eximstats: added the -include_original_destination flag
3154 SC/07 Eximstats: removed tabs and trailing whitespace.
3156 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3158 TK/06 MBOX spool code: Add real "From " MBOX separator line
3159 so the .eml file is really in mbox format (even though
3160 most programs do not really care). Patch from Alex Miller.
3162 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3163 The latter is generated from $received_to and is only set if the
3164 message has one envelope recipient. SA can use these headers,
3165 obviously out-of-the-box. Patch from Alex Miller.
3167 PH/08 The ${def test on a variable was returning false if the variable's
3168 value was "0", contrary to what the specification has always said!
3169 The result should be true unless the variable is empty.
3171 PH/09 The syntax error of a character other than { following "${if
3172 def:variable_name" (after optional whitespace) was not being diagnosed.
3173 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3174 accidental colon was present, for example, could give incorrect results.
3176 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3177 result is used (not including appendfile, where other changes are about
3180 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3181 This involved changing a lot of size variables from int to off_t. It
3182 should work with maildirs and everything.
3184 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3185 spamd dying while we are connected to it.
3187 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3190 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3191 function for BDB 4.3.
3193 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3195 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3196 the sources for PCRE have been re-arranged and more files are now
3199 PH/15 The code I had for printing potentially long long variables in PH/11
3200 above was not the best (it lost precision). The length of off_t variables
3201 is now inspected at build time, and an appropriate printing format (%ld
3202 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3203 to be "long long int" or "long int". This is needed for the internal
3204 formatting function string_vformat().
3206 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3207 the configuration file to be ":syslog", then the script "guesses" where
3208 the logs files are, rather than using the compiled in default. In our
3209 case the guess is not the same as the compiled default, so the script
3210 suddenly stopped working when I started to use syslog. The patch checks
3211 to see if log_file_path is "". If so, it attempts to read it from exim
3212 with no configuration file to get the compiled in version, before it
3213 falls back to the previous guessing code."
3215 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3216 implementing BATV in an Exim configuration. See NewStuff for the gory
3219 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3220 Makefile that are specific to HP-UX.
3222 PH/18 If the "use_postmaster" option was set for a recipient callout together
3223 with the "random" option, the postmaster address was used as the MAIL
3224 FROM address for the random test, but not for the subsequent recipient
3225 test. It is now used for both.
3227 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3228 patch removes a few documentation additions to RFC 3028, because the
3229 latest draft now contains them. It adds the new en;ascii-case comparator
3230 and a new error check for 8bit text in MIME parts. Comparator and
3231 require names are now matched exactly. I enabled the subaddress
3232 extension, but it is not well tested yet (read: it works for me)."
3234 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3235 rework some of the code of TK/09 above to avoid the hardwired use of
3236 "%lld" and "long long". Replaced the call to snprintf() with a call to
3239 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3240 records point to non-existent hosts", "retry timeout exceeded", and
3241 "retry time not reached for any host after a long failure period".
3243 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3244 experimental DomainKeys support:
3246 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3247 (2) On an error such as an illegally used "control", the wrong name for
3248 the control was given.
3250 These problems did NOT occur unless DomainKeys support was compiled.
3252 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3254 PH/24 Added ${if match_ip condition.
3256 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3257 databases so that it will be absolutely obvious if a crash occurs in the
3258 DB library. This is a regular occurrence (often caused by mis-matched
3261 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3262 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3263 picked up on a user's system that detects such things. There doesn't seem
3264 to be a gcc warning option for this - only an attribute that has to be
3265 put on the function's prototype. It seems that in Fedora Core 4 they have
3266 set this on a number of new functions. No doubt there will be more in due
3269 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3270 specify a transport. However, if an address that was verified by such a
3271 router was the subject of a callout, Exim crashed because it tried to
3272 read the rcpt_include_affixes from the non-existent transport. Now it
3273 just assumes that the setting of that option is false. This bug was
3274 introduced by 4.51/PH/31.
3276 PH/28 Changed -d+all to exclude +memory, because that information is very
3277 rarely of interest, but it makes the output a lot bigger. People tend to
3278 do -d+all out of habit.
3280 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3281 code in os-type was giving problems when libc.so lives in lib64, like on
3284 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3285 aren't the modern standard, and it seems that some systems' include files
3286 don't always have them. Exim was already checking for some of the newer
3287 ones like T_AAAA, and defining it itself. I've added checks for all the
3288 record types that Exim uses.
3290 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3291 not automatically generating a new one, as it is supposed to. This
3292 prevented TLS from working. If the file did exist, but contained invalid
3293 data, a new version was generated, as expected. It was only the case of a
3294 non-existent file that was broken.
3296 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3297 with a change in libdomainkeys > 0.64.
3299 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3300 from DNS. If the selector record carries the flag, it now has
3301 precedence over the domain-wide flag.
3303 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3305 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3306 the use of a port name that isn't defined in /etc/services) occurred, the
3307 message was deferred as in a normal delivery, and thus remained on the
3308 spool, instead of being failed because of the mua_wrapper setting. This
3309 is now fixed, and I tidied up some of the mua_wrapper messages at the
3312 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3313 the messages in a hash of arrays rather than using individual hashes.
3314 This is a bit cleaner and results in dramatic memory savings, albeit
3315 at a slight CPU cost.
3317 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3318 as requested by Marc Sherman.
3320 SC/10 Eximstats: added histograms for user specified patterns as requested
3323 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3325 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3326 fopen() in the content-scanning modules that did not already have it.
3332 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3333 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3335 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3337 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3339 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3340 to test Sieve filters that use "vacation".
3342 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3343 that changes the way the GnuTLS parameters are stored in the cache file.
3344 The new format can be generated externally. For backward compatibility,
3345 if the data in the cache doesn't make sense, Exim assumes it has read an
3346 old-format file, and it generates new data and writes a new file. This
3347 means that you can't go back to an older release without removing the
3350 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3351 work if there are any delivery delays because "one_time" forces the
3352 parent to be marked "delivered", so its unseen clone is never tried
3353 again. For this reason, Exim now forbids the simultaneous setting of
3356 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3357 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3358 entry if you want to know the details. The fix, however, neglected to
3359 consider the case when local delivery batching is involved. The test for
3360 "previously delivered" was not happening when checking to see if an
3361 address could be batched with a previous (undelivered) one; under
3362 certain circumstances this could lead to multiple deliveries to the same
3365 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3366 in its include files, and this causes problems building Exim.
3368 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3369 header_syntax) but Exim was just ignoring anything given after a slash.
3370 In particular, this caused confusion with an attempt to use "verify =
3371 reverse_host_lookup/defer_ok". An error is now given when options are
3372 supplied for verify items that do not have them. (Maybe reverse_host_
3373 lookup should have a defer_ok option, but that's a different point.)
3375 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3376 defined by RFC 821) to 2048, because there were problems with some AUTH
3377 commands, and RFC 1869 says the size should be increased for extended
3378 SMTP commands that take arguments.
3380 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3383 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3384 "unknown" error; now it says that the functionality isn't in the binary.
3386 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3387 an address' error message when a string expansion fails (syntax or
3388 whatever). Otherwise the password may appear in the log. Following change
3389 PH/42 below, there is no longer a chance of it appearing in a bounce
3392 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3394 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3395 of its addresses was ever tried. (Bugzilla bug #2.)
3397 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3398 the result incorrectly in the debug output. (It correctly added a newline
3399 to what was transported.)
3401 TF/01 Added $received_time.
3403 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3404 commented out examples of how to interface to a virus scanner and to
3405 SpamAssassin. Also added commented examples of av_scanner and
3406 spamd_address settings.
3408 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3409 and controls are allowed in which ACLs. There were a couple of minor
3410 errors. Some of the entries in the conditions table (which is a table of
3411 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3412 negation of where the condition IS allowed.
3414 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3416 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3417 header file does not have a version number, so I've had to invent a new
3418 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3419 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3420 radiusclient), but it was contributed by a Radius user.
3422 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3423 files or directories, for OpenSSL.
3425 PH/19 When an Exim process that is running as root has to create an Exim log
3426 file, it does so in a subprocess that runs as exim:exim so as to get the
3427 ownership right at creation (otherwise, other Exim processes might see
3428 the file with the wrong ownership). There was no test for failure of this
3429 fork() call, which would lead to the process getting stuck as it waited
3430 for a non-existent subprocess. Forks do occasionally fail when resources
3431 run out. I reviewed all the other calls to fork(); they all seem to check
3434 PH/20 When checking for unexpected SMTP input at connect time (before writing
3435 the banner), Exim was not dealing correctly with a non-positive return
3436 from the read() function. If the client had disconnected by this time,
3437 the result was a log entry for a synchronization error with an empty
3438 string after "input=" when read() returned zero. If read() returned -1
3439 (an event I could not check), uninitialized data bytes were printed.
3440 There were reports of junk text (parts of files, etc) appearing after
3443 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3445 PH/22 Added support for macro redefinition, and (re)definition in between
3446 driver and ACL definitions.
3448 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3449 forgetting to use the resulting value; it was using the unexpanded value.
3451 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3452 hadn't been configured. The fix is from Juergen Kreileder, who
3453 understands it better than I do:
3455 "Here's what I see happening with three configured cyrus_sasl
3456 authenticators configured (plain, login, cram-md5):
3458 On startup auth_cyrus_sasl_init() gets called for each of these.
3459 This means three calls to sasl_listmech() without a specified mech_list.
3460 => SASL tests which mechs of all available mechs actually work
3461 => three warnings about OTP not working
3462 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3464 With the patch, sasl_listmech() also gets called three times. But now
3465 SASL's mech_list option is set to the server_mech specified in the the
3466 authenticator. Or in other words, the answer from sasl_listmech()
3467 gets limited to just the mech you're testing for (which is different
3469 => the return list contains just 'plain' or 'login', 'cram-md5' or
3470 nothing depending on the value of ob->server_mech.
3472 I've just tested the patch: Authentication still works fine,
3473 unavailable mechs specified in the exim configuration are still
3474 caught, and the auth.log warnings about OTP are gone."
3476 PH/25 When debugging is enabled, the contents of the command line are added
3477 to the debugging output, even when log_selector=+arguments is not
3480 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3481 answer is "GNU", and only if the return is "GNU/something" is the answer
3484 PH/27 $acl_verify_message is now set immediately after the failure of a
3485 verification in an ACL, and so is available in subsequent modifiers. In
3486 particular, the message can be preserved by coding like this:
3488 warn !verify = sender
3489 set acl_m0 = $acl_verify_message
3491 Previously, $acl_verify_message was set only while expanding "message"
3492 and "log_message" when a very denied access.
3494 PH/28 Modified OS/os.c-Linux with
3496 -#ifndef OS_LOAD_AVERAGE
3497 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3499 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3500 nomenclature these days.)
3502 PH/29 Installed patch from the Sieve maintainer that adds the options
3503 sieve_useraddress and sieve_subaddress to the redirect router.
3505 PH/30 In these circumstances:
3506 . Two addresses routed to the same list of hosts;
3507 . First host does not offer TLS;
3508 . First host accepts first address;
3509 . First host gives temporary error to second address;
3510 . Second host offers TLS and a TLS session is established;
3511 . Second host accepts second address.
3512 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3513 and peerdn, if requested) that were in fact used only for the second
3516 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3517 attention to any local part prefix or suffix that was matched by the
3518 router that accepted the address. It now behaves in the same way as it
3519 does for delivery: the affixes are removed from the local part unless
3520 rcpt_include_affixes is set on the transport.
3522 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3523 timeout during the DATA phase of an incoming message.
3525 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3526 applied a patch sanctioned by the Sieve maintainer.
3528 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3529 the uid or gid is negative. A case of a negative gid caused this to be
3530 noticed. The fix allows for either to be negative.
3532 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3533 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3536 PH/36 The change PH/12 above was broken. Fixed it.
3538 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3539 the grounds that routing the same address twice would always produce the
3540 same answer. This might have been true once, but it is certainly no
3541 longer true now. Routing a child address may depend on the previous
3542 routing that produced that child. Some complicated redirection strategies
3543 went wrong when messages had multiple recipients, and made Exim's
3544 behaviour dependent on the order in which the addresses were given.
3546 I have moved the duplicate checking until after the routing is complete.
3547 Exim scans the addresses that are assigned to local and remote
3548 transports, and removes any duplicates. This means that more work will be
3549 done, as duplicates will always all be routed, but duplicates are
3550 presumably rare, so I don't expect this is of any significance.
3552 For deliveries to pipes, files, and autoreplies, the duplicate checking
3553 still happens during the routing process, since they are not going to be
3556 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3557 It corrects a timeout issue with spamd. This is Ian's comment: "The
3558 background is that sometimes spamd either never reads data from a
3559 connection it has accepted, or it never writes response data. The exiscan
3560 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3561 blindly assumes that writes won't block so it may never time out."
3563 PH/39 Allow G after quota size as well as K and M.
3565 PH/40 The value set for $authenticated_id in an authenticator may not contain
3566 binary zeroes or newlines because the value is written to log lines and
3567 to spool files. There was no check on this. Now the value is run through
3568 the string_printing() function so that such characters are converted to
3569 printable escape sequences.
3571 PH/41 $message_linecount is a new variable that contains the total number of
3572 lines in the message. Compare $body_linecount, which is the count for the
3575 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3576 bounce and delay warning messages, except in certain special cases, which
3579 (a) An SMTP error message from a remote host;
3580 (b) A message specified in a :fail: redirection;
3581 (c) A message specified in a "fail" command in a system filter;
3582 (d) A message specified in a FAIL return from the queryprogram router;
3583 (e) A message specified by the cannot_route_message router option.
3585 In these cases only, Exim does include the error details in bounce and
3586 warning messages. There are also a few cases where bland messages such
3587 as "unrouteable address" or "local delivery error" are given.
3589 PH/43 $value is now also set for the "else" part of a ${run expansion.
3591 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3592 being worked on, but at least Exim now implements the latest version to
3595 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3596 process to complete was treated as a delivery failure, a timeout while
3597 writing the message to the pipe was logged, but erroneously treated as a
3598 successful delivery. Such timeouts include transport filter timeouts. For
3599 consistency with the overall process timeout, these timeouts are now
3600 treated as errors, giving rise to delivery failures by default. However,
3601 there is now a new Boolean option for the pipe transport called
3602 timeout_defer, which, if set TRUE, converts the failures into defers for
3603 both kinds of timeout. A transport filter timeout is now identified in
3606 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3607 systems where "make" and "gmake" are different, calling "gmake" at top
3608 level broke things. I've arranged for the value of $(MAKE) to be passed
3609 from the Makefile to this script so that it can call the same version of
3613 A note about Exim versions 4.44 and 4.50
3614 ----------------------------------------
3616 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3617 changes of various kinds. As a consequence, a big documentation update was
3618 needed. This delayed the release for rather longer than seemed good, especially
3619 in the light of a couple of (minor) security issues. Therefore, the changes
3620 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3621 release. So 4.44 and 4.50 are in effect two different branches that both start
3624 I have left the 4.50 change log unchanged; it contains all the changes since
3625 4.43. The change log for 4.44 is below; many of its items are identical to
3626 those for 4.50. This seems to be the most sensible way to preserve the
3627 historical information.
3633 1. Minor wording change to the doc/README.SIEVE file.
3635 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3636 computation of the current number of files was incorrect.
3638 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3639 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3642 4. Give more explanation in the error message when the command for a transport
3643 filter fails to execute.
3645 5. There are several places where Exim runs a non-Exim command in a
3646 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3647 was being done only for the command run by the queryprogram router. It is
3648 now done for all such subprocesses. The other cases are: ${run, transport
3649 filters, and the commands run by the lmtp and pipe transports.
3651 6. Added CONFIGURE_GROUP build-time option.
3653 7. Some older OS have a limit of 256 on the maximum number of file
3654 descriptors. Exim was using setrlimit() to set 1000 as a large value
3655 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3656 systems. I've change it so that if it can't get 1000, it tries for 256.
3658 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3659 was an oversight, and furthermore, ever since the addition of extra
3660 controls (e.g. 4.43/32), the checks on when to allow different forms of
3661 "control" were broken. There should now be diagnostics for all cases when a
3662 control that does not make sense is encountered.
3664 9. Added the /retain_sender option to "control=submission".
3666 10. $recipients is now available in the predata ACL (oversight).
3668 11. Tidy the search cache before the fork to do a delivery from a message
3669 received from the command line. Otherwise the child will trigger a lookup
3670 failure and thereby defer the delivery if it tries to use (for example) a
3671 cached ldap connection that the parent has called unbind on.
3673 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3674 of $address_data from the recipient verification was clobbered by the
3675 sender verification.
3677 13. The value of address_data from a sender verification is now available in
3678 $sender_address_data in subsequent conditions in the ACL statement.
3680 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3682 15. Added a new option "connect=<time>" to callout options, to set a different
3685 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3686 was its contents. (It was OK if the option was not defined at all.)
3688 17. A "Completed" log line is now written for messages that are removed from
3689 the spool by the -Mrm option.
3691 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3692 information about exactly what failed.
3694 19. Added -dd to debug only the daemon process.
3696 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3697 handles timeouts, both on the server side and network timeouts. Renamed the
3698 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3700 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3701 It is now set to "smtps".
3703 22. $host_address is now set to the target address during the checking of
3704 ignore_target_hosts.
3706 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3707 being passed; this would have caused $sender_host_name to have been used if
3708 matching the list had actually called for a host name (not very likely,
3709 since this list is usually IP addresses). A host name is now passed as
3712 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3713 code that specifies a non-restarting handler (typically sigaction() in
3714 modern systems) in an attempt to fix a rare and obscure crash bug.
3716 25. Narrowed the window for a race in the daemon that could cause it to ignore
3717 SIGCHLD signals. This is not a major problem, because they are used only to
3718 wake it up if nothing else does.
3720 26. A malformed maildirsize file could cause Exim to calculate negative values
3721 for the mailbox size or file count. Odd effects could occur as a result.
3722 The maildirsize information is now recalculated if the size or filecount
3725 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3726 support for a long time. Removed HAVE_SYS_VFS_H.
3728 28. Installed the latest version of exipick from John Jetmore.
3730 29. In an address list, if the pattern was not a regular expression, an empty
3731 subject address (from a bounce message) matched only if the pattern was an
3732 empty string. Non-empty patterns were not even tested. This was the wrong
3733 because it is perfectly reasonable to use an empty address as part of a
3734 database query. An empty address is now tested by patterns that are
3735 lookups. However, all the other forms of pattern expect the subject to
3736 contain a local part and a domain, and therefore, for them, an empty
3737 address still always fails if the pattern is not itself empty.
3739 30. Exim went into a mad DNS loop when attempting to do a callout where the
3740 host was specified on an smtp transport, and looking it up yielded more
3741 than one IP address.
3743 31. Re-factored the code for checking spool and log partition space into a
3744 function that finds that data and another that does the check. The former
3745 is then used to implement four new variables: $spool_space, $log_space,
3746 $spool_inodes, and $log_inodes.
3748 32. The RFC2047 encoding function was originally intended for short strings
3749 such as real names; it was not keeping to the 75-character limit for
3750 encoded words that the RFC imposes. It now respects the limit, and
3751 generates multiple encoded words if necessary. To be on the safe side, I
3752 have increased the buffer size for the ${rfc2047: expansion operator from
3755 33. It is now permitted to omit both strings after an "if" condition; if the
3756 condition is true, the result is "true". As before, when the second string
3757 is omitted, a false condition yields an empty string. This makes it less
3758 cumbersome to write custom ACL and router conditions.
3760 34. Failure to deliver a bounce message always caused it to be frozen, even if
3761 there was an errors_to setting on the router. The errors_to setting is now
3764 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3765 canonical form (fully expanded) before being placed in
3766 $sender_host_address.
3768 36. The table in the code that translates DNS record types into text (T_A to
3769 "A" for instance) was missing entries for NS and CNAME. It is just possible
3770 that this could have caused confusion if both these types were looked up
3771 for the same domain, because the text type is used as part of Exim's
3772 per-process caching. But the chance of anyone hitting this buglet seems
3775 37. The dnsdb lookup has been extended in a number of ways.
3777 (1) There is a new type, "zns", which walks up the domain tree until it
3778 finds some nameserver records. It should be used with care.
3780 (2) There is a new type, "mxh", which is like "mx" except that it returns
3781 just the host names, not the priorities.
3783 (3) It is now possible to give a list of domains (or IP addresses) to be
3784 looked up. The behaviour when one of the lookups defers can be
3785 controlled by a keyword.
3787 (4) It is now possible to specify the separator character for use when
3788 multiple records are returned.
3790 38. The dnslists ACL condition has been extended: it is now possible to supply
3791 a list of IP addresses and/or domains to be looked up in a particular DNS
3794 39. Added log_selector=+queue_time_overall.
3796 40. When running the queue in the test harness, wait just a tad after forking a
3797 delivery process, to get repeatability of debugging output.
3799 41. Include certificate and key file names in error message when GnuTLS fails
3800 to set them up, because the GnuTLS error message doesn't include the name
3801 of the failing file when there is a problem reading it.
3803 42. Allow both -bf and -bF in the same test run.
3805 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3807 44. The "Exiscan patch" is now merged into the mainline Exim source.
3809 45. Sometimes the final signoff response after QUIT could fail to get
3810 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3811 < 0 before doing a fflush(). This bug looks as though it goes back to the
3812 introduction of TLS in release 3.20, but "sometimes" must have been rare
3813 because the tests only now provoked it.
3815 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3816 (this can affect the format of dates).
3818 47. exim_tidydb, when checking for the continued existence of a message for
3819 which it has found a message-specific retry record, was not finding
3820 messages that were in split spool directories. Consequently, it was
3821 deleting retry records that should have stayed in existence.
3823 48. Steve fixed some bugs in eximstats.
3825 49. The SPA authentication driver was not abandoning authentication and moving
3826 on to the next authenticator when an expansion was forced to fail,
3827 contradicting the general specification for all authenticators. Instead it
3828 was generating a temporary error. It now behaves as specified.
3830 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3831 (the order specifies the preference for clients). The order is now AES256,
3832 AES128, 3DES, ARCFOUR128.
3834 51. Small patch to Sieve code - explicitly set From: when generating an
3837 52. Exim crashed if a remote delivery caused a very long error message to be
3838 recorded - for instance if somebody sent an entire SpamAssassin report back
3839 as a large number of 550 error lines. This bug was coincidentally fixed by
3840 increasing the size of one of Exim's internal buffers (big_buffer) that
3841 happened as part of the Exiscan merge. However, to be on the safe side, I
3842 have made the code more robust (and fixed the comments that describe what
3845 53. Now that there can be additional text after "Completed" in log lines (if
3846 the queue_time_overall log selector is set), a one-byte patch to exigrep
3847 was needed to allow it to recognize "Completed" as not the last thing in
3850 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3851 patch that reportedly fixes this has been added. I am not expert enough to
3852 create a test for it. This is what the patch creator wrote:
3854 "I found a little strange behaviour of ldap code when working with
3855 Windows 2003 AD Domain, where users was placed in more than one
3856 Organization Units. When I tried to give exim partial DN, the exit code
3857 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3858 But simultaneously result of request was absolutely normal ldap result,
3859 so I produce this patch..."
3861 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3862 REFERENCE, so I have modified the code to exclude the patch when that macro
3865 55. Some experimental protocols are using DNS PTR records for new purposes. The
3866 keys for these records are domain names, not reversed IP addresses. The
3867 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3868 leaves it alone. Component reversal etc. now happens only for IP addresses.
3871 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3873 57. Double the size of the debug message buffer (to 2048) so that more of very
3874 long debug lines gets shown.
3876 58. The exicyclog utility now does better if the number of log files to keep
3877 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3879 59. Two changes related to the smtp_active_hostname option:
3881 (1) $smtp_active_hostname is now available as a variable.
3882 (2) The default for smtp_banner uses $smtp_active_hostname instead
3883 of $primary_hostname.
3885 60. The host_aton() function is supposed to be passed a string that is known
3886 to be a valid IP address. However, in the case of IPv6 addresses, it was
3887 not checking this. This is a hostage to fortune. Exim now panics and dies
3888 if the condition is not met. A case was found where this could be provoked
3889 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3890 components; fortuitously, this particular loophole had already been fixed
3891 by change 4.50/55 above.
3893 If there are any other similar loopholes, the new check in host_aton()
3894 itself should stop them being exploited. The report I received stated that
3895 data on the command line could provoke the exploit when Exim was running as
3896 exim, but did not say which command line option was involved. All I could
3897 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3898 running as the user.
3901 61. There was a buffer overflow vulnerability in the SPA authentication code
3902 (which came originally from the Samba project). I have added a test to the
3903 spa_base64_to_bits() function which I hope fixes it.
3906 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3907 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3909 63. The daemon start-up calls getloadavg() while still root for those OS that
3910 need the first call to be done as root, but it missed one case: when
3911 deliver_queue_load_max is set with deliver_drop_privilege. This is
3912 necessary for the benefit of the queue runner, because there is no re-exec
3913 when deliver_drop_privilege is set.
3915 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3916 This has been fixed.
3918 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3919 was in use, was not putting the data itself into the right store pool;
3920 consequently, it could be overwritten for a subsequent message in the same
3921 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3924 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3926 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3927 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3928 the function were treating the return as a boolean value, which happened to
3929 work because 0=false and not-0=true, but is not correct code.
3931 68. The host_aton() function was not handling scoped IPv6 addresses (those
3932 with, for example, "%eth0" on the end) correctly.
3934 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3935 negated items (that is, ~something) in unsigned ints. Some compilers
3936 apparently mutter when there is no cast.
3938 70. If an address verification called from an ACL failed, and did not produce a
3939 user-specific message (i.e. there was only a "system" message), nothing was
3940 put in $acl_verify_message. In this situation, it now puts the system
3943 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3944 session; change 4.31/43 added the unwanted input to the log line - except
3945 that it did not do this in the start of session case. It now does.
3947 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3948 This is wrong and can cause the other end to generate a synchronization
3949 error if it is another Exim or anything else that does the synchronization
3950 check. A QUIT command is no longer sent after a timeout.
3952 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3953 during host lookups.
3955 74. The defer_ok option of callout verification was not working if it was used
3956 when verifying addresses in header lines, that is, for this case:
3958 verify = header_sender/callout=defer_ok
3960 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3961 those file descriptors could be used for SMTP connections. If anything
3962 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3963 could be sent to the SMTP client, causing chaos. The daemon now opens
3964 stdin, stdout, and stderr to /dev/null when it puts itself into the
3967 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3968 log by default. The user can override this with suitable Perl magic.
3970 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3971 the log message when discard triggers, was not working for the DATA ACL or
3972 for the non-SMTP ACL.
3974 78. Error message wording change in sieve.c.
3976 79. If smtp_accept_max_per_host was set, the number of connections could be
3977 restricted to fewer than expected, because the daemon was trying to set up
3978 a new connection before checking whether the processes handling previous
3979 connections had finished. The check for completed processes is now done
3980 earlier. On busy systems, this bug wouldn't be noticed because something
3981 else would have woken the daemon, and it would have reaped the completed
3984 80. If a message was submitted locally by a user whose login name contained one
3985 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3986 It caused a spool format error. I have fixed the spool reading code. A
3987 related problem was that the "from" clause in the Received: line became
3988 illegal because of the space(s). It is now covered by ${quote_local_part.
3990 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3993 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3995 83. Patch from Sieve maintainer to fix unterminated string problem in
3996 "vacation" handling.
3998 84. Some minor changes to the Linux configuration files to help with other
3999 OS variants using glibc.
4001 85. One more patch for Sieve to update vacation handling to latest spec.
4004 ----------------------------------------------------
4005 See the note above about the 4.44 and 4.50 releases.
4006 ----------------------------------------------------
4012 1. Change 4.43/35 introduced a bug that caused file counts to be
4013 incorrectly computed when quota_filecount was set in an appendfile
4016 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4017 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4020 3. Give more explanation in the error message when the command for a transport
4021 filter fails to execute.
4023 4. There are several places where Exim runs a non-Exim command in a
4024 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4025 was being done only for the command run by the queryprogram router. It is
4026 now done for all such subprocesses. The other cases are: ${run, transport
4027 filters, and the commands run by the lmtp and pipe transports.
4029 5. Some older OS have a limit of 256 on the maximum number of file
4030 descriptors. Exim was using setrlimit() to set 1000 as a large value
4031 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4032 systems. I've change it so that if it can't get 1000, it tries for 256.
4034 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4035 was an oversight, and furthermore, ever since the addition of extra
4036 controls (e.g. 4.43/32), the checks on when to allow different forms of
4037 "control" were broken. There should now be diagnostics for all cases when a
4038 control that does not make sense is encountered.
4040 7. $recipients is now available in the predata ACL (oversight).
4042 8. Tidy the search cache before the fork to do a delivery from a message
4043 received from the command line. Otherwise the child will trigger a lookup
4044 failure and thereby defer the delivery if it tries to use (for example) a
4045 cached ldap connection that the parent has called unbind on.
4047 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4048 of $address_data from the recipient verification was clobbered by the
4049 sender verification.
4051 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4052 was its contents. (It was OK if the option was not defined at all.)
4054 11. A "Completed" log line is now written for messages that are removed from
4055 the spool by the -Mrm option.
4057 12. $host_address is now set to the target address during the checking of
4058 ignore_target_hosts.
4060 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4061 being passed; this would have caused $sender_host_name to have been used if
4062 matching the list had actually called for a host name (not very likely,
4063 since this list is usually IP addresses). A host name is now passed as
4066 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4067 code that specifies a non-restarting handler (typically sigaction() in
4068 modern systems) in an attempt to fix a rare and obscure crash bug.
4070 15. Narrowed the window for a race in the daemon that could cause it to ignore
4071 SIGCHLD signals. This is not a major problem, because they are used only to
4072 wake it up if nothing else does.
4074 16. A malformed maildirsize file could cause Exim to calculate negative values
4075 for the mailbox size or file count. Odd effects could occur as a result.
4076 The maildirsize information is now recalculated if the size or filecount
4079 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4080 support for a long time. Removed HAVE_SYS_VFS_H.
4082 18. Updated exipick to current release from John Jetmore.
4084 19. Allow an empty sender to be matched against a lookup in an address list.
4085 Previously the only cases considered were a regular expression, or an
4088 20. Exim went into a mad DNS lookup loop when doing a callout where the
4089 host was specified on the transport, if the DNS lookup yielded more than
4092 21. The RFC2047 encoding function was originally intended for short strings
4093 such as real names; it was not keeping to the 75-character limit for
4094 encoded words that the RFC imposes. It now respects the limit, and
4095 generates multiple encoded words if necessary. To be on the safe side, I
4096 have increased the buffer size for the ${rfc2047: expansion operator from
4099 22. Failure to deliver a bounce message always caused it to be frozen, even if
4100 there was an errors_to setting on the router. The errors_to setting is now
4103 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4104 canonical form (fully expanded) before being placed in
4105 $sender_host_address.
4107 24. Updated eximstats to version 1.33
4109 25. Include certificate and key file names in error message when GnuTLS fails
4110 to set them up, because the GnuTLS error message doesn't include the name
4111 of the failing file when there is a problem reading it.
4113 26. Expand error message when OpenSSL has problems setting up cert/key files.
4116 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4117 (this can affect the format of dates).
4119 28. exim_tidydb, when checking for the continued existence of a message for
4120 which it has found a message-specific retry record, was not finding
4121 messages that were in split spool directories. Consequently, it was
4122 deleting retry records that should have stayed in existence.
4124 29. eximstats updated to version 1.35
4125 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4126 1.35 - bugfix such that pie charts by volume are generated correctly
4128 30. The SPA authentication driver was not abandoning authentication and moving
4129 on to the next authenticator when an expansion was forced to fail,
4130 contradicting the general specification for all authenticators. Instead it
4131 was generating a temporary error. It now behaves as specified.
4133 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4134 (the order specifies the preference for clients). The order is now AES256,
4135 AES128, 3DES, ARCFOUR128.
4137 31. Small patch to Sieve code - explicitly set From: when generating an
4140 32. Exim crashed if a remote delivery caused a very long error message to be
4141 recorded - for instance if somebody sent an entire SpamAssassin report back
4142 as a large number of 550 error lines. This bug was coincidentally fixed by
4143 increasing the size of one of Exim's internal buffers (big_buffer) that
4144 happened as part of the Exiscan merge. However, to be on the safe side, I
4145 have made the code more robust (and fixed the comments that describe what
4148 33. Some experimental protocols are using DNS PTR records for new purposes. The
4149 keys for these records are domain names, not reversed IP addresses. The
4150 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4151 leaves it alone. Component reversal etc. now happens only for IP addresses.
4154 34. The host_aton() function is supposed to be passed a string that is known
4155 to be a valid IP address. However, in the case of IPv6 addresses, it was
4156 not checking this. This is a hostage to fortune. Exim now panics and dies
4157 if the condition is not met. A case was found where this could be provoked
4158 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4159 components; fortuitously, this particular loophole had already been fixed
4160 by change 4.50/55 or 4.44/33 above.
4162 If there are any other similar loopholes, the new check in host_aton()
4163 itself should stop them being exploited. The report I received stated that
4164 data on the command line could provoke the exploit when Exim was running as
4165 exim, but did not say which command line option was involved. All I could
4166 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4167 running as the user.
4170 35. There was a buffer overflow vulnerability in the SPA authentication code
4171 (which came originally from the Samba project). I have added a test to the
4172 spa_base64_to_bits() function which I hope fixes it.
4175 36. The daemon start-up calls getloadavg() while still root for those OS that
4176 need the first call to be done as root, but it missed one case: when
4177 deliver_queue_load_max is set with deliver_drop_privilege. This is
4178 necessary for the benefit of the queue runner, because there is no re-exec
4179 when deliver_drop_privilege is set.
4181 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4182 was in use, was not putting the data itself into the right store pool;
4183 consequently, it could be overwritten for a subsequent message in the same
4184 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4187 38. Sometimes the final signoff response after QUIT could fail to get
4188 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4189 < 0 before doing a fflush(). This bug looks as though it goes back to the
4190 introduction of TLS in release 3.20, but "sometimes" must have been rare
4191 because the tests only now provoked it.
4197 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4198 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4199 Now it returns an integer. A number of calls were still expecting a T/F
4200 return. Fortuitously, in all cases, the tests worked in OK situations,
4201 which is the norm. However, things would have gone wrong on any write
4202 failures on the smtp file descriptor. This function is used when sending
4203 messages over SMTP and also when doing verify callouts.
4205 2. When Exim is called to do synchronous delivery of a locally submitted
4206 message (the -odf or -odi options), it no longer closes stderr before doing
4209 3. Implemented the mua_wrapper option.
4211 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4213 5. Implemented the functions header_remove(), header_testname(),
4214 header_add_at_position(), and receive_remove_recipient(), and exported them
4217 6. If an ACL "warn" statement specified the addition of headers, Exim already
4218 inserted X-ACL-Warn: at the start if there was no header name. However, it
4219 was not making this test for the second and subsequent header lines if
4220 there were newlines in the string. This meant that an invalid header could
4221 be inserted if Exim was badly configured.
4223 7. Allow an ACL "warn" statement to add header lines at the start or after all
4224 the Received: headers, as well as at the end.
4226 8. Added the rcpt_4xx retry error code.
4228 9. Added postmaster_mailfrom=xxx to callout verification option.
4230 10. Added mailfrom=xxxx to the callout verification option, for verify=
4233 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4234 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4236 12. Inserted some casts to stop certain compilers warning when using pointer
4237 differences as field lengths or precisions in printf-type calls (mostly
4238 affecting debugging statements).
4240 13. Added optional readline() support for -be (dynamically loaded).
4242 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4243 same clock tick as a message's arrival, so that its received time was the
4244 same as the "first fail" time on the retry record, and that message
4245 remained on the queue past the ultimate address timeout, every queue runner
4246 would try a delivery (because it was past the ultimate address timeout) but
4247 after another failure, the ultimate address timeout, which should have then
4248 bounced the address, did not kick in. This was a "< instead of <=" error;
4249 in most cases the first failure would have been in the next clock tick
4250 after the received time, and all would be well.
4252 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4253 being recognized when the domain list was tested by the match_domain
4254 condition in an expansion string.
4256 16. Added the ${str2b64: operator.
4258 17. Exim was always calling setrlimit() to set a large limit for the number of
4259 processes, without checking whether the existing limit was already
4260 adequate. (It did check for the limit on file descriptors.) Furthermore,
4261 errors from getrlimit() and setrlimit() were being ignored. Now they are
4262 logged to the main and panic logs, but Exim does carry on, to try to do its
4263 job under whatever limits there are.
4265 18. Imported PCRE 5.0.
4267 19. Trivial typo in log message " temporarily refused connection" (the leading
4270 20. If the log selector return_path_on_delivery was set and an address was
4271 redirected to /dev/null, the delivery process crashed because it assumed
4272 that a return path would always be set for a "successful" delivery. In this
4273 case, the whole delivery is bypassed as an optimization, and therefore no
4276 21. Internal re-arrangement: the function for sending a challenge and reading
4277 a response while authentication was assuming a zero-terminated challenge
4278 string. It's now changed to take a pointer and a length, to allow for
4279 binary data in such strings.
4281 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4283 23. Exim was not respecting finduser_retries when seeking the login of the
4284 uid under which it was called; it was always trying 10 times. (The default
4285 setting of finduser_retries is zero.) Also, it was sleeping after the final
4286 failure, which is pointless.
4288 24. Implemented tls_on_connect_ports.
4290 25. Implemented acl_smtp_predata.
4292 26. If the domain in control=submission is set empty, Exim assumes that the
4293 authenticated id is a complete email address when it generates From: or
4294 Sender: header lines.
4296 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4297 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4298 chown and chgrp in /bin and hostname in /usr/bin.
4300 28. Exim was keeping the "process log" file open after each use, just as it
4301 does for the main log. This opens the possibility of it remaining open for
4302 long periods when the USR1 signal hits a daemon. Occasional processlog
4303 errors were reported, that could have been caused by this. Anyway, it seems
4304 much more sensible not to leave this file open at all, so that is what now
4307 29. The long-running daemon process does not normally write to the log once it
4308 has entered its main loop, and it closes the log before doing so. This is
4309 so that log files can straightforwardly be renamed and moved. However,
4310 there are a couple of unusual error situations where the daemon does write
4311 log entries, and I had neglected to close the log afterwards.
4313 30. The text of an SMTP error response that was received during a remote
4314 delivery was being truncated at 512 bytes. This is too short for some of
4315 the long messages that one sometimes sees. I've increased the limit to
4318 31. It is now possible to make retry rules that apply only when a message has a
4319 specific sender, in particular, an empty sender.
4321 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4322 it possible to be selective about when SMTP synchronization is enforced.
4324 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4326 32. Implemented hosts_connection_nolog.
4328 33. Added an ACL for QUIT.
4330 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4333 35. Added mailbox_size and mailbox_filecount to appendfile.
4335 36. Added control = no_multiline_responses to ACLs.
4337 37. There was a bug in the logic of the code that waits for the clock to tick
4338 in the case where the clock went backwards by a substantial amount such
4339 that the microsecond fraction of "now" was more than the microsecond
4340 fraction of "then" (but the whole seconds number was less).
4342 38. Added support for the libradius Radius client library this is found on
4343 FreeBSD (previously only the radiusclient library was supported).
4349 1. When certain lookups returned multiple values in the form name=value, the
4350 quoting of the values was not always being done properly. Specifically:
4351 (a) If the value started with a double quote, but contained no whitespace,
4353 (b) If the value contained whitespace other than a space character (i.e.
4354 tabs or newlines or carriage returns) it was not quoted.
4355 This fix has been applied to the mysql and pgsql lookups by writing a
4356 separate quoting function and calling it from the lookup code. The fix
4357 should probably also be applied to nisplus, ibase and oracle lookups, but
4358 since I cannot test any of those, I have not disturbed their existing code.
4360 2. A hit in the callout cache for a specific address caused a log line with no
4361 reason for rejecting RCPT. Now it says "Previous (cached) callout
4362 verification failure".
4364 3. There was an off-by-one bug in the queryprogram router. An over-long
4365 return line was truncated at 256 instead of 255 characters, thereby
4366 overflowing its buffer with the terminating zero. As well as fixing this, I
4367 have increased the buffer size to 1024 (and made a note to document this).
4369 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4370 when Exim is waiting for an SMTP response from a remote server, Exim
4371 restarts its select() call on the socket, thereby resetting its timeout.
4372 This is not a problem when such interrupts are rare. Somebody set up a cron
4373 job to run exiwhat every 2 minutes, which is less than the normal select()
4374 timeout (5 or 10 minutes). This meant that the select() timeout never
4375 kicked in because it was always reset. I have fixed this by comparing the
4376 time when an interrupt arrives with the time at the start of the first call
4377 to select(). If more time than the timeout has elapsed, the interrupt is
4378 treated as a timeout.
4380 5. Some internal re-factoring in preparation for the addition of Sieve
4381 extensions (by MH). In particular, the "personal" test is moved to a
4382 separate function, and given an option for scanning Cc: and Bcc: (which is
4383 not set for Exim filters).
4385 6. When Exim created an email address using the login of the caller as the
4386 local part (e.g. when creating a From: or Sender: header line), it was not
4387 quoting the local part when it contained special characters such as @.
4389 7. Installed new OpenBSD configuration files.
4391 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4392 try to make them clearer.
4394 9. Callout options, other than the timeout value, were being ignored when
4395 verifying sender addresses in header lines. For example, when using
4397 verify = header_sender/callout=no_cache
4399 the cache was (incorrectly) being used.
4401 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4402 only the Cygwin environment.
4404 11. When return_path_on_delivery was set as a log selector, if different remote
4405 addresses in the same message used different return paths and parallel
4406 remote delivery occurred, the wrong values would sometimes be logged.
4407 (Whenever a remote delivery process finished, the return path value from
4408 the most recently started remote delivery process was logged.)
4410 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4411 lines when AUTH and/or TLS are in use. This is the "received protocol"
4412 field. Exim used to use "asmtp" for authenticated SMTP, without any
4413 indication (in the protocol name) for TLS use. Now it follows the RFC and
4414 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4415 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4416 names appear in log lines as well as in Received: header lines.
4418 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4419 extensions, and comparison tests, and to fix some bugs.
4421 14. Changes to the "personal" filter test:
4423 (1) The test was buggy in that it was just doing the equivalent of
4424 "contains" tests on header lines. For example, if a user's address was
4425 anne@some.where, the "personal" test would incorrectly be true for
4427 To: susanne@some.where
4429 This test is now done by extracting each address from the header in turn,
4430 and checking the entire address. Other tests that are part of "personal"
4431 are now done using regular expressions (for example, to check local parts
4432 of addresses in From: header lines).
4434 (2) The list of non-personal local parts in From: addresses has been
4435 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4436 taken from the Sieve specification recommendations.
4438 (3) If the message contains any header line starting with "List-" it is
4439 treated as non-personal.
4441 (4) The test for "circular" in the Subject: header line has been removed
4442 because it now seems ill-conceived.
4444 15. Minor typos in src/EDITME comments corrected.
4446 16. Installed latest exipick from John Jetmore.
4448 17. If headers_add on a router specified a text string that was too long for
4449 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4450 of string_sprintf() is now avoided.
4452 18. $message_body_size was not set (it was always zero) when running the DATA
4453 ACL and the local_scan() function.
4455 19. For the "mail" command in an Exim filter, no default was being set for
4456 the once_repeat time, causing a random time value to be used if "once" was
4457 specified. (If the value happened to be <= 0, no repeat happened.) The
4458 default is now 0s, meaning "never repeat". The "vacation" command was OK
4459 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4460 (I found it when inspecting the code).
4462 20. There is now an overall timeout for performing a callout verification. It
4463 defaults to 4 times the callout timeout, which applies to individual SMTP
4464 commands during the callout. The overall timeout applies when there is more
4465 than one host that can be tried. The timeout is checked before trying the
4466 next host. This prevents very long delays if there are a large number of
4467 hosts and all are timing out (e.g. when the network connections are timing
4468 out). The value of the overall timeout can be changed by specifying an
4469 additional sub-option for "callout", called "maxwait". For example:
4471 verify = sender/callout=5s,maxwait=20s
4473 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4474 to the end before writing, but this should make it even safer).
4476 22. Exim was forgetting that it had advertised PIPELINING for the second and
4477 subsequent messages on an SMTP connection. It was also not resetting its
4478 memory on STARTTLS and an internal HELO.
4480 23. When Exim logs an SMTP synchronization error within a session, it now
4481 records whether PIPELINING has been advertised or not.
4483 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4484 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4485 rather than long int.
4487 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4489 26. Added the never_mail option to autoreply.
4495 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4496 crash if the getsockname() call failed; this can happen if a connection is
4497 closed very soon after it is established. The problem was simply in the
4498 order in which certain operations were done, causing Exim to try to write
4499 to the SMTP stream before it had set up the file descriptor. The bug has
4500 been fixed by making things happen in the correct order.
4506 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4507 before the connection was closed, thus losing the rejection response.
4509 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4510 some early Solaris releases, but causes trouble in current releases where
4511 socklen_t is defined.
4513 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4516 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4519 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4520 that releasing the top of it at the end releases what was used for sub-
4521 expansions (unless the block got too big). However, discard this block if
4522 the first thing is a variable or header, so that we can use its block when
4523 it is dynamic (useful for very large $message_headers, for example).
4525 6. Lookups now cache *every* query, not just the most recent. A new, separate
4526 store pool is used for this. It can be recovered when all lookup caches are
4527 flushed. Lookups now release memory at the end of their result strings.
4528 This has involved some general refactoring of the lookup sources.
4530 7. Some code has been added to the store_xxx() functions to reduce the amount
4531 of flapping under certain conditions.
4533 8. log_incoming_interface used to affect only the <= reception log lines. Now
4534 it causes the local interface and port to be added to several more SMTP log
4535 lines, for example "SMTP connection from", and rejection lines.
4537 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4539 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4541 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4542 could be overwritten at the end of the current message (or the start of a
4543 new message if it was set in a HELO ACL). The value is now preserved for
4544 the duration of the SMTP connection.
4546 12. If a transport had a headers_rewrite setting, and a matching header line
4547 contained an unqualified address, that address was qualified, even if it
4548 did not match any rewriting rules. The underlying bug was that the values
4549 of the flags that permit the existence of unqualified sender and recipient
4550 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4551 non-local messages, and by -bnq for local messages) were not being
4552 preserved with the message after it was received.
4554 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4555 "next input=" as part of the text comprising the host identity instead of
4556 the correct text. The code was using the same buffer for two different
4557 strings. However, depending on which order the printing function evaluated
4558 its arguments, the bug did not always show up. Under Linux, for example, my
4559 test suite worked just fine.
4561 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4562 doesn't work with some older versions of Perl. It has been changed to "my",
4563 which in any case is probably the better facility to use.
4565 15. A really picky compiler found some instances of statements for creating
4566 error messages that either had too many or two few arguments for the format
4569 16. The size of the buffer for calls to the DNS resolver has been increased
4570 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4571 for addresses that have a lot of PTR records. This alleviates a problem; it
4572 does not fully solve it.
4574 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4575 buffer now truncates the list and logs the incident, which is the same
4576 action as happens when Exim is looking up a host name and its aliases.
4577 Previously in this situation something unpredictable would happen;
4578 sometimes it was "internal error: store_reset failed".
4580 18. If a server dropped the connection unexpectedly when an Exim client was
4581 using GnuTLS and trying to read a response, the client delivery process
4582 crashed while trying to generate an error log message.
4584 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4585 string, for example:
4587 warn message = H1: something\nH2: something
4589 the text was added as a single header line from Exim's point of view
4590 though it ended up OK in the delivered message. However, searching for the
4591 second and subsequent header lines using $h_h2: did not work. This has been
4592 fixed. Similarly, if a system filter added multiple headers in this way,
4593 the routers could not see them.
4595 20. Expanded the error message when iplsearch is called with an invalid key to
4596 suggest using net-iplsearch in a host list.
4598 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4599 ACLs are no longer actually imposed (and a message to that effect is
4602 22. If a "gecos" field in a passwd entry contained escaped characters, in
4603 particular, if it contained a \" sequence, Exim got it wrong when building
4604 a From: or a Sender: header from that name. A second bug also caused
4605 incorrect handling when an unquoted " was present following a character
4606 that needed quoting.
4608 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4609 was not being matched caselessly.
4611 24. Arranged for all hyphens in the exim.8 source to be escaped with
4614 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4615 properly. Recipient callout cache records were still being keyed to include
4616 the sender, even when use_sender was set false. This led to far more
4617 callouts that were necessary. The sender is no longer included in the key
4618 when use_sender is false.
4620 26. Added "control = submission" modifier to ACLs.
4622 27. Added the ${base62d: operator to decode base 62 numbers.
4624 28. dnsdb lookups can now access SRV records.
4626 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4627 the configuration file.
4629 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4630 (-v) mode. This makes the output for a verbose queue run more intelligible.
4632 31. Added a use_postmaster feature to recipient callouts.
4634 32. Added the $body_zerocount variable, containing the number of binary zero
4635 bytes in the message body.
4637 33. The time of last modification of the "new" subdirectory is now used as the
4638 "mailbox time last read" when there is a quota error for a maildir
4641 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4643 35. Added +ignore_unknown as a special item in host lists.
4645 36. Code for decoding IPv6 addresses in host lists is now included, even if
4646 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4647 address was recognized as an IP address, but was then not correctly decoded
4648 into binary, causing unexpected and incorrect effects when compared with
4655 1. Very minor rewording of debugging text in manualroute to say "list of
4656 hosts" instead of "hostlist".
4658 2. If verify=header_syntax was set, and a header line with an unqualified
4659 address (no domain) and a large number of spaces between the end of the
4660 name and the colon was received, the reception process suffered a buffer
4661 overflow, and (when I tested it) crashed. This was caused by some obsolete
4662 code that should have been removed. The fix is to remove it!
4664 3. When running in the test harness, delay a bit after writing a bounce
4665 message to get a bit more predictability in the log output.
4667 4. Added a call to search_tidyup() just before forking a reception process. In
4668 theory, someone could use a lookup in the expansion of smtp_accept_max_
4669 per_host which, without the tidyup, could leave open a database connection.
4671 5. Added the variables $recipient_data and $sender_data which get set from a
4672 lookup success in an ACL "recipients" or "senders" condition, or a router
4673 "senders" option, similar to $domain_data and $local_part_data.
4675 6. Moved the writing of debug_print from before to after the "senders" test
4678 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4679 problems for message scanning, either using a data ACL, or using
4680 local_scan() because the Received: header was not generated till after they
4681 were called (in order to set the time as the time of reception completion).
4682 I have revised the way this works. The header is now generated after the
4683 body is received, but before the ACL or local_scan() are called. After they
4684 are run, the timestamp in the header is updated.
4690 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4691 before starting a queue runner without re-exec. This happened only when
4692 deliver_drop_privilege was set or when the Exim user was set to root. The
4693 effect of the bug was that timeouts during subsequent deliveries caused
4694 crashes instead of being properly handled. The handler is now left at its
4695 default (and expected) setting.
4697 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4698 message, again when deliver_drop_privilege is set or Exim is run as root.
4699 The bug described in (1) was not present in this case, but the tidying up
4700 of the other signals was missing. I have made the two cases consistent.
4702 3. The ignore_target_hosts setting on a manualroute router was being ignored
4703 for hosts that were looked up using the /MX notation.
4705 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4708 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4709 operated on the sender address. After changing the $sender_address to <>
4710 for the sender address verify, Exim was re-instated it as the original
4711 (before rewriting) address, but remembering that it had rewritten it, so it
4712 wasn't rewriting it again. This bug also had the effect of breaking the
4713 sender address verification caching when the sender address was rewritten.
4715 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4716 This has been changed so that if the ip literal address matches
4717 ignore_target_hosts, the router declines.
4719 7. Added expansion conditions match_domain, match_address, and match_local_
4720 part (NOT match_host).
4722 8. The placeholder for the Received: header didn't have a length field set.
4724 9. Added code to Exim itself and to exim_lock to test for a specific race
4725 condition that could lead to file corruption when using MBX delivery. The
4726 issue is with the lockfile that is created in /tmp. If this file is removed
4727 after a process has opened it but before that process has acquired a lock,
4728 there is the potential for a second process to recreate the file and also
4729 acquire a lock. This could lead to two Exim processes writing to the file
4730 at the same time. The added code performs the same test as UW imapd; it
4731 checks after acquiring the lock that its file descriptor still refers to
4732 the same named file.
4734 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4735 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4738 11. Added the smtp_active_hostname option. If used, this will typically be made
4739 to depend on the incoming interface address. Because $interface_address is
4740 not set up until the daemon has forked a reception process, error responses
4741 that can happen earlier (such as "too many connections") no longer contain
4744 12. If an expansion in a condition on a "warn" statement fails because a lookup
4745 defers, the "warn" statement is abandoned, and the next ACL statement is
4746 processed. Previously this caused the whole ACL to be aborted.
4748 13. Added the iplsearch lookup type.
4750 14. Added ident_timeout as a log selector.
4752 15. Added tls_certificate_verified as a log selector.
4754 16. Added a global option tls_require_ciphers (compare the smtp transport
4755 option of the same name). This controls incoming TLS connections.
4757 17. I finally figured out how to make tls_require_ciphers do a similar thing
4758 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4759 before starting the TLS session.
4761 18. Tabs are now shown as \t in -bP output.
4763 19. If the log selector return_path_on_delivery was set, Exim crashed when
4764 bouncing a message because it had too many Received: header lines.
4766 20. If two routers both had headers_remove settings, and the first one included
4767 a superfluous trailing colon, the final name in the first list and the
4768 first name in the second list were incorrectly joined into one item (with a
4769 colon in the middle).
4775 1. Added -C and -D options to the exinext utility, mainly to make it easier
4776 to include in the automated testing, but these could be helpful when
4777 multiple configurations are in use.
4779 2. The exinext utility was not formatting the output nicely when there was
4780 an alternate port involved in the retry record key, nor when there was a
4781 message id as well (for retries that were specific to a specific message
4782 and a specific host). It was also confused by IPv6 addresses, because of
4783 the additional colons they contain. I have fixed the IPv4 problem, and
4784 patched it up to do a reasonable job for IPv6.
4786 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4787 delivery, the log line now contains "pipelined" if PIPELINING was used.
4789 4. An SMTP transport process used to panic and die if the bind() call to set
4790 an explicit outgoing interface failed. This has been changed; it is now
4791 treated in the same way as a connect() failure.
4793 5. A reference to $sender_host_name in the part of a conditional expansion
4794 that was being skipped was still causing a DNS lookup. This no longer
4797 6. The def: expansion condition was not recognizing references to header lines
4798 that used bh_ and bheader_.
4800 7. Added the _cache feature to named lists.
4802 8. The code for checking quota_filecount in the appendfile transport was
4803 allowing one more file than it should have been.
4805 9. For compatibility with Sendmail, the command line option
4813 and sets the incoming protocol and host name (for trusted callers). The
4814 host name and its colon can be omitted when only the protocol is to be set.
4815 Note the Exim already has two private options, -pd and -ps, that refer to
4816 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4817 "s", but I don't think that's a major issue.
4819 10. A number of refactoring changes to the code, none of which should affect
4822 (a) The number of logging options was getting close to filling up the
4823 32-bit word that was used as a bit map. I have split them into two classes:
4824 those that are passed in the argument to log_write(), and those that are
4825 only ever tested independently outside of that function. These are now in
4826 separate 32-bit words, so there is plenty of room for expansion again.
4827 There is no change in the user interface or the logging behaviour.
4829 (b) When building, for example, log lines, the code previously used a
4830 macro that called string_cat() twice, in order to add two strings. This is
4831 not really sufficiently general. Furthermore, there was one instance where
4832 it was actually wrong because one of the argument was used twice, and in
4833 one call a function was used. (As it happened, calling the function twice
4834 did not affect the overall behaviour.) The macro has been replaced by a
4835 function that can join an arbitrary number of extra strings onto a growing
4838 (c) The code for expansion conditions now uses a table and a binary chop
4839 instead of a serial search (which was left over from when there were very
4840 few conditions). Also, it now recognizes conditions like "pam" even when
4841 the relevant support is not compiled in: a suitably worded error message is
4842 given if an attempt is made to use such a condition.
4844 11. Added ${time_interval:xxxxx}.
4846 12. A bug was causing one of the ddress fields not to be passed back correctly
4847 from remote delivery subprocesses. The field in question was not being
4848 subsequently used, so this caused to problems in practice.
4850 13. Added new log selectors queue_time and deliver_time.
4852 14. Might have fixed a bug in maildirsizefile handling that threw up
4853 "unexpected character" debug warnings, and recalculated the data
4854 unnecessarily. In any case, I expanded the warning message to give more
4857 15. Added the message "Restricted characters in address" to the statements in
4858 the default ACL that block characters like @ and % in local parts.
4860 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4861 Three changes have been made:
4863 (a) There was a serious bug; a negative response to MAIL caused the whole
4864 recipient domain to be cached as invalid, thereby blocking all messages
4865 to all local parts at the same domain, from all senders. This bug has
4866 been fixed. The domain is no longer cached after a negative response to
4867 MAIL if the sender used is not empty.
4869 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4872 (c) A new callout option, "use_sender" has been added for people who want
4873 the modified behaviour.
4879 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4882 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4883 indeed breaks things for older releases.
4885 3. Added additional logging to the case where there is a problem reading data
4886 from a filter that is running in a subprocess using a pipe, in order to
4887 try to track down a specific problem.
4889 4. Testing facility fudge: when running in the test harness and attempting
4890 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4891 getting "No route to host". Convert this to a timeout.
4893 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4896 6. Some OS don't have socklen_t but use size_t instead. This affects the
4897 fifth argument of getsockopt() amongst other things. This is now
4898 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4899 can be set for individual OS. I have set it for SunOS5, OSF1, and
4900 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4901 some earlier ones do not.
4903 7. Change 4.30/15 was not doing the test caselessly.
4905 8. The standard form for an IPv6 address literal was being rejected by address
4906 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4907 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4908 this, as well as the form without the "IPv6" on the front (but only when
4909 address literals are enabled, of course).
4911 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4913 10. Exim crashed if a message with an empty sender address specified by -f
4914 encountered a router with an errors_to setting. This could be provoked only
4915 by a command such as
4919 where an empty string was supplied; "<>" did not hit this bug.
4921 11. Installed PCRE release 4.5.
4923 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4924 remained set. It is now erased.
4926 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4927 times from message ids (which are base 36 rather than the normal 62).
4929 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4930 were being counted as actual protocol errors, and logged if the log
4931 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4932 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4933 and DATA following a set of rejected RCPTs do not count as protocol errors.
4934 In other words, Exim assumes they were pipelined, though this may not
4935 actually be the case. Of course, in all cases the client gets an
4936 appropriate error code.
4938 15. If a lookup fails in an ACL condition, a message about the failure may
4939 be available; it is used if testing the ACL cannot continue, because most
4940 such messages specify what the cause of the deferral is. However, some
4941 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4942 that caused an old message to be retained and used if a later statement
4943 caused a defer, replacing the real cause of the deferral.
4945 16. If an IP address had so many PTR records that the DNS lookup buffer
4946 was not large enough to hold them, Exim could crash while trying to process
4947 the truncated data. It now detects and logs this case.
4949 17. Further to 4.21/58, another change has been made: if (and only if) the
4950 first line of a message (the first header line) ends with CRLF, a bare LF
4951 in a subsequent header line has a space inserted after it, so as not to
4952 terminate the header.
4954 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4955 unnecessarily, used atoi() instead of strtol(), and didn't check the
4956 termination when getting file sizes from file names by regex.
4958 19. Completely re-implemented the support for maildirsize files, in the light
4959 of a number of problems with the previous contributed implementation
4960 (4.30/29). In particular:
4962 . If the quota is zero, the maildirsize file is maintained, but no quota is
4965 . If the maildir directory does not exist, it is created before any attempt
4966 to write a maildirsize file.
4968 . The quota value in the file is just a cache; if the quota is changed in
4969 the transport, the new value overrides.
4971 . A regular expression is available for excluding directories from the
4974 20. The autoreply transport checks the characters in options that define the
4975 message's headers; it allows continued headers, but it was checking with
4976 isspace() after an embedded newline instead of explicitly looking for a
4979 21. If all the "regular" hosts to which an address was routed had passed their
4980 expiry times, and had not reached their retry times, the address was
4981 bounced, even if fallback hosts were defined. Now Exim should go on to try
4984 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4985 equivalent code in the SMTP transport. Some hosts send humungous responses
4986 to HELO/EHLO, more than 1024 it seems.
4988 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4989 gives compiler warnings in some environments. I've now done it "properly",
4992 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4993 (because of problems with the built-in one) was declared to return uschar *
4994 instead of char *, causing compiler failure.
4996 25. Fixed a file descriptor leak when processing alias/forward files.
4998 26. Fixed a minor format string issue in dbfn.c.
5000 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5002 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5003 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5006 29. When a sender address is verified, it is cached, to save repeating the test
5007 when there is more than one recipient in a message. However, when the
5008 verification involves a callout, it is possible for different callout
5009 options to be set for different recipients. It is too complicated to keep
5010 track of this in the cache, so now Exim always runs a verification when a
5011 callout is required, relying on the callout cache for the optimization.
5012 The overhead is duplication of the address routing, but this should not be
5015 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5016 to be verified with callout=postmaster, and the main callout worked but the
5017 postmaster check failed, the verification correctly failed. However, if a
5018 subsequent RCPT command asked for sender verification *without* the
5019 postmaster check, incorrect caching caused this verification also to fail,
5022 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5023 it was not caching the DNS options (qualify_single, search_parents) that
5024 were used when the lookup failed. A subsequent lookup with different
5025 options therefore always gave the same answer, though there were cases
5026 where it should not have. (Example: a "domains = !$mx_any" option on a
5027 dnslookup router: the "domains" option is always processed without any
5028 widening, but the router might have qualify_single set.) Now Exim uses the
5029 cached value only when the same options are set.
5031 32. Added John Jetmore's "exipick" utility to the distribution.
5033 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5034 than a timeout (e.g. a certificate is required, and is not provided), an
5035 Exim server now closes the connection immediately. Previously it waited for
5036 the client to close - but if the client is SSL, it seems that they each
5037 wait for each other, leading to a delay before one of them times out.
5039 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5040 maintained 0.8.x compatibility because I don't think many are using it, and
5041 it is clearly obsolete.
5043 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5046 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5047 client certificate was expired. A simple patch fixes this, though I don't
5048 understand the full logic of why the verify callback is called multiple
5051 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5052 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5053 which causes problems with some clients (such as the Certicom SSL Plus
5054 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5055 disables the coutermeasure allowing Eudora to connect."
5057 38. Exim was not checking that a write() to a log file succeeded. This could
5058 lead to Bad Things if a log got too big, in particular if it hit a file
5059 size limit. Exim now panics and dies if it cannot write to a log file, just
5060 as it does if it cannot open a log file.
5062 39. Modified OS/Makefile-Linux so that it now contains
5064 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5066 The two -D definitions ensure that Exim is compiled with large file
5067 support, which makes it possible to handle log files that are bigger than
5070 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5071 instance) a domain was checked against a named list that involved a lookup,
5072 causing $domain_data to be set, then another domain was checked against the
5073 same list, then the first domain was re-checked, the value of $domain_data
5074 after the final check could be wrong. In particular, if the second check
5075 failed, it could be set empty. This bug probably also applied to
5078 41. The strip_trailing_dot option was not being applied to the address given
5079 with the -f command-line option.
5081 42. The code for reading a message's header from the spool was incrementing
5082 $received_count, but never initializing it. This meant that the value was
5083 incorrect (doubled) while delivering a message in the same process in which
5084 it was received. In the most common configuration of Exim, this never
5085 happens - a fresh exec is done - but it can happen when
5086 deliver_drop_privilege is set.
5088 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5089 it now includes up to 150 characters of the unexpected data in the log
5092 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5093 and building data strings. The size of both of these buffers was 10 000
5094 bytes - far larger than anybody would *ever* want, thought I. Needless to
5095 say, somebody hit the limit. I have increased the maximum line length to
5096 20 000 and the maximum data length of concatenated lines to 100 000. I have
5097 also fixed two bugs, because there was no checking on these buffers. Tsk,
5098 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5099 buffer is too small.
5101 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5102 lsearch lookups. Now it does.
5104 46. When parsing a route_list item in a manualroute router, a fixed-length
5105 buffer was used for the list of hosts. I made this 1024 bytes long,
5106 thinking that nobody would ever have a list of hosts that long. Wrong.
5107 Somebody had a whole pile of complicated expansion conditions, and the
5108 string was silently truncated, leading to an expansion error. It turns out
5109 that it is easier to change to an unlimited length (owing to other changes
5110 that have happened since this code was originally written) than to build
5111 structure for giving a limitation error. The length of the item that
5112 expands into the list of hosts is now unlimited.
5114 47. The lsearch lookup could not handle data where the length of text line was
5115 more than 4095 characters. Such lines were truncated, leading to shortened
5116 data being returned. It should now handle lines of any length.
5118 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5119 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5122 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5123 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5124 no longer happens in comment lines. A list of replacements is now placed
5125 at the head of all of the source files, except those whose only change is
5126 to replace PERL_COMMAND in the very first #! line.
5128 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5129 messages on the queue, with a bottom-up merge sort, using code contributed
5130 by Michael Haardt. This should make operations like -bp somewhat faster on
5131 large queues. It won't affect queue runners, except when queue_run_in_order
5134 51. Installed eximstats 1.31 in the distribution.
5136 52. Added support for SRV lookups to the dnslookup router.
5138 53. If an ACL referred to $message_body or $message_body_end, the value was not
5139 reset for any messages that followed in the same SMTP session.
5141 54. The store-handling optimization for building very long strings was not
5142 differentiating between the different store pools. I don't think this
5143 actually made any difference in practice, but I've tidied it.
5145 55. While running the routers to verify a sender address, $sender_address
5146 was still set to the sender address. This is wrong, because when routing to
5147 send a bounce to the sender, it would be empty. Therefore, I have changed
5148 it so that, while verifying a sender address, $sender_address is set to <>.
5149 (There is no change to what happens when verifying a recipient address.)
5151 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5152 target A or AAAA records (if not already returned) without resetting the
5153 qualify_single or search_parents options of the DNS resolver. These are
5154 inappropriate in this case because the targets of MX and SRV records must
5155 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5156 target that, when qualified, matched something in the local domain. These
5157 two options are now turned off when doing these lookups.
5159 57. It seems that at least some releases of Reiserfs (which does not have the
5160 concept of a fixed number of inodes) returns zero and not -1 for the
5161 number of available inodes. This interacted badly with check_spool_inodes,
5162 which assumed that -1 was the "no such thing" setting. What I have done is
5163 to check that the total number of inodes is greater than zero before doing
5164 the test of how many are available.
5166 58. When a "warn" ACL statement has a log_message modifier, the message is
5167 remembered, and not repeated. This is to avoid a lot of repetition when a
5168 message has many recipients that cause the same warning to be written.
5169 Howewer, Exim was preserving the list of already written lines for an
5170 entire SMTP session, which doesn't seem right. The memory is now reset if a
5171 new message is started.
5173 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5174 debugging output unless log_rewrite was also set.
5176 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5177 of (int)(handle) when we know that handle contains (void *)(-1).
5179 61. The Exim daemon panic-logs an error return when it closes the incoming
5180 connection. However "connection reset by peer" seems to be common, and
5181 isn't really an error worthy of noting specially, so that particular error
5184 62. When Exim is trying to find all the local interfaces, it used to panic and
5185 die if the ioctl to get the interface flags failed. However, it seems that
5186 on at least one OS (Solaris 9) it is possible to have an interface that is
5187 included in the list of interfaces, but for which you get a failure error
5188 for this call. This happens when the interface is not "plumbed" into a
5189 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5190 failure of the "get flags" call assumes that the interface is down.
5192 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5193 makes life easier for people who are doing arithmetic on fields extracted
5194 from dates, where you often get leading zeros that should not be
5195 interpreted as octal.
5197 64. Added qualify_domain to the redirect router, to override the global
5200 65. If a pathologically long header line contained very many addresses (the
5201 report of this problem mentioned 10 000) and each of them was rewritten,
5202 Exim could use up a very large amount of memory. (It kept on making new
5203 copies of the header line as it rewrote, and never released the old ones.)
5204 At the expense of a bit more processing, the header rewriting function has
5205 been changed so that it no longer eats memory in this way.
5207 66. The generation of the Received: header has been moved from the time that a
5208 message starts to be received, to the time that it finishes. The timestamp
5209 in the Received: header should now be very close to that of the <= log
5210 line. There are two side-effects of this change:
5212 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5213 logged header lines no longer include the local Received: line, because
5214 it has not yet been created. The same applies to a copy of the message
5215 that is returned to a non-SMTP sender when a message is rejected.
5217 (b) When a filter file is tested using -bf, no additional Received: header
5218 is added to the test message. After some thought, I decided that this
5221 This change does not affect the value of $received_for. It is still set
5222 after address rewriting, but before local_scan() is called.
5224 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5226 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5227 gave an unhelpful panic error message, and a defer error. I have managed to
5228 change this behaviour so that it now rejects any supplied certificate,
5229 which seems right, as the list of acceptable certificates is empty.
5231 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5232 gave an unhelpful defer error. I have not managed to make this reject any
5233 supplied certificates, but the error message it gives is "no certificate
5234 supplied", which is not helpful.
5236 70. exigrep's output now also includes lines that are not associated with any
5237 message, but which match the given pattern. Implemented by a patch from
5238 Martin Sluka, which also tidied up the Perl a bit.
5240 71. Recipient callout verification, like sender verification, was using <> in
5241 the MAIL FROM command. This isn't really the right thing, since the actual
5242 sender may affect whether the remote host accepts the recipient or not. I
5243 have changed it to use the actual sender in the callout; this means that
5244 the cache record is now keyed on a recipient/sender pair, not just the
5245 recipient address. There doesn't seem to be a real danger of callout loops,
5246 since a callout by the remote host to check the sender would use <>.
5247 [SEE ABOVE: changed after hitting problems.]
5249 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5250 temporary errors. However, in the case of such a code being given after
5251 the end of a data transmission (i.e. after ".") Exim was failing to write
5252 a retry record for the message. (Yes, there was some broken host that was
5253 actually sending 8xx at this point.)
5255 73. An unknown lookup type in a host list could cause Exim to panic-die when
5256 the list was checked. (An example that provoked this was putting <; in the
5257 middle of a list instead of at the start.) If this happened during a DATA
5258 ACL check, a -D file could be left lying around. This kind of configuration
5259 error no longer causes Exim to die; instead it causes a defer errror. The
5260 incident is still logged to the main and panic logs.
5262 74. Buglet left over from Exim 3 conversion. The message "too many messages
5263 in one connection" was written to the rejectlog but not the mainlog, except
5264 when address rewriting (yes!) was being logged.
5266 75. Added write_rejectlog option.
5268 76. When a system filter was run not as root (that is, when system_filter_user
5269 was set), the values of the $n variables were not being returned to the
5270 main process; thus, they were not subsequently available in the $sn
5273 77. Added +return_path_on_delivery log selector.
5275 78. A connection timeout was being treated differently from recipients deferred
5276 when testing hosts_max_try with a message that was older than the host's
5277 retry timeout. (The host should not be counted, thus allowing all hosts to
5278 be tried at least once before bouncing.) This may have been the cause of an
5279 occasionally reported bug whereby a message would remain on the queue
5280 longer than the retry timeout, but would be bounced if a delivery was
5281 forced. I say "may" because I never totally pinned down the problem;
5282 setting up timeout/retry tests is difficult. See also the next item.
5284 79. The ultimate address timeout was not being applied to errors that involved
5285 a combination of host plus message (for example, a timeout on a MAIL
5286 command). When an address resolved to a number of possible hosts, and they
5287 were not all tried for each delivery (e.g. because of hosts_max_try), a
5288 message could remain on the queue longer than the retry timeout.
5290 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5293 81. Fixed an obscure SMTP outgoing bug which required at least the following
5294 conditions: (a) there was another message waiting for the same server;
5295 (b) the server returned 5xx to all RCPT commands in the first message so
5296 that the message was not completed; (c) the server dropped the connection
5297 or gave a negative response to the RSET that Exim sends to abort the
5298 transaction. The observed case was a dropped connection after DATA that had
5299 been sent in pipelining mode. That is, the server had advertised PIPELINING
5300 but was not implementing it correctly. The effect of the bug was incorrect
5301 behaviour, such as trying another host, and this could lead to a crash.
5307 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5308 and daemon.c were passed as pointers to ints; they should have been
5309 pointers to socklen_t variables (which are typically unsigned ints).
5311 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5314 3. Fixed a really odd bug that affected only the testing scheme; patching a
5315 certain fixed string in the binary changed the value of another string that
5316 happened to be identical to the end of the original first string.
5318 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5319 name", it returns that address as the IP address. On some operating
5320 systems (e.g. Solaris), it also passes back the IP address string as the
5321 "host name". However, on others (e.g. Linux), it passes back an empty
5322 string. Exim wasn't checking for this, and was changing the host name to an
5323 empty string, assuming it had been canonicized.
5325 5. Although rare, it is permitted to have more than one PTR record for a given
5326 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5327 all the names associated with an address, because they do in Solaris.
5328 However, it seems that they do not in Linux for data that comes from the
5329 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5330 I found this out when I moved to a new Linux workstation and tried to run
5331 the Exim test suite.
5333 To get round this problem I have changed the code so that it now does its
5334 own call to the DNS to look up PTR records when searching for a host name.
5335 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5336 addresses that are only in /etc/hosts are still found.
5338 This behaviour is, however, controlled by an option called host_lookup_
5339 order, which defaults to "bydns:byaddr". If people want to use the other
5340 order, or indeed, just use one or the other means of lookup, they can
5341 specify it in this variable.
5343 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5344 some operating systems, this comes back from gethostbyaddr() as an empty
5345 string, and this is what Exim used to test for. However, it seems that in
5346 other systems, "." is yielded. Exim now tests for this case too.
5348 7. The values of check_spool_space and check_log_space are now held internally
5349 as a number of kilobytes instead of an absolute number of bytes. If a
5350 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5351 kilobyte. This means that much larger values can be stored.
5353 8. Exim monitor: an attempt to get the action menu when not actually pointing
5354 at a message produces an empty menu entitled "No message selected". This
5355 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5356 no entries in it ("Shell widget menu has zero width and/or height"). So I
5357 have added a single, blank menu entry in this case.
5359 9. Added ${quote_local_part.
5361 10. MIME decoding is now applied to the contents of Subject: header lines when
5364 11. Now that a reference to $sender_host_address automatically causes a reverse
5365 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5366 host lookup before query-style lookups in lists that might use this
5367 variable. This has therefore been abolished, and the "net-" prefix is no
5368 longer necessary for query-style lookups.
5370 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5371 have been a typo for LFLAGS, so it has been changed.
5373 13. The install script calls Exim with "-C /dev/null" in order to find the
5374 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5375 to be output. Howeve, since Exim outputs its version number before the
5376 error, it didn't break the script. It just looked ugly. I fixed this by
5377 always allowing "-C /dev/null" if the caller is root.
5379 14. Ignore overlarge ACL variable number when reading spool file - insurance
5380 against a later release with more variables having written the file.
5382 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5383 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5384 this, as well as the form without the "IPv6" on the front.
5386 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5387 OS/Makefile-Darwin file.
5389 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5390 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5392 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5393 variable. However, some OS use ints for inodes. I've added cast to long int
5394 to get rid of the compiler warning.
5396 19. I had forgotten to lock out "/../" in configuration file names when
5397 ALT_CONFIG_PREFIX was set.
5399 20. Routers used for verification do not need to specify transports. However,
5400 if such a router generated a host list, and callout was configured, Exim
5401 crashed, because it could not find a port number from the (non-existent)
5402 transport. It now assumes port 25 in this circumstance.
5404 21. Added the -t option to exigrep.
5406 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5407 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5408 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5409 them is set, LOOKUP_LSEARCH is forced.
5411 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5413 24. Added sender and host information to the "rejected by local_scan()" log
5414 line; previously there was no indication of these.
5416 25. Added .include_if_exists.
5418 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5419 Linux. It turns out that not all file systems support this. Apparently some
5420 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5421 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5422 not supported on the file descriptor, is now ignored when Exim is trying to
5423 sync a directory. This applies only to Linux.
5425 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5427 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5428 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5431 29. Added support for maildirsize files from supplied patch (modified a bit).
5433 30. The use of :fail: followed by an empty string could lead Exim to respond to
5434 sender verification failures with (e.g.):
5436 550 Verification failed for <xxx>
5437 550 Sender verify failed
5439 where the first response line was missing the '-' that indicates it is not
5440 the final line of the response.
5442 31. The loop for finding the name of the user that called Exim had a hardwired
5443 limit of 10; it now uses the value of finduser_retries, which is used for
5444 all other user lookups.
5446 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5449 33. Exim was neglecting to zero errno before one call of strtol() when
5450 expanding a string and expecting an integer value. On some systems this
5451 resulted in spurious "integer overflow" errors. Also, it was casting the
5452 result into an int without checking.
5454 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5455 did not work. The code looks as if it has *never* worked, though it appears
5456 to have been documented since at least releast 1.62. I have made it work.
5458 35. The "timeout_DNS" error in retry rules, also documented since at least
5459 1.62, also never worked. As it isn't clear exactly what this means, and
5460 clearly it isn't a major issue, I have abolished the feature by treating it
5461 as "timeout", and writing a warning to the main and panic logs.
5463 36. The display of retry rules for -brt wasn't always showing the error code
5466 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5467 timeout_connect_A, timeout_connect_MX.
5469 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5470 to the empty sender.
5472 39. The daemon was not analysing the content of -oX till after it had closed
5473 stderr and disconnected from the controlling terminal. This meant that any
5474 syntax errors were only noted on the panic log, and the return code from
5475 the command was 0. By re-arranging the code a little, I've made the
5476 decoding happen first, so such errors now appear on stderr, and the return
5477 code is 1. However, the actual setting up of the sockets still happens in
5478 the disconnected process, so errors there are still only recorded on the
5481 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5482 connections (as happens on some IP stacks) was logged at start up time as
5483 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5484 it from "IPv6 and IPv4", which means that two separate sockets are being
5487 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5488 says whether AF_INET or AF_INET6 was passed as an argument.
5490 42. Exiwhat output was messed up when time zones were included in log
5493 43. Exiwhat now gives more information about the daemon's listening ports,
5494 and whether -tls-on-connect was used.
5496 44. The "port" option of the smtp transport is now expanded.
5498 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5499 silently ignored. Now an error message is written to the main and panic
5502 46. There's a new ACL modifier called "logwrite" which writes to a log file
5503 as soon as it is encountered.
5505 47. Added $local_user_uid and $local_user_gid at routing time.
5507 48. Exim crashed when trying to verify a sender address that was being
5510 49. Exim was recognizing only a space character after ".include". It now also
5511 recognizes a tab character.
5513 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5514 extracting the relevant information from the specification. The man page no
5515 longer contains scrambled data for the -d option, and I've added a section
5516 at the front about calling Exim under different names.
5518 51. Added "extra_headers" argument to the "mail" command in filter files.
5520 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5523 53. Installed eximstats 1.29.
5525 54. Added transport_filter_timeout as a generic transport option.
5527 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5528 Cc: header lines. This was required by RFC 822, but it not required by RFC
5531 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5532 incoming messages that did not have them. Now it does so only if the
5533 message originates locally, that is, if there is no associated remote host
5534 address. When Resent- header lines are present, this applies to the Resent-
5535 lines rather than the non-Resent- lines.
5537 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5538 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5540 58. Messages for configuration errors now include the name of the main
5541 configuration file - useful now that there may be more than one file in a
5542 list (.included file names were always shown).
5544 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5545 for those rare installations that do not start the daemon as root or run it
5546 setuid root. I've cut out the call to initgroups() if the daemon is not
5549 60. The Exim user and group can now be bound into the binary as text strings
5550 that are looked up at the start of Exim's processing.
5552 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5554 62. Added $mailstore_basename variable.
5556 63. Installed patch to sieve.c from Michael Haardt.
5558 64. When Exim failed to open the panic log after failing to open the main log,
5559 the original message it was trying to log was written to stderr and debug
5560 output, but if they were not available (the usual case in production), it
5561 was lost. Now it is written to syslog before the two lines that record the
5562 failures to open the logs.
5564 65. Users' Exim filters run in subprocesses under the user's uid. It is
5565 possible for a "deliver" command or an alias in a "personal" command to
5566 provoke an address rewrite. If logging of address rewriting is configured,
5567 this fails because the process is not running as root or exim. There may be
5568 a better way of dealing with this, but for the moment (because 4.30 needs
5569 to be released), I have disabled address rewrite logging when running a
5570 filter in a non-root, non-exim process.
5576 1. The buildconfig auxiliary program wasn't quoting the value set for
5577 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5578 not defined. This bug was masked in 4.22 by the effect that was fixed in
5581 2. Some messages that were rejected after a message id was allocated were
5582 shown as "incomplete" by exigrep. It no longer does this for messages that
5583 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5585 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5586 have allow_domain_literals set, the ID did not get logged in the <= line.
5587 Domain literals are now always recognized in Message-ID: header lines.
5589 4. The first argument for a ${extract expansion item is the key name or field
5590 number. Leading and trailing spaces in this item were not being ignored,
5591 causing some misleading effects.
5593 5. When deliver_drop_privilege was set, single queue runner processes started
5594 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5595 same command in the process it spins off) were not dropping privilege.
5597 6. When the daemon running as "exim" started a queue runner, it always
5598 re-executed Exim in the spun-off process. This is a waste of effort when
5599 deliver_drop_privilege is set. The new process now just calls the
5600 queue-runner function directly.
5606 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5609 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5610 ignored. Though the use of -oP was forcing the writing of a pid file, it
5611 was always written to the default place.
5613 3. If the message "no IP address found for host xxxx" is generated during
5614 incoming verification, it is now followed by identification of the incoming
5615 connection (so you can more easily find what provoked it).
5617 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5619 5. Added some features to "harden" Exim a bit more against certain attacks:
5621 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5622 be put in Local/Makefile. This is like the never_users runtime option,
5623 but it cannot be overridden. The default setting is "root".
5625 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5626 prefix string with which any file named in a -C command line option
5629 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5630 is retained for -C and -D only if the caller of Exim is root. Without
5631 it, the exim user may also use -C and -D and retain privilege.
5633 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5634 command line option is disabled.
5636 6. Macro names set by the -D option must start with an upper case letter, just
5637 like macro names defined in the configuration file.
5639 7. Added "dereference=" facility to LDAP.
5641 8. Two instances of the typo "uknown" in the source files are fixed.
5643 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5644 the Configure-Makefile script screwed up while processing it.
5646 10. Incorporated PCRE 4.4.
5648 11. The SMTP synchronization check was not operating right at the start of an
5649 SMTP session. For example, it could not catch a HELO sent before the client
5650 waited for the greeting. There is now a check for outstanding input at the
5651 point when the greeting is written. Because of the duplex, asynchronous
5652 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5653 way, but not yet received, when the check is performed.
5655 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5656 on TCP/IP sockets, because this apparently causes some broken clients to
5659 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5660 unchanged) from the Cygwin maintainer.
5662 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5663 was supported instead of testing for mbx. Effectively a typo.
5665 15. The spa authenticator server code was not checking that the input it
5666 received was valid base64.
5668 16. The debug output line for the "set" modifier in ACLs was not showing the
5669 name of the variable that was being set.
5671 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5673 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5674 lookup on its own. Something else was needed to trigger the lookup. For
5675 example, a match in host_lookup or the need for a host name in a host list.
5676 Now, if $sender_host_name is referenced and the host name has not yet been
5677 looked up, a lookup is performed. If the lookup fails, the variable remains
5678 empty, and $host_lookup_failed is set to "1".
5680 19. Added "eqi" as a case-independent comparison operator.
5682 20. The saslauthd authentication condition could segfault if neither service
5683 nor realm was specified.
5685 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5686 error message that was logged was misleading, and incoming SMTP
5687 connections were dropped. The message is now more accurate, and temporary
5688 errors are given to SMTP connections.
5690 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5691 (except RSET) with a 421 error, until QUIT is received. However, it was
5692 failing to send a response to QUIT.
5694 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5695 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5696 ACL is now run after the helo_try_verify_hosts code.
5698 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5699 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5700 case-independent, so other case variants are also recognized). Apparently
5701 some systems use these upper case variants.
5703 25. If more than two messages were waiting for the same host, and a transport
5704 filter was specified for the transport, Exim sent two messages over the
5705 same TCP/IP connection, and then failed with "socket operation on non-
5706 socket" when it tried to send the third.
5708 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5710 27. The extern definition of crypt16() in expand.c was not being excluded when
5711 the OS had its own crypt16() function.
5713 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5714 as a preferred synonym for return_size_limit, both as an option and as an
5717 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5719 30. Changed the default configuration ACL to relax the local part checking rule
5720 for addresses that are not in any local domains. For these addresses,
5721 slashes and pipe symbols are allowed within local parts, but the sequence
5722 /../ is explicitly forbidden.
5724 31. SPA server authentication was not clearing the challenge buffer before
5727 32. log_message in a "warn" ACL statement was writing to the reject log as
5728 well as to the main log, which contradicts the documentation and doesn't
5729 seem right (because no rejection is happening). So I have stopped it.
5731 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5732 However, I am unable to do any testing of this.
5734 34. Fixed an infelicity in the appendfile transport. When checking directories
5735 for a mailbox, to see if any needed to be created, it was accidentally
5736 using path names with one or more superfluous leading slashes; tracing
5737 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5739 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5740 contents are added to the log line that is written for every discarded
5741 recipient. (Previously a log_message setting was ignored.)
5743 36. The ${quote: operator now quotes the string if it is empty.
5745 37. The install script runs exim in order to find its version number. If for
5746 some reason other than non-existence or emptiness, which it checks, it
5747 could not run './exim', it was installing it with an empty version number,
5748 i.e. as "exim-". This error state is now caught, and the installation is
5751 38. An argument was missing from the function that creates an error message
5752 when Exim fails to connect to the socket for saslauthd authentication.
5753 This could cause Exim to crash, or give a corrupted message.
5755 39. Added isip, isip4, and isip6 to ${if conditions.
5757 40. The ACL variables $acl_xx are now saved with the message, and can be
5758 accessed later in routers, transports, and filters.
5760 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5761 strings in the file are not string-expanded.
5763 42. If a MAIL command specified a SIZE value that was too large to fit into an
5764 int variable, the check against message_size_limit failed. Such values are
5765 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5766 day this will have to be increased, but I don't think I want to be around
5767 when emails are that large.
5774 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5775 iconv() is not standard in FreeBSD.
5777 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5778 IPv6 enabled. The observed symptom was a segmentation fault on return from
5779 the function os_common_find_running_interfaces() in src/os.c.
5781 3. In the check_special_case() function in daemon.c I had used "errno" as an
5782 argument name, which causes warnings on some systems. This was basically a
5783 typo, since it was named "eno" in the comments!
5785 4. The code that waits for the clock to tick (at a resolution of some fraction
5786 of a second) so as to ensure message-id uniqueness was always waiting for
5787 at least one whole tick, when it could have waited for less. [This is
5788 almost certainly not relevant at current processor speeds, where it is
5789 unlikely to ever wait at all. But we try to future-proof.]
5791 5. The function that sleeps for a time interval that includes fractions of a
5792 second contained a race. It did not block SIGALRM between setting the
5793 timer, and suspending (a couple of lines later). If the interval was short
5794 and the sigsuspend() was delayed until after it had expired, the suspension
5795 never ended. On busy systems this could lead to processes getting stuck for
5798 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5799 process, before it forks any delivery processes. The open lookup caching
5800 mechanism meant that the open file or database connection was passed into
5801 the delivery process. The problem was that delivery processes always tidy
5802 up cached lookup data. This could cause a problem for the next delivery
5803 process started by the queue runner, because the external queue runner
5804 process does not know about the closure. So the next delivery process
5805 still has data in the lookup cache. In the case of a file lookup, there was
5806 no problem because closing a file descriptor in a subprocess doesn't affect
5807 the parent. However, if the lookup was caching a connection to a database,
5808 the connection was closed, and the second delivery process was likely to
5809 see errors such as "PGSQL: query failed: server closed the connection
5810 unexpectedly". The problem has been fixed by closing all cached lookups
5811 in a queue runner before running a delivery process.
5813 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5814 seem to have the "const" qualifier which it has on other OS. I've
5817 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5818 *of the same type* (client or server) with the same public name that an
5819 error should be diagnosed.
5821 9. When Exim looked up a host name for an IP address, but failed to find the
5822 original IP address when looking up the host name (a safety check), it
5823 output the message "<ip address> does not match any IP for NULL", which was
5824 confusing, to say the least. The bug was that the host name should have
5825 appeared instead of "NULL".
5827 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5828 user that is built into the binary, and the -C or -D options is used, root
5829 privilege is dropped before the configuration file is read. In addition,
5830 logging is switched to stderr instead of the normal log files. If the
5831 configuration then re-defines the Exim user, the unprivileged environment
5832 is probably not what is expected, so Exim logs a panic warning message (but
5835 However, if deliver_drop_privilege is set, the unprivileged state may well
5836 be exactly what is intended, so the warning has been cut out in that case,
5837 and Exim is allowed to try to write to its normal log files.
5843 1. smtp_return_error_details was not giving details for temporary sender
5844 or receiver verification errors.
5846 2. Diagnose a configuration error if two authenticators have the same public
5849 3. Exim used not to create the message log file for a message until the first
5850 delivery attempt. This could be confusing when incoming messages were held
5851 for policy or load reasons. The message log file is now created at the time
5852 the message is received, and an initial "Received" line is written to it.
5854 4. The automatically generated man page for command line options had a minor
5855 bug that caused no ill effects; however, a more serious problem was that
5856 the procedure for building the man page automatically didn't always
5857 operate. Consequently, release 4.20 contains an out-of-date version. This
5858 shouldn't happen again.
5860 5. When building Exim with embedded Perl support, the script that builds the
5861 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5862 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5864 6. The freeze_tell option was not being used for messages that were frozen on
5865 arrival, either by an ACL or by local_scan().
5867 7. Added the smtp_incomplete_transaction log selector.
5869 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5870 was accepting AUTH without a new EHLO.
5872 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5873 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5876 10. Logging of TCP/IP connections (when configured) now happens in the main
5877 daemon process instead of the child process, so that the TCP/IP connection
5878 count is more accurate (but it can never be perfect).
5880 11. The use of "drop" in a nested ACL was not being handled correctly in the
5881 outer ACL. Now, if condition failure induced by the nested "drop" causes
5882 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5883 or "require"), the connection is dropped.
5885 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5886 that yield "discard" can now be used with an "accept" or a "discard" verb,
5887 but an error is generated for any others (because I can't see a useful way
5888 to define what should happen).
5890 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5891 are now processed in the same way as lines in the Exim configuration file.
5892 In particular, continuation lines are supported.
5894 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5896 15. Added -ti meaning -t -i.
5898 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5899 domains, and warn by logging if others are found.
5901 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5902 returned from GIFCONF when Exim is trying to find the list of interfaces on
5903 a host. The code in os.c has been modified to copy each ifreq to an aligned
5904 structure in all cases.
5906 Also, in some cases, the returned ifreq's were being copied to a 'struct
5907 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5908 means the last couple of bytes of an IPv6 address could be chopped if the
5909 ifreq contained only a normal sockaddr (14 bytes storage).
5911 18. Named domain lists were not supported in the hosts_treat_as_local option.
5912 An entry such as +xxxx was not recognized, and was treated as a literal
5915 19. Ensure that header lines added by a DATA ACL are included in the reject log
5916 if the ACL subsequently rejects the message.
5918 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5919 MD5 (which is deprecated).
5921 21. When testing a filter file using -bf, Exim was writing a message when it
5922 took the sender from a "From " line in the message, but it was not doing so
5923 when it took $return_path from a Return-Path: header line. It now does.
5925 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5926 with a valid header line field name (a series of printing characters
5927 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5929 23. Changed "disc" in the source to "disk" to conform to the documentation and
5930 the book and for uniformity.
5932 24. Ignore Sendmail's -Ooption=value command line item.
5934 25. When execve() failed while trying to run a command in a pipe transport,
5935 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5936 could be confused with a return value of 69 from the command itself. This
5937 has been changed to 127, the value the shell returns if it is asked to run
5938 a non-existent command. The wording for the related log line suggests a
5939 non-existent command as the problem.
5941 26. If received_header_text expands to an empty string, do not add a Received:
5942 header line to the message. (Well, it adds a token one on the spool, but
5943 marks it "old" so that it doesn't get used or transmitted.)
5945 27. Installed eximstats 1.28 (addition of -nt option).
5947 28. There was no check for failure on the call to getsockname() in the daemon
5948 code. This can fail if there is a shortage of resources on the system, with
5949 ENOMEM, for example. A temporary error is now given on failure.
5951 29. Contrary to the C standard, it seems that in some environments, the
5952 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5953 program. Exim now does this explicitly; it affects the formatting of
5954 timestamps using strftime().
5956 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5957 complaints. I've now initialized all the variables, to avoid this.
5959 32. Header lines added by a system filter were not being "seen" during
5960 transport-time rewrites.
5962 33. The info_callback() function passed to OpenSSL is set up with type void
5963 (*)(SSL *, int, int), as described somewhere. However, when calling the
5964 function (actually a macro) that sets it up, the type void(*)() is
5965 expected. I've put in a cast to prevent warnings from picky compilers.
5967 34. If a DNS black list lookup found a CNAME record, but there were no A
5968 records associated with the domain it pointed at, Exim crashed.
5970 35. If a DNS black list lookup returned more than one A record, Exim ignored
5971 all but the first. It now scans all returned addresses if a particular IP
5972 value is being sought. In this situation, the contents of the
5973 $dnslist_value variable are a list of all the addresses, separated by a
5976 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5977 to accept a host name and all its aliases if the forward lookup for any of
5978 them yielded the IP address of the incoming connection. Now it accepts only
5979 those names whose forward lookup yields the correct IP address. Any other
5980 names are discarded. This closes a loophole whereby a rogue DNS
5981 administrator could create reverse DNS records to break through a
5982 wildcarded host restriction in an ACL.
5984 37. If a user filter or a system filter that ran in a subprocess used any of
5985 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5986 the wrong values were passed to the pipe command ($thisaddress had the
5987 value of $0, $0 had the value of $1, etc). This bug was introduced by
5988 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5990 38. Improved the line breaking for long SMTP error messages from ACLs.
5991 Previously, if there was no break point between 40 and 75 characters, Exim
5992 left the rest of the message alone. Two changes have been made: (a) I've
5993 reduced the minimum length to 35 characters; (b) if it can't find a break
5994 point between 35 and 75 characters, it looks ahead and uses the first one
5995 that it finds. This may give the occasional overlong line, but at least the
5996 remaining text gets split now.
5998 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5999 file descriptors might be low, and that setting 1000 would always raise it.
6000 It turns out that in some environments, the limit is already over 1000 and
6001 that lowering it causes trouble. So now Exim takes care not to decrease it.
6003 40. When delivering a message, the value of $return_path is set to $sender_
6004 address at the start of routing (routers may change the value). By an
6005 oversight, this default was not being set up when an address was tested by
6006 -bt or -bv, which affected the outcome if any router or filter referred to
6009 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6010 log only when "message" or "log_message" are set. However, if one of the
6011 conditions was an address verification, or a call to a nested ACL, the
6012 messages generated by the underlying test were being passed through. This
6013 no longer happens. The underlying message is available in $acl_verify_
6014 message for both "message" and "log_message" expansions, so it can be
6015 passed through if needed.
6017 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6018 new expansion $bh_ to give the encoded byte string without charset
6019 translation. Translation happens only if iconv() is available; HAVE_ICONV
6020 indicates this at build time. HEADERS_CHARSET gives the charset to
6021 translate to; headers_charset can change it in the configuration, and
6022 "headers charset" can change it in an individual filter file.
6024 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6025 that creates RFC 2047 encoded "words" labels them as that charset instead
6026 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6027 expansion operator; (ii) when Exim creates a From: line for a local
6028 message; (iii) when a header line is rewritten to include a "phrase" part.
6030 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6031 buggy, causing it to skip the first lines of messages whose message ID
6032 ended in 'D'. This would not have bitten before Exim release 4.14, because
6033 message IDs were unlikely to end in 'D' before then. The effect was to have
6034 incorrect size information for certain domains.
6036 45. #include "config.h" was missing at the start of the crypt16.c module. This
6037 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6040 46. If there was a timeout during a "random" callout check, Exim treated it as
6041 a failure of the random address, and carried on sending RSET and the real
6042 address. If the delay was just some slowness somewhere, the response to the
6043 original RCPT would be taken as a response to RSET and so on, causing
6044 mayhem of various kinds.
6046 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6047 when I implemented it. It didn't allow for the fact that some option values
6048 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6049 the right test for positive values.
6051 48. Domain names in DNS records are case-independent. Exim always looks them up
6052 in lower case. Some resolvers return domain names in exactly the case they
6053 appear in the zone file, that is, they may contain uppercase letters. Not
6054 all resolvers do this - some return always lower case. Exim was treating a
6055 change of case by a resolver as a change of domain, similar to a widening
6056 of a domain abbreviation. This triggered its re-routing code and so it was
6057 trying to route what was effectively the same domain again. This normally
6058 caused routing to fail (because the router wouldn't handle the domain
6059 twice). Now Exim checks for this case specially, and just changes the
6060 casing of the domain that it ultimately uses when it transmits the message
6063 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6066 50. If a filter generated a file delivery with a non-absolute name (possible if
6067 no home directory exists for the router), the forbid_file option was not
6070 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6071 the existing equality matching.
6073 52. Exim was using ints instead of ino_t variables in some places where it was
6074 dealing with inode numbers.
6076 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6077 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6078 called TMPDIR, and if it finds it is different, it changes its value.
6080 54. The smtp_printf() function is now made available to local_scan() so
6081 additional output lines can be written before returning. There is also an
6082 smtp_fflush() function to enable the detection of a dropped connection.
6083 The variables smtp_input and smtp_batched_input are exported to
6086 55. Changed the default runtime configuration: the message "Unknown user"
6087 has been removed from the ACL, and instead placed on the localuser router,
6088 using the cannot_route_message feature. This means that any verification
6089 failures that generate their own messages won't get overridden. Similarly,
6090 the "Unrouteable address" message that was in the ACL for unverifiable
6091 relay addresses has also been removed.
6093 56. Added hosts_avoid_esmtp to the smtp transport.
6095 57. The exicyclog script was not checking for the esoteric option
6096 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6097 will work only if exicyclog is run under the appropriate euid.
6099 58. Following a discussion on the list, the rules by which Exim recognises line
6100 endings on incoming messages have been changed. The -dropcr and drop_cr
6101 options are now no-ops, retained only for backwards compatibility. The
6102 following line terminators are recognized: LF CRLF CR. However, special
6103 processing applies to CR:
6105 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6106 nor a local message in the state where . is a terminator.
6108 (ii) If a bare CR is encountered in a header line, an extra space is added
6109 after the line terminator so as not to end the header. The reasoning
6110 behind this is that bare CRs in header lines are most likely either
6111 to be mistakes, or people trying to play silly games.
6113 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6114 was being incorrectly given as 18 bytes larger than it should have been.
6115 This is a VOB (very old bug).
6117 60. This may never have affected anything current, but just in case it has:
6118 When the local host is found other than at the start of a list of hosts,
6119 the local host, those with the same MX, and any that follow, are discarded.
6120 When the list in question was part of a longer list of hosts, the following
6121 hosts (not currently being processed) were also being discarded. This no
6122 longer happens. I'm not sure if this situation could ever has previously
6125 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6128 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6129 header. This is something that is recommended in a new Internet Draft, and
6130 is something that is documented as being done by Sendmail. There are two
6131 possible values. For messages generated by the autoreply transport, Exim
6134 Auto-Submitted: auto-replied
6136 whereas for all other generated messages (e.g. bounces) it adds
6138 Auto-Submitted: auto-generated
6140 63. The "personal" condition in filters now includes a test for the
6141 Auto-Submitted: header. If it contains the string "auto-" the message it
6142 not considered personal.
6144 64. Added rcpt_include_affixes as a generic transport option.
6146 65. Added queue_only_override (default true).
6148 66. Added the syslog_duplication option.
6150 67. If what should have been the first header line of a message consisted of
6151 a space followed by a colon, Exim was mis-interpreting it as a header line.
6152 It isn't of course - it is syntactically invalid and should therefore be
6153 treated as the start of the message body. The misbehaviour could have
6154 caused a number of strange effects, including loss of data in subsequent
6155 header lines, and spool format errors.
6157 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6158 client host had authenticated. This control can now be exercised by an ACL
6159 for more flexibility.
6161 69. By default, callouts do not happen when testing with -bh. There is now a
6162 variant, -bhc, which does actually run the callout code, including
6163 consulting and updating the callout cache.
6165 70. Added support for saslauthd authentication, courtesy of Alexander
6168 71. If statvfs() failed on the spool or log directories while checking their
6169 size for availability, Exim confusingly gave the error "space shortage".
6170 Furthermore, in debugging mode it crashed with a floating point exception.
6171 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6172 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6173 serious problem, Exim now writes to the main and panic logs when this
6174 happens, with details of the failure. It then refuses to accept the
6175 incoming message, giving the message "spool directory problem" or "log
6176 directory problem" with a 421 code for SMTP messages.
6178 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6179 0, 1, and 2 exist, because some OS complain for execs without them (see
6180 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6181 descriptors. However, the code omitted to check that the open succeeded,
6182 causing mysterious errors if for some reason the permissions on /dev/null
6183 got screwed. Now Exim writes a message to the main and panic logs, and
6184 bombs out if it can't open /dev/null.
6186 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6187 interact so that it is all more flexible. It is supposed to remain
6188 backwards compatible. Also added extra_local_interfaces.
6190 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6191 to bomb out with an assertion failure - to the client this appears as a
6192 connection drop. This problem occurs in the part of the code that was taken
6193 from the Samba project. Fortunately, the assertion is in a very simple
6194 function, so I have fixed this by reproducing the function inline in the
6195 one place where it is called, and arranging for authentication to fail
6196 instead of killing the process with assert().
6198 75. The SPA client code was not working when the server requested OEM rather
6199 than Unicode encoding.
6201 76. Added code to make require_files with a specific uid setting more usable in
6202 the case where statting the file as root fails - usually a non-root-mounted
6203 NFS file system. When this happens and the failure is EACCES, Exim now
6204 forks a subprocess and does the per-uid checking as the relevant uid.
6206 77. Added process_log_path.
6208 78. If log_file_path was not explicitly set, a setting of check_log_space or
6209 check_log_inodes was ignored.
6211 79. If a space check for the spool or log partitions fails, the incident is now
6212 logged. Of course, in the latter case the data may get lost...
6214 80. Added the %p formatting code to string_format() so that it can be used to
6215 print addresses in debug_print(). Adjusted all the address printing in the
6216 debugging in store.c to use %p rather than %d.
6218 81. There was a concern that a line of code in smtp_in.c could overflow a
6219 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6220 initially expressed, the concern was not well-founded, because trailing
6221 spaces are removed early. However, if the trailing spaces were followed by
6222 a NULL, they did not get removed, so the overflow was possible. Two fixes
6225 (a) I re-wrote the offending code in a cleaner fashion.
6226 (b) If an incoming SMTP command contains a NULL character, it is rejected
6229 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6230 runs initgroups(), so that if the Exim user is in any additional groups,
6231 they will be used during message reception.
6237 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob