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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
48 TL/01 When running the test suite, the README says that variables such as
49 no_msglog_check are global and can be placed anywhere in a specific
50 test's script, however it was observed that placement needed to be near
51 the beginning for it to behave that way. Changed the runtest perl
52 script to read through the entire script once to detect and set these
53 variables, reset to the beginning of the script, and then run through
54 the script parsing/test process like normal.
56 TL/02 The BSD's have an arc4random API. One of the functions to induce
57 adding randomness was arc4random_stir(), but it has been removed in
58 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
59 function when detected.
61 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
62 cause callback expansion.
64 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
65 syntax errors in an expansion can be treated as a string instead of
66 logging or causing an error, due to the internal use of bool_lax
67 instead of bool when processing it.
69 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
70 server certificates when making smtp deliveries.
72 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
74 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
76 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
78 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
79 Merged patch from Sebastian Wiedenroth.
81 JH/05 Fix results-pipe from transport process. Several recipients, combined
82 with certificate use, exposed issues where response data items split
83 over buffer boundaries were not parsed properly. This eventually
84 resulted in duplicates being sent. This issue only became common enough
85 to notice due to the introduction of conection certificate information,
86 the item size being so much larger. Found and fixed by Wolfgang Breyha.
88 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
89 size buffer was used, resulting in syntax errors when an expansion
92 JH/07 Add support for directories of certificates when compiled with a GnuTLS
93 version 3.3.6 or later.
95 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
96 is EXPERIMENTAL_EVENT, the main-configuration and transport options
97 both become "event_action", the variables become $event_name, $event_data
98 and $event_defer_errno. There is a new variable $verify_mode, usable in
99 routers, transports and related events. The tls:cert event is now also
100 raised for inbound connections, if the main configuration event_action
103 TL/06 In test suite, disable OCSP for old versions of openssl which contained
104 early OCSP support, but no stapling (appears to be less than 1.0.0).
106 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
107 server certificate names available under the smtp transport option
108 "tls_verify_cert_hostname" now do not permit multi-component wildcard
111 JH/10 Time-related extraction expansions from certificates now use the main
112 option "timezone" setting for output formatting, and are consistent
113 between OpenSSL and GnuTLS compilations. Bug 1541.
115 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
116 encoded parameter in the incoming message. Bug 1558.
118 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
119 include certificate info, eximon was claiming there were spoolfile
122 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
124 JH/14 Log delivery-related information more consistently, using the sequence
125 "H=<name> [<ip>]" wherever possible.
127 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
128 are problematic for Debian distribution, omit them from the release
131 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
133 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
136 JH/17 Fix a null-indirection in certextract expansions when a nondefault
137 output list separator was used.
142 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
143 checkers that were complaining about end of non-void function with no
146 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
147 This was a regression intruduced in 4.83 by another bugfix.
149 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
151 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
152 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
158 TF/01 Correctly close the server side of TLS when forking for delivery.
160 When a message was received over SMTP with TLS, Exim failed to clear up
161 the incoming connection properly after forking off the child process to
162 deliver the message. In some situations the subsequent outgoing
163 delivery connection happened to have the same fd number as the incoming
164 connection previously had. Exim would try to use TLS and fail, logging
165 a "Bad file descriptor" error.
167 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
168 utilities have not been installed.
170 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
171 temporary space as the ACL may create new global variables.
173 TL/01 LDAP support uses per connection or global context settings, depending
174 upon the detected version of the libraries at build time.
176 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
177 to extract and use the src ip:port in logging and expansions as if it
178 were a direct connection from the outside internet. PPv2 support was
179 updated based on HAProxy spec change in May 2014.
181 JH/02 Add ${listextract {number}{list}{success}{fail}}.
183 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
184 Properly escape header and check for NULL return.
186 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
189 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
191 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
192 characters in header names, implemented as a verify condition.
193 Contributed by Michael Fischer v. Mollard.
195 TL/05 Rename SPF condition results err_perm and err_temp to standardized
196 results permerror and temperror. Previous values are deprecated but
197 still accepted. In a future release, err_perm and err_temp will be
198 completely removed, which will be a backward incompatibility if the
199 ACL tests for either of these two old results. Patch contributed by
200 user bes-internal on the mailing list.
202 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
204 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
205 selectors, in both main and reject logs.
207 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
210 JH/07 Add malware type "sock" for talking to simple daemon.
212 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
214 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
215 routers/transports under cutthrough routing.
217 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
218 numbers. Touch up "bool" conditional to keep the same definition.
220 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
222 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
224 JH/12 Expand items in router/transport headers_add or headers_remove lists
225 individually rather than the list as a whole. Bug 1452.
227 Required for reasonable handling of multiple headers_ options when
228 they may be empty; requires that headers_remove items with embedded
229 colons must have them doubled (or the list-separator changed).
231 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
232 view the policy declared in the DMARC record. Currently, $dmarc_status
233 is a combined value of both the record presence and the result of the
236 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
238 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
239 dnslookup router and the smtp transport (applying to the forward
242 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
243 of ldap servers used for a specific lookup. Patch provided by Heiko
246 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
247 New variable $lookup_dnssec_authenticated for observability.
249 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
250 Patch submitted by Lars Timman.
252 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
254 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
255 Requires trusted mode and valid format message id, aborts otherwise.
256 Patch contributed by Heiko Schlichting.
258 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
259 certextract with support for various fields. Bug 1358.
261 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
262 is requested by default, modifiable by smtp transport option
265 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
266 operate on certificate variables to give certificate fingerprints
267 Also new ${sha256:cert_variable}.
269 JH/23 The PRDR feature is moved from being Experimental into the mainline.
271 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
272 Christian Aistleitner.
274 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
276 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
277 file. Patch from Wolfgang Breyha.
279 JH/25 Expand the coverage of the delivery $host and $host_address to
280 client authenticators run in verify callout. Bug 1476.
282 JH/26 Port service names are now accepted for tls_on_connect_ports, to
283 align with daemon_smtp_ports. Bug 72.
285 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
286 support and error reporting did not work properly.
288 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
289 and is readable. Patch from Andrew Colin Kissa.
291 TL/14 Enhance documentation of ${run expansion and how it parses the
292 commandline after expansion, particularly in the case when an
293 unquoted variable expansion results in an empty value.
295 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
297 PP/02 Fix internal collision of T_APL on systems which support RFC3123
298 by renaming away from it. Addresses GH issue 15, reported by
301 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
303 TL/15 SECURITY: prevent double expansion in math comparison functions
304 (can expand unsanitized data). Not remotely exploitable.
311 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
313 PP/02 Make -n do something, by making it not do something.
314 When combined with -bP, the name of an option is not output.
316 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
319 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
320 $sender_host_name and config options to manage this, and basic check
323 PP/05 DSCP support for outbound connections and control modifier for inbound.
325 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
326 (Only plugin which currently uses this is kerberos4, which nobody should
327 be using, but we should make it available and other future plugins might
328 conceivably use it, even though it would break NAT; stuff *should* be
329 using channel bindings instead).
331 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
332 name; added for Sendmail compatibility; requires admin caller.
333 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
334 just ignore it); requires trusted caller.
335 Also parse but ignore: -Ac -Am -X<logfile>
338 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
340 TL/02 Add +smtp_confirmation as a default logging option.
342 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
343 Patch by Magnus Holmgren from 2007-02-20.
345 TL/04 Bugzilla 1281 - Spec typo.
346 Bugzilla 1283 - Spec typo.
347 Bugzilla 1290 - Spec grammar fixes.
349 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
351 TL/06 Add Experimental DMARC support using libopendmarc libraries.
353 TL/07 Fix an out of order global option causing a segfault. Reported to dev
354 mailing list by by Dmitry Isaikin.
356 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
358 JH/02 Support "G" suffix to numbers in ${if comparisons.
360 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
362 NM/01 Bugzilla 1197 - Spec typo
363 Bugzilla 1196 - Spec examples corrections
365 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
367 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
368 gnutls_enable_pkcs11, but renamed to more accurately indicate its
371 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
372 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
374 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
375 "acl {{name}{arg}...}", and optional args on acl condition
378 JH/05 Permit multiple router/transport headers_add/remove lines.
380 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
382 JH/07 Avoid using a waiting database for a single-message-only transport.
383 Performance patch from Paul Fisher. Bugzilla 1262.
385 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
388 JH/09 Add $headers_added variable, with content from use of ACL modifier
389 add_header (but not yet added to the message). Bugzilla 199.
391 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
392 Pulled from Bugzilla 817 by Wolfgang Breyha.
394 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
396 (nb: this is the same fix as in Exim 4.80.1)
398 JH/11 Add A= logging on delivery lines, and a client_set_id option on
401 JH/12 Add optional authenticated_sender logging to A= and a log_selector
404 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
406 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
407 advertise SMTP AUTH mechanism to us, instead of a generic
408 protocol violation error. Also, make Exim more robust to bad
409 data from the Dovecot auth socket.
411 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
413 When a queue runner is handling a message, Exim first routes the
414 recipient addresses, during which it prunes them based on the retry
415 hints database. After that it attempts to deliver the message to
416 any remaining recipients. It then updates the hints database using
419 So if a recipient address works intermittently, it can get repeatedly
420 deferred at routing time. The retry hints record remains fresh so the
421 address never reaches the final cutoff time.
423 This is a fairly common occurrence when a user is bumping up against
424 their storage quota. Exim had some logic in its local delivery code
425 to deal with this. However it did not apply to per-recipient defers
426 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
428 This change adds a proper retry rule check during routing so that the
429 final cutoff time is checked against the message's age. We only do
430 this check if there is an address retry record and there is not a
431 domain retry record; this implies that previous attempts to handle
432 the address had the retry_use_local_parts option turned on. We use
433 this as an approximation for the destination being like a local
434 delivery, as in LMTP.
436 I suspect this new check makes the old local delivery cutoff check
437 redundant, but I have not verified this so I left the code in place.
439 TF/02 Correct gecos expansion when From: is a prefix of the username.
441 Test 0254 submits a message to Exim with the header
445 When I ran the test suite under the user fanf2, Exim expanded
446 the header to contain my full name, whereas it should have added
447 a Resent-Sender: header. It erroneously treats any prefix of the
448 username as equal to the username.
450 This change corrects that bug.
452 GF/01 DCC debug and logging tidyup
453 Error conditions log to paniclog rather than rejectlog.
454 Debug lines prefixed by "DCC: " to remove any ambiguity.
456 TF/03 Avoid unnecessary rebuilds of lookup-related code.
458 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
459 Bug spotted by Jeremy Harris; was flawed since initial commit.
460 Would have resulted in OCSP responses post-SNI triggering an Exim
461 NULL dereference and crash.
463 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
465 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
466 Bug detection, analysis and fix by Samuel Thibault.
467 Bugzilla 1331, Debian bug #698092.
469 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
471 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
472 Server implementation by Todd Lyons, client by JH.
473 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
474 config variable "prdr_enable" controls whether the server
475 advertises the facility. If the client requests PRDR a new
476 acl_data_smtp_prdr ACL is called once for each recipient, after
477 the body content is received and before the acl_smtp_data ACL.
478 The client is controlled by bolth of: a hosts_try_prdr option
479 on the smtp transport, and the server advertisement.
480 Default client logging of deliveries and rejections involving
481 PRDR are flagged with the string "PRDR".
483 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
484 fclose(). Diagnosis by Todd Lyons.
486 PP/17 Update configure.default to handle IPv6 localhost better.
487 Patch by Alain Williams (plus minor tweaks).
490 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
491 This is now consistent with GnuTLS, and is now documented: the
492 previous undocumented portable approach to treating the option as
493 unset was to force an expansion failure. That still works, and
494 an empty string is now equivalent.
496 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
497 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
498 not performing validation itself.
500 PP/20 Added force_command boolean option to pipe transport.
501 Patch from Nick Koston, of cPanel Inc.
503 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
506 TF/04 Added udpsend ACL modifer and hexquote expansion operator
508 PP/21 Fix eximon continuous updating with timestamped log-files.
509 Broken in a format-string cleanup in 4.80, missed when I repaired the
510 other false fix of the same issue.
511 Report and fix from Heiko Schlichting.
514 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
515 Report from Prashanth Katuri.
517 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
518 It's SecureTransport, so affects any MacOS clients which use the
519 system-integrated TLS libraries, including email clients.
521 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
522 using a MIME ACL for non-SMTP local injection.
523 Report and assistance in diagnosis by Warren Baker.
525 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
527 JH/16 Fix comparisons for 64b. Bugzilla 1385.
529 TL/09 Add expansion variable $authenticated_fail_id to keep track of
530 last id that failed so it may be referenced in subsequent ACL's.
532 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
535 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
536 ldap library initialization, allowing self-signed CA's to be
537 used. Also properly sets require_cert option later in code by
538 using NULL (global ldap config) instead of ldap handle (per
539 session). Bug diagnosis and testing by alxgomz.
541 TL/12 Enhanced documentation in the ratelimit.pl script provided in
542 the src/util/ subdirectory.
544 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
545 renamed to Transport Post Delivery Action by Jeremy Harris, as
548 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
549 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
550 redis_servers = needs to be configured which will be used by the redis
551 lookup. Patch from Warren Baker, of The Packet Hub.
553 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
555 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
556 hostname or reverse DNS when processing a host list. Used suggestions
557 from multiple comments on this bug.
559 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
561 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
562 Missed a few lines, added it to make the runtest require no keyboard
565 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
566 contains upper case chars. Make router use caseful_local_part.
568 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
569 support when GnuTLS has been built with p11-kit.
575 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
577 This, or similar/improved, will also be change PP/11 of 4.82.
583 PP/01 Handle short writes when writing local log-files.
584 In practice, only affects FreeBSD (8 onwards).
585 Bugzilla 1053, with thanks to Dmitry Isaikin.
587 NM/01 Bugzilla 949 - Documentation tweak
589 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
592 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
594 PP/02 Implemented gsasl authenticator.
596 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
598 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
599 `pkg-config foo` for cflags/libs.
601 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
602 with rest of GSASL and with heimdal_gssapi.
604 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
605 `pkg-config foo` for cflags/libs for the TLS implementation.
607 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
608 properties get this fed in as external SSF. A number of robustness
609 and debugging improvements to the cyrus_sasl authenticator.
611 PP/08 cyrus_sasl server now expands the server_realm option.
613 PP/09 Bugzilla 1214 - Log authentication information in reject log.
614 Patch by Jeremy Harris.
616 PP/10 Added dbmjz lookup type.
618 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
620 PP/12 MAIL args handles TAB as well as SP, for better interop with
621 non-compliant senders.
622 Analysis and variant patch by Todd Lyons.
624 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
625 Bug report from Lars Müller <lars@samba.org> (via SUSE),
626 Patch from Dirk Mueller <dmueller@suse.com>
628 PP/13 tls_peerdn now print-escaped for spool files.
629 Observed some $tls_peerdn in wild which contained \n, which resulted
630 in spool file corruption.
632 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
633 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
634 or write after TLS renegotiation, which otherwise led to messages
637 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
638 as a tracking header (ie: a signed header comes before the signature).
639 Patch from Wolfgang Breyha.
641 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
642 comma-sep list; embedded commas doubled.
644 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
646 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
648 Report and patch from Dmitry Banschikov.
650 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
651 Removed SSL_clear() after SSL_new() which led to protocol negotiation
652 failures. We appear to now support TLS1.1+ with Exim.
654 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
655 lets Exim select keys and certificates based upon TLS SNI from client.
656 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
657 before an outbound SMTP session. New log_selector, +tls_sni.
659 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
660 NULL dereference. Report and patch from Alun Jones.
662 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
663 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
664 Not seeing resolver debug output on NetBSD, but suspect this is a
665 resolver implementation change.
667 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
668 Left warnings. Added "eximon gdb" invocation mode.
670 PP/21 Defaulting "accept_8bitmime" to true, not false.
672 PP/22 Added -bw for inetd wait mode support.
674 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
675 locate the relevant includes and libraries. Made this the default.
677 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
678 Bugzilla 1246, report and most of solution from Tomasz Kusy.
680 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
681 This may cause build issues on older platforms.
683 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
684 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
685 gnutls_require_mac & gnutls_require_protocols (no longer supported).
686 Added SNI support via GnuTLS too.
687 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
689 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
691 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
692 Applied second patch from Janne, implementing suggestion to default
693 multiple-strings-in-record handling to match SPF spec.
695 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
697 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
698 read-only, out of scope).
699 Patch from Wolfgang Breyha, report from Stuart Northfield.
701 PP/29 Fix three issues highlighted by clang analyser static analysis.
702 Only crash-plausible issue would require the Cambridge-specific
703 iplookup router and a misconfiguration.
704 Report from Marcin Mirosław.
706 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
708 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
709 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
710 As part of this, removing so much warning spew let me fix some minor
711 real issues in debug logging.
713 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
714 assignment on my part. Fixed.
716 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
717 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
718 Janne Snabb (who went above and beyond: thank you).
720 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
721 string otherwise requires a connection and a bunch more work and it's
722 relatively easy to get wrong. Should also expose TLS library linkage
725 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
726 64-bit ${eval} (JH/03).
728 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
729 GNU libc to support some of the 64-bit stuff, should not lead to
730 conflicts. Defined before os.h is pulled in, so if a given platform
731 needs to override this, it can.
733 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
734 protection layer was required, which is not implemented.
735 Bugzilla 1254, patch from Wolfgang Breyha.
737 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
738 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
739 tls_dhparam take prime identifiers. Also unbreak combination of
740 OpenSSL+DH_params+TLSSNI.
742 PP/39 Disable SSLv2 by default in OpenSSL support.
748 PP/01 Solaris build fix for Oracle's LDAP libraries.
749 Bugzilla 1109, patch from Stephen Usher.
751 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
753 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
756 TF/02 Fix a couple more cases where we did not log the error message
757 when unlink() failed. See also change 4.74-TF/03.
759 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
760 lock up or crash if it happened to be inside a call to libc when it
761 got a SIGUSR1 from exiwhat.
763 The SIGUSR1 handler appends the current process status to the process
764 log which is later printed by exiwhat. It used to use the general
765 purpose logging code to do this, but several functions it calls are
766 not safe for signals.
768 The new output code in the SIGUSR1 handler is specific to the process
769 log, and simple enough that it's easy to inspect for signal safety.
770 Removing some special cases also simplifies the general logging code.
771 Removing the spurious timestamps from the process log simplifies
774 TF/04 Improved ratelimit ACL condition.
776 The /noupdate option has been deprecated in favour of /readonly which
777 has clearer semantics. The /leaky, /strict, and /readonly update modes
778 are mutually exclusive. The update mode is no longer included in the
779 database key; it just determines when the database is updated. (This
780 means that when you upgrde Exim will forget old rate measurements.)
782 Exim now checks that the per_* options are used with an update mode that
783 makes sense for the current ACL. For example, when Exim is processing a
784 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
785 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
786 must specify per_mail/readonly. If you omit the update mode it defaults to
787 /leaky where that makes sense (as before) or /readonly where required.
789 The /noupdate option is now undocumented but still supported for
790 backwards compatibility. It is equivalent to /readonly except that in
791 ACLs where /readonly is required you may specify /leaky/noupdate or
792 /strict/noupdate which are treated the same as /readonly.
794 A useful new feature is the /count= option. This is a generalization
795 of the per_byte option, so that you can measure the throughput of other
796 aggregate values. For example, the per_byte option is now equivalent
797 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
799 The per_rcpt option has been generalized using the /count= mechanism
800 (though it's more complicated than the per_byte equivalence). When it is
801 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
802 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
803 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
804 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
805 Note that using per_rcpt with a non-readonly update mode in more than
806 one ACL will cause the recipients to be double-counted. (The per_mail
807 and per_byte options don't have this problem.)
809 The handling of very low rates has changed slightly. If the computed rate
810 is less than the event's count (usually one) then this event is the first
811 after a long gap. In this case the rate is set to the same as this event's
812 count, so that the first message of a spam run is counted properly.
814 The major new feature is a mechanism for counting the rate of unique
815 events. The new per_addr option counts the number of different
816 recipients that someone has sent messages to in the last time period. It
817 behaves like per_rcpt if all the recipient addresses are different, but
818 duplicate recipient addresses do not increase the measured rate. Like
819 the /count= option this is a general mechanism, so the per_addr option
820 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
821 example, measure the rate that a client uses different sender addresses
822 with the options per_mail/unique=$sender_address. There are further
823 details in the main documentation.
825 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
827 TF/06 Removed a few PCRE remnants.
829 TF/07 Automatically extract Exim's version number from tags in the git
830 repository when doing development or release builds.
832 PP/02 Raise smtp_cmd_buffer_size to 16kB.
833 Bugzilla 879. Patch from Paul Fisher.
835 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
836 Heavily based on revision 40f9a89a from Simon Arlott's tree.
839 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
841 PP/05 Variable $av_failed, true if the AV scanner deferred.
842 Bugzilla 1078. Patch from John Horne.
844 PP/06 Stop make process more reliably on build failure.
845 Bugzilla 1087. Patch from Heiko Schlittermann.
847 PP/07 Make maildir_use_size_file an _expandable_ boolean.
848 Bugzilla 1089. Patch from Heiko Schlittermann.
850 PP/08 Handle ${run} returning more data than OS pipe buffer size.
851 Bugzilla 1131. Patch from Holger Weiß.
853 PP/09 Handle IPv6 addresses with SPF.
854 Bugzilla 860. Patch from Wolfgang Breyha.
856 PP/10 GnuTLS: support TLS 1.2 & 1.1.
858 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
861 PP/11 match_* no longer expand right-hand-side by default.
862 New compile-time build option, EXPAND_LISTMATCH_RHS.
863 New expansion conditions, "inlist", "inlisti".
865 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
867 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
869 PP/14 fix log_write() format string regression from TF/03.
870 Bugzilla 1152. Patch from Dmitry Isaikin.
876 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
878 PP/02 Harmonised TLS library version reporting; only show if debugging.
879 Layout now matches that introduced for other libraries in 4.74 PP/03.
881 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
883 PP/04 New "dns_use_edns0" global option.
885 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
888 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
889 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
891 TK/01 Updated PolarSSL code to 0.14.2.
892 Bugzilla 1097. Patch from Andreas Metzler.
894 PP/07 Catch divide-by-zero in ${eval:...}.
897 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
900 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
901 format-string attack -- SECURITY: remote arbitrary code execution.
903 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
904 time unintentionally subject to list matching rules, letting the header
905 cause arbitrary Exim lookups (of items which can occur in lists, *not*
906 arbitrary string expansion). This allowed for information disclosure.
908 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
909 INT_MIN/-1 -- value coerced to INT_MAX.
915 NM/01 Workround for PCRE version dependency in version reporting
918 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
919 This fixes portability to compilers other than gcc, notably
920 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
922 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
923 makefiles for portability to HP-UX and POSIX correctness.
925 PP/01 Permit LOOKUP_foo enabling on the make command-line.
926 Also via indented variable definition in the Makefile.
927 (Debugging by Oliver Heesakkers).
929 PP/02 Restore caching of spamd results with expanded spamd_address.
930 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
932 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
933 Improves build reliability. Fix from: Frank Elsner
935 NM/02 Fix wide character breakage in the rfc2047 coding
936 Fixes bug 1064. Patch from Andrey N. Oktyabrski
938 NM/03 Allow underscore in dnslist lookups
939 Fixes bug 1026. Patch from Graeme Fowler
941 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
942 Code patches from Adam Ciarcinski of NetBSD.
944 NM/04 Fixed exiqgrep to cope with mailq missing size issue
947 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
948 is logged, to avoid truncation. Patch from John Horne.
950 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
951 Patch from Jakob Hirsch.
953 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
954 SQL string expansion failure details.
955 Patch from Andrey Oktyabrski.
957 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
958 Patch from Simon Arlott.
960 PP/09 New lookups functionality failed to compile on old gcc which rejects
961 extern declarations in function scope.
962 Patch from Oliver Fleischmann
964 PP/10 Use sig_atomic_t for flags set from signal handlers.
965 Check getgroups() return and improve debugging.
966 Fixed developed for diagnosis in bug 927 (which turned out to be
969 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
970 Patch from Mark Zealey.
972 PP/12 Bugzilla 1056: Improved spamd server selection.
973 Patch from Mark Zealey.
975 PP/13 Bugzilla 1086: Deal with maildir quota file races.
976 Based on patch from Heiko Schlittermann.
978 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
979 Patch from Uwe Doering, sign-off by Michael Haardt.
981 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
982 variable declaration deep within a block. Bug and patch from
985 PP/15 lookups-Makefile IRIX compatibilty coercion.
987 PP/16 Make DISABLE_DKIM build knob functional.
989 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
990 Patch by Simon Arlott
992 TF/03 Fix valgrind.h portability to C89 compilers that do not support
993 variable argument macros. Our copy now differs from upstream.
999 TF/01 Failure to get a lock on a hints database can have serious
1000 consequences so log it to the panic log.
1002 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1003 controlled using the smtp_confirmation log selector.
1005 TF/03 Include the error message when we fail to unlink a spool file.
1007 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1008 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1009 for maintaining out-of-tree patches for some time.
1011 PP/01 Bugzilla 139: Documentation and portability issues.
1012 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1013 Handle per-OS dynamic-module compilation flags.
1015 PP/02 Let /dev/null have normal permissions.
1016 The 4.73 fixes were a little too stringent and complained about the
1017 permissions on /dev/null. Exempt it from some checks.
1018 Reported by Andreas M. Kirchwitz.
1020 PP/03 Report version information for many libraries, including
1021 Exim version information for dynamically loaded libraries. Created
1022 version.h, now support a version extension string for distributors
1023 who patch heavily. Dynamic module ABI change.
1025 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1026 privilege escalation vulnerability whereby the Exim run-time user
1027 can cause root to append content of the attacker's choosing to
1030 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1033 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1034 If dropping privileges for untrusted macros, we disabled normal logging
1035 on the basis that it would fail; for the Exim run-time user, this is not
1036 the case, and it resulted in successful deliveries going unlogged.
1037 Fixed. Reported by Andreas Metzler.
1043 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1044 only prepend for the Resent-* case. Fixes regression introduced in
1045 Exim 4.70 by NM/22 for Bugzilla 607.
1047 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1048 increasing numbers of administrators be bitten by this.
1050 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1052 PP/03 Bugzilla 994: added openssl_options main configuration option.
1054 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1056 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1058 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1060 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1061 without a peer certificate, leading to a segfault because of an
1062 assumption that peers always have certificates. Be a little more
1063 paranoid. Problem reported by Martin Tscholak.
1065 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1066 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1067 NB: ClamAV planning to remove STREAM in "middle of 2010".
1068 CL also introduces -bmalware, various -d+acl logging additions and
1069 more caution in buffer sizes.
1071 PP/09 Implemented reverse_ip expansion operator.
1073 PP/10 Bugzilla 937: provide a "debug" ACL control.
1075 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1077 PP/12 Bugzilla 973: Implement --version.
1079 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1081 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1083 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1085 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1086 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1087 ignore trailing whitespace.
1089 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1091 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1094 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1095 Notification from Dr Andrew Aitchison.
1097 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1098 ExtendedDetectionInfo response format.
1099 Notification from John Horne.
1101 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1104 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1105 XSL and documented dependency on system catalogs, with examples of how
1108 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1111 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1112 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1113 configuration file which is writeable by the Exim user or group.
1115 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1116 of configuration files to cover files specified with the -C option if
1117 they are going to be used with root privileges, not just the default
1120 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1121 option (effectively making it always true).
1123 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1124 files to be used while preserving root privileges.
1126 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1127 that rogue child processes cannot use them.
1129 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1130 run-time user, instead of root.
1132 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1133 Exim run-time user without dropping privileges.
1135 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1136 result string, instead of calling string_vformat() twice with the same
1139 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1140 for other users. Others should always drop root privileges if they use
1141 -C on the command line, even for a whitelisted configure file.
1143 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1145 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1151 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1152 $data_path, and $header_path variables; fixed documentation bugs and
1155 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1156 exipick to access non-standard spools, including the "frozen" queue
1159 NM/01 Bugzilla 965: Support mysql stored procedures.
1160 Patch from Alain Williams
1162 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1164 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1165 Patch from Andreas Metzler
1167 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1168 Patch from Kirill Miazine
1170 NM/05 Bugzilla 671: Added umask to procmail example.
1172 JJ/03 installed exipick 20100323.0, fixing doc bug
1174 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1175 directory. Notification and patch from Dan Rosenberg.
1177 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1179 TK/02 Improve log output when DKIM signing operation fails.
1181 MH/01 Treat the transport option dkim_domain as a colon separated
1182 list, not as a single string, and sign the message with each element,
1183 omitting multiple occurences of the same signer.
1185 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1186 Bugzilla 985, 986. Patch by Simon Arlott
1188 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1189 Patch by Simon Arlott
1191 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1192 MBX locking. Notification from Dan Rosenberg.
1198 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1200 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1202 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1204 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1206 NM/04 Bugzilla 915: Flag broken perl installation during build.
1212 TK/01 Added patch by Johannes Berg that expands the main option
1213 "spamd_address" if it starts with a dollar sign.
1215 TK/02 Write list of recipients to X-Envelope-Sender header when building
1216 the mbox-format spool file for content scanning (suggested by Jakob
1219 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1220 (http://www.dcc-servers.net/) support via dccifd. Activated by
1221 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1223 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1224 by Mark Daniel Reidel <mr@df.eu>.
1226 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1227 When building exim an external PCRE library is now needed -
1228 PCRE is a system library on the majority of modern systems.
1229 See entry on PCRE_LIBS in EDITME file.
1231 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1232 conversation. Added nologin parameter to request.
1233 Patch contributed by Kirill Miazine.
1235 TF/01 Do not log submission mode rewrites if they do not change the address.
1237 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1239 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1240 log files in place. Contributed by Roberto Lima.
1242 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1244 TF/03 Bugzilla 615: When checking the local_parts router precondition
1245 after a local_part_suffix or local_part_prefix option, Exim now
1246 does not use the address's named list lookup cache, since this
1247 contains cached lookups for the whole local part.
1249 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1250 Robert Millan. Documentation is in experimental-spec.txt.
1252 TF/04 Bugzilla 668: Fix parallel build (make -j).
1254 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1256 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1257 Patch provided by Jan Srzednicki.
1259 TF/05 Leading white space used to be stripped from $spam_report which
1260 wrecked the formatting. Now it is preserved.
1262 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1263 that they are available at delivery time.
1265 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1267 TF/08 TLS error reporting now respects the incoming_interface and
1268 incoming_port log selectors.
1270 TF/09 Produce a more useful error message if an SMTP transport's hosts
1271 setting expands to an empty string.
1273 NM/06 Bugzilla 744: EXPN did not work under TLS.
1274 Patch provided by Phil Pennock.
1276 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1277 Patch provided by Richard Godbee.
1279 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1280 acl_smtp_notquit, added index entry.
1282 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1283 Patch provided by Eugene Bujak.
1285 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1286 accept(). Patch provided by Maxim Dounin.
1288 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1289 Patch provided by Phil Pennock.
1291 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1293 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1294 Patch provided by Brad "anomie" Jorsch.
1296 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1297 Patch provided by Dean Brooks.
1299 TK/05 Add native DKIM support (does not depend on external libraries).
1301 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1302 Patch provided by Graeme Fowler.
1304 NM/16 Bugzilla 851: Documentation example syntax fix.
1306 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1308 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1311 NM/19 Bugzilla 745: TLS version reporting.
1312 Patch provided by Phil Pennock.
1314 NM/20 Bugzilla 167: bool: condition support.
1315 Patch provided by Phil Pennock.
1317 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1318 clients. Patch provided by Phil Pennock.
1320 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1321 Patch provided by Brad "anomie" Jorsch.
1323 NM/23 Bugzilla 687: Fix misparses in eximstats.
1324 Patch provided by Heiko Schlittermann.
1326 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1327 Patch provided by Heiko Schlittermann.
1329 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1330 plus update to original patch.
1332 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1334 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1335 Patch provided by David Brownlee.
1337 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1339 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1341 NM/30 Bugzilla 888: TLS documentation bugfixes.
1343 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1345 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1346 Unlike the original bugzilla I have changed all shell scripts in src tree.
1348 NM/33 Bugzilla 898: Transport filter timeout fix.
1349 Patch by Todd Rinaldo.
1351 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1352 Patch by Serge Demonchaux.
1354 NM/35 Bugzilla 39: Base64 decode bug fixes.
1355 Patch by Jakob Hirsch.
1357 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1359 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1361 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1363 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1369 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1370 ALT-N's libdkim that I have put here:
1371 http://duncanthrax.net/exim-experimental/
1373 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1374 were called 'true' and it seems that C99 defines that as a reserved
1375 keyword to be used with 'bool' variable types. That means you could
1376 not include C99-style headers which use bools without triggering
1377 build errors in sieve.c.
1379 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1380 as mailq or other aliases. Changed the --help handling significantly
1381 to do whats expected. exim_usage() emits usage/help information.
1383 SC/01 Added the -bylocaldomain option to eximstats.
1385 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1387 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1389 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1395 PH/01 Another patch from the Sieve maintainer.
1397 PH/02 When an IPv6 address is converted to a string for single-key lookup
1398 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1399 dots are used instead of colons so that keys in lsearch files need not
1400 contain colons. This was done some time before quoting was made available
1401 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1402 (notated using the quote facility) so as to distinguish them from IPv4
1403 keys. This meant that lookups for IP addresses in host lists did not work
1404 for iplsearch lookups.
1406 This has been fixed by arranging for IPv6 addresses to be expressed with
1407 colons if the lookup type is iplsearch. This is not incompatible, because
1408 previously such lookups could never work.
1410 The situation is now rather anomolous, since one *can* have colons in
1411 ordinary lsearch keys. However, making the change in all cases is
1412 incompatible and would probably break a number of configurations.
1414 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1417 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1418 conversion specification without a maximum field width, thereby enabling
1419 a rogue spamd server to cause a buffer overflow. While nobody in their
1420 right mind would setup Exim to query an untrusted spamd server, an
1421 attacker that gains access to a server running spamd could potentially
1422 exploit this vulnerability to run arbitrary code as the Exim user.
1424 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1425 $primary_hostname instead of what libspf2 thinks the hosts name is.
1427 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1428 a directory entry by the name of the lookup key. Previously, if a
1429 symlink pointed to a non-existing file or a file in a directory that
1430 Exim lacked permissions to read, a lookup for a key matching that
1431 symlink would fail. Now it is enough that a matching directory entry
1432 exists, symlink or not. (Bugzilla 503.)
1434 PH/03 The body_linecount and body_zerocount variables are now exported in the
1437 PH/04 Added the $dnslist_matched variable.
1439 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1440 This means they are set thereafter only if the connection becomes
1443 PH/06 Added the client_condition to authenticators so that some can be skipped
1444 by clients under certain conditions.
1446 PH/07 The error message for a badly-placed control=no_multiline_responses left
1447 "_responses" off the end of the name.
1449 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1451 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1452 (without spaces) instead of just copying the configuration text.
1454 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1456 PH/11 Added $max_received_linelength.
1458 PH/12 Added +ignore_defer and +include_defer to host lists.
1460 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1461 way in which PCRE > 7.0 is built.
1463 PH/14 Implemented queue_only_load_latch.
1465 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1466 MAIL command. The effect was to mangle the value on 64-bit systems.
1468 PH/16 Another patch from the Sieve maintainer.
1470 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1472 PH/18 If a system quota error occurred while trying to create the file for
1473 a maildir delivery, the message "Mailbox is full" was not appended to the
1474 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1475 only to a quota excession during the actual writing of the file.
1477 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1478 characters?) which causes problems in log lines. The DN values are now
1479 passed through string_printing() before being added to log lines.
1481 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1482 and InterBase are left for another time.)
1484 PH/21 Added message_body_newlines option.
1486 PH/22 Guard against possible overflow in moan_check_errorcopy().
1488 PH/23 POSIX allows open() to be a macro; guard against that.
1490 PH/24 If the recipient of an error message contained an @ in the local part
1491 (suitably quoted, of course), incorrect values were put in $domain and
1492 $local_part during the evaluation of errors_copy.
1498 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1499 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1502 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1503 issue a MAIL command.
1505 PH/02 In an ACL statement such as
1507 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1509 if a client was not listed at all, or was listed with a value other than
1510 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1511 the condition was not true (as it should be), so access was not denied.
1512 The bug was that the ! inversion was incorrectly passed on to the second
1513 item. This has been fixed.
1515 PH/03 Added additional dnslists conditions == and =& which are different from
1516 = and & when the dns lookup returns more than one IP address.
1518 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1519 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1521 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1522 FSYNC, which compiles an option called disable_fsync that allows for
1523 bypassing fsync(). The documentation is heavily laced with warnings.
1525 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1527 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1528 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1529 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1530 including adding "make clean"; (3) Added -fPIC when compiling the test
1531 dynamically loaded module, to get rid of a warning.
1533 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1534 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1535 = 0s. The bug is otherwise harmless.
1537 PH/07 There was a bug in the dovecot authenticator such that the value of
1538 $auth1 could be overwritten, and so not correctly preserved, after a
1539 successful authentication. This usually meant that the value preserved by
1540 the server_setid option was incorrect.
1542 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1544 PH/09 Installed PCRE release 7.0.
1546 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1547 run for batched SMTP input. It is now run at the start of every message
1548 in the batch. While fixing this I discovered that the process information
1549 (output by running exiwhat) was not always getting set for -bs and -bS
1550 input. This is fixed, and it now also says "batched" for BSMTP.
1552 PH/11 Added control=no_pipelining.
1554 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1555 patch, slightly modified), and move the expansion of helo_data till after
1556 the connection is made in the smtp transport (so it can use these
1559 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1561 PH/14 Added log_selector = +pid.
1563 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1565 PH/16 Add ${if forany and ${if forall.
1567 PH/17 Added dsn_from option to vary the From: line in DSNs.
1569 PH/18 Flush SMTP output before performing a callout, unless control =
1570 no_callout_flush is set.
1572 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1573 was true (the default) a successful delivery failed to delete the retry
1574 item, thus causing premature timeout of the address. The bug is now
1577 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1579 PH/21 Long custom messages for fakedefer and fakereject are now split up
1580 into multiline reponses in the same way that messages for "deny" and
1581 other ACL rejections are.
1583 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1584 with slight modification.
1586 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1587 draft, changing the syntax and factoring some duplicate code".
1589 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1590 for deliveries of the second and subsequent messages over the same SMTP
1593 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1594 ${reduce, with only minor "tidies".
1596 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1598 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1599 expansion side effects.
1601 PH/27 When a message times out after an over-quota error from an Exim-imposed
1602 quota, the bounce message says "mailbox is full". This message was not
1603 being given when it was a system quota that was exceeded. It now should
1606 MH/03 Made $recipients available in local_scan(). local_scan() already has
1607 better access to the recipient list through recipients_list[], but
1608 $recipients can be useful in postmaster-provided expansion strings.
1610 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1611 in the case of a MAIL command with additional options following the
1612 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1613 were accidentally chopped off.
1615 PH/29 SMTP synchronization checks are implemented when a command is read -
1616 there is a check that no more input is waiting when there shouldn't be
1617 any. However, for some commands, a delay in an ACL can mean that it is
1618 some time before the response is written. In this time, more input might
1619 arrive, invalidly. So now there are extra checks after an ACL has run for
1620 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1621 pipelining has not been advertised.
1623 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1625 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1626 returned in the SMTP response when smtp_return_error_details was false.
1627 This has been fixed.
1629 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1630 instead of the C I/O that was originally supplied, because problems were
1631 reported on Solaris.
1633 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1634 Exim which did not show up earlier: it was assuming that a call to
1635 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1636 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1637 was a macro that became an assignment, so it seemed to work. This has
1638 changed to a proper function call with a void return, hence the compile
1639 error. Exim's code has been fixed.
1641 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1644 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1646 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1648 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1649 because it turns out to be too big or there is a timeout) there is no
1650 "Completed" line in the log. When some messages of this type were
1651 selected by exigrep, they were listed as "not completed". Others were
1652 picked up by some special patterns. I have improved the selection
1653 criteria to be more general.
1655 PH/38 The host_find_failed option in the manualroute router can now be set
1656 to "ignore", to completely ignore a host whose IP address cannot be
1657 found. If all hosts are ignored, the behaviour is controlled by the new
1658 host_all_ignored option.
1660 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1661 homing or because of multiple MX records with /mx) generated more than
1662 one IP address, and the following item turned out to be the local host,
1663 all the secondary addresses of the first item were incorrectly removed
1664 from the list, along with the local host and any following hosts (which
1665 is what is supposed to happen).
1667 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1668 whoever called Exim into the -H file. In the case of the daemon it was
1669 behaving confusingly. When first started, it used values for whoever
1670 started the daemon, but after a SIGHUP it used the Exim user (because it
1671 calls itself on a restart). I have changed the code so that it now always
1674 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1675 message are rejected with the same error (e.g. no authentication or bad
1676 sender address), and a DATA command is nevertheless sent (as can happen
1677 with PIPELINING or a stupid MUA), the error message that was given to the
1678 RCPT commands is included in the rejection of the DATA command. This is
1679 intended to be helpful for MUAs that show only the final error to their
1682 PH/42 Another patch from the Sieve maintainer.
1684 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1685 Eximstats - Fixed some broken HTML links and added missing column headers
1687 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1688 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1690 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1692 PH/43 Yet another patch from the Sieve maintainer.
1694 PH/44 I found a way to check for a TCP/IP connection going away before sending
1695 the response to the final '.' that terminates a message, but only in the
1696 case where the client has not sent further data following the '.'
1697 (unfortunately, this is allowed). However, in many cases there won't be
1698 any further data because there won't be any more messages to send. A call
1699 to select() can be used: if it shows that the input is "ready", there is
1700 either input waiting, or the socket has been closed. An attempt to read
1701 the next input character can distinguish the two cases. Previously, Exim
1702 would have sent an OK response which the client would never have see.
1703 This could lead to message repetition. This fix should cure that, at
1704 least in a lot of common cases.
1706 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1707 advertised in response to EHLO.
1713 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1714 fixed by 4.65/MH/01 (is this a record?) are fixed:
1716 (i) An empty string was always treated as zero by the numeric comparison
1717 operators. This behaviour has been restored.
1719 (ii) It is documented that the numeric comparison operators always treat
1720 their arguments as decimal numbers. This was broken in that numbers
1721 starting with 0 were being interpreted as octal.
1723 While fixing these problems I realized that there was another issue that
1724 hadn't been noticed. Values of message_size_limit (both the global option
1725 and the transport option) were treated as octal if they started with 0.
1726 The documentation was vague. These values are now always treated as
1727 decimal, and I will make that clear in the documentation.
1733 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1734 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1737 MH/01 Don't check that the operands of numeric comparison operators are
1738 integers when their expansion is in "skipping" mode (fixes bug
1739 introduced by 4.64-PH/07).
1741 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1742 child addresses, Exim now panics and dies. Previously, because the count
1743 is held in a short int, deliveries were likely to be lost. As such a
1744 large number of recipients for a single message is ridiculous
1745 (performance will be very, very poor), I have chosen to impose a limit
1746 rather than extend the field.
1752 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1753 leftover -K file (the existence of which was triggered by #402).
1754 While we were at it, introduced process PID as part of the -K
1755 filename. This should rule out race conditions when creating
1758 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1759 processing considerably. Previous code took too long for large mails,
1760 triggering a timeout which in turn triggers #401.
1762 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1763 in the DK code in transports.c. sendfile() is not really portable,
1764 hence the _LINUX specificness.
1766 TF/01 In the add_headers option to the mail command in an Exim filter,
1767 there was a bug that Exim would claim a syntax error in any
1768 header after the first one which had an odd number of characters
1771 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1772 callout verification, Exim cached a "reject" for the entire domain. This
1773 is correct for most verifications, but it is not correct for a recipient
1774 verification with use_sender or use_postmaster set, because in that case
1775 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1776 case of MAIL FROM:<> rejection from other early rejections (e.g.
1777 rejection of HELO). When verifying a recipient using a non-null MAIL
1778 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1779 Whatever the result of the callout, the value of the domain cache is
1780 left unchanged (for any other kind of callout, getting as far as trying
1781 RCPT means that the domain itself is ok).
1783 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1786 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1787 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1788 session, the daemon ignores ECONNECT errors and logs others; it now
1789 ignores EPIPE as well.
1791 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1792 (quoted-printable decoding).
1794 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1795 later the small subsequent patch to fix an introduced bug.
1797 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1799 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1801 PH/08 An error is now given if message_size_limit is specified negative.
1803 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1804 to be given (somewhat) arbitrary names.
1806 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1809 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1810 miscellaneous code fixes
1812 PH/10 Added the log_reject_target ACL modifier to specify where to log
1815 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1816 hostname. This is wrong, because it relates to the incoming message (and
1817 probably the interface on which it is arriving) and not to the outgoing
1818 callout (which could be using a different interface). This has been
1819 changed to use the value of the helo_data option from the smtp transport
1820 instead - this is what is used when a message is actually being sent. If
1821 there is no remote transport (possible with a router that sets up host
1822 addresses), $smtp_active_hostname is used.
1824 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1825 tweaks were necessary in order to get it to work (see also 21 below):
1826 (a) The code assumed that strncpy() returns a negative number on buffer
1827 overflow, which isn't the case. Replaced with Exim's string_format()
1829 (b) There were several signed/unsigned issues. I just did the minimum
1830 hacking in of casts. There is scope for a larger refactoring.
1831 (c) The code used strcasecmp() which is not a standard C function.
1832 Replaced with Exim's strcmpic() function.
1833 (d) The code set only $1; it now sets $auth1 as well.
1834 (e) A simple test gave the error "authentication client didn't specify
1835 service in request". It would seem that Dovecot has changed its
1836 interface. Fortunately there's a specification; I followed it and
1837 changed what the client sends and it appears to be working now.
1839 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1842 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1843 address A is aliased to B and C, where B exists and C does not. Without
1844 -v the output is "A verified" because verification stops after a
1845 successful redirection if more than one address is generated. However,
1846 with -v the child addresses are also verified. Exim was outputting "A
1847 failed to verify" and then showing the successful verification for C,
1848 with its parentage. It now outputs "B failed to verify", showing B's
1849 parentage before showing the successful verification of C.
1851 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1852 look up a TXT record in a specific list after matching in a combined
1855 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1856 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1857 they consult the DNS. I had assumed they would set it the way they
1858 wanted; and indeed my experiments on Linux seem to show that in some
1859 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1860 To be on the safe side, however, I have now made the interface to
1861 host_find_byname() similar to host_find_bydns(), with an argument
1862 containing the DNS resolver options. The host_find_byname() function now
1863 sets these options at its start, just as host_find_bydns() does. The smtp
1864 transport options dns_qualify_single and dns_search_parents are passed to
1865 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1866 of host_find_byname() use the default settings of RES_DEFNAMES
1867 (qualify_single) but not RES_DNSRCH (search_parents).
1869 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1870 spool_read_header() do less string testing, by means of a preliminary
1871 switch on the second character of optional "-foo" lines. (This is
1872 overdue, caused by the large number of possibilities that now exist.
1873 Originally there were few.) While I was there, I also converted the
1874 str(n)cmp tests so they don't re-test the leading "-" and the first
1875 character, in the hope this might squeeze out yet more improvement.
1877 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1878 flag allowing group syntax was set by the header_syntax check but not
1879 turned off, possible causing trouble later; (2) The flag was not being
1880 set at all for the header_verify test, causing "group"-style headers to
1881 be rejected. I have now set it in this case, and also caused header_
1882 verify to ignore an empty address taken from a group. While doing this, I
1883 came across some other cases where the code for allowing group syntax
1884 while scanning a header line wasn't quite right (mostly, not resetting
1885 the flag correctly in the right place). These bugs could have caused
1886 trouble for malformed header lines. I hope it is now all correct.
1888 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1889 with the "reply" argument non-NULL. The code, however (which originally
1890 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1891 but it didn't always do it. This confused somebody who was copying the
1892 code for some other use. I have removed all the tests.
1894 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1895 feature that was used to support insecure browsers during the U.S. crypto
1896 embargo. It requires special client support, and Exim is probably the
1897 only MTA that supported it -- and would never use it because real RSA is
1898 always available. This code has been removed, because it had the bad
1899 effect of slowing Exim down by computing (never used) parameters for the
1900 RSA_EXPORT functionality.
1902 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1903 authenticator to fail if there's a tab character in the incoming data
1904 (there should never be unless someone is messing about, as it's supposed
1905 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1906 if the connection is using TLS or if the remote IP is the same as the
1907 local IP, and the "valid-client-cert option" if a client certificate has
1910 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1911 authenticators. This can be used for authorization after authentication
1912 succeeds. (In the case of plaintext, it servers for both authentication
1915 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1916 if any retry times were supplied.
1918 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1919 connection, where there is no client IP address to check. In this
1920 situation, the verify now always succeeds.
1922 PH/25 Applied John Jetmore's -Mset patch.
1924 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1926 PH/27 In a string expansion for a processed (not raw) header when multiple
1927 headers of the same name were present, leading whitespace was being
1928 removed from all of them, but trailing whitespace was being removed only
1929 from the last one. Now trailing whitespace is removed from each header
1930 before concatenation. Completely empty headers in a concatenation (as
1931 before) are ignored.
1933 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1934 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1936 PH/29 [Removed. This was a change that I later backed out, and forgot to
1937 correct the ChangeLog entry (that I had efficiently created) before
1938 committing the later change.]
1940 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1941 address errors (4xx response to RCPT) over the same connection as other
1942 messages routed to the same hosts. Such deliveries are always "forced",
1943 so retry times are not inspected. This resulted in far too many retries
1944 for the affected addresses. The effect occurred only when there were more
1945 hosts than the hosts_max_try setting in the smtp transport when it had
1946 the 4xx errors. Those hosts that it had tried were not added to the list
1947 of hosts for which the message was waiting, so if all were tried, there
1948 was no problem. Two fixes have been applied:
1950 (i) If there are any address or message errors in an SMTP delivery, none
1951 of the hosts (tried or untried) are now added to the list of hosts
1952 for which the message is waiting, so the message should not be a
1953 candidate for sending over the same connection that was used for a
1954 successful delivery of some other message. This seems entirely
1955 reasonable: after all the message is NOT "waiting for some host".
1956 This is so "obvious" that I'm not sure why it wasn't done
1957 previously. Hope I haven't missed anything, but it can't do any
1958 harm, as the worst effect is to miss an optimization.
1960 (ii) If, despite (i), such a delivery is accidentally attempted, the
1961 routing retry time is respected, so at least it doesn't keep
1962 hammering the server.
1964 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1965 in ${readsocket because some servers need this prod.
1967 PH/32 Added some extra debug output when updating a wait-xxx database.
1969 PH/33 The hint "could be header name not terminated by colon", which has been
1970 given for certain expansion errors for a long time, was not being given
1971 for the ${if def:h_colon_omitted{... case.
1973 PH/34 The spec says: "With one important exception, whenever a domain list is
1974 being scanned, $domain contains the subject domain." There was at least
1975 one case where this was not true.
1977 PH/35 The error "getsockname() failed: connection reset by peer" was being
1978 written to the panic log as well as the main log, but it isn't really
1979 panic-worthy as it just means the connection died rather early on. I have
1980 removed the panic log writing for the ECONNRESET error when getsockname()
1983 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1984 runs only) independently of the message's sender address. This meant
1985 that, if the 4xx error was in fact related to the sender, a different
1986 message to the same recipient with a different sender could confuse
1987 things. In particualar, this can happen when sending to a greylisting
1988 server, but other circumstances could also provoke similar problems.
1989 I have changed the default so that the retry time for these errors is now
1990 based a combination of the sender and recipient addresses. This change
1991 can be overridden by setting address_retry_include_sender=false in the
1994 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1995 remote server are returned as part of bounce messages. This was not
1996 happening for LMTP over a pipe (the lmtp transport), but now it is the
1997 same for both kinds of LMTP.
1999 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2000 in header lines that were in fact CNAMEs. This is no longer the case.
2002 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2003 and queue runs started by the daemon processed all messages. This has
2004 been fixed so that -R and -S can now usefully be given with -q<time>.
2006 PH/40 Import PCRE release 6.7 (fixes some bugs).
2008 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2010 PH/42 Give an error if -q is specified more than once.
2012 PH/43 Renamed the variables $interface_address and $interface_port as
2013 $received_ip_address and $received_port, to make it clear that these
2014 values apply to message reception, and not to the outgoing interface when
2015 a message is delivered. (The old names remain recognized, of course.)
2017 PH/44 There was no timeout on the connect() call when using a Unix domain
2018 socket in the ${readsocket expansion. There now is.
2020 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2021 be meaningful with "accept".
2023 SC/01 Eximstats V1.43
2024 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2026 SC/02 Eximstats V1.44
2027 Use a glob alias rather than an array ref in the generated
2028 parser. This improves both readability and performance.
2030 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2031 Collect SpamAssassin and rejection statistics.
2032 Don't display local sender or destination tables unless
2033 there is data to show.
2034 Added average volumes into the top table text output.
2036 SC/04 Eximstats V1.46
2037 Collect data on the number of addresses (recipients)
2038 as well as the number of messages.
2040 SC/05 Eximstats V1.47
2041 Added 'Message too big' to the list of mail rejection
2042 reasons (thanks to Marco Gaiarin).
2044 SC/06 Eximstats V1.48
2045 Mainlog lines which have GMT offsets and are too short to
2046 have a flag are now skipped.
2048 SC/07 Eximstats V1.49 (Alain Williams)
2049 Added the -emptyok flag.
2051 SC/08 Eximstats V1.50
2052 Fixes for obtaining the IP address from reject messages.
2054 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2055 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2056 whitesspace changes from 4.64-PH/27
2058 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2061 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2062 are found, allow negative numbers in numeric criteria)
2064 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2066 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2067 to match changes made in 4.64-PH/43
2069 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2071 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2072 "domain" is actually an IP address, and force a failure. This locks out
2073 those revolvers/nameservers that support "A-for-A" lookups, in
2074 contravention of the specifications.
2076 PH/48 When a host name was looked up from an IP address, and the subsequent
2077 forward lookup of the name timed out, the host name was left in
2078 $sender_host_name, contrary to the specification.
2080 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2081 restricted to single-key lookups, Exim was not diagnosing an error if
2082 * or *@ was used with a query-style lookup.
2084 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2086 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2087 long ago, but noone interested enough thought of it. Let's just say that
2088 the "1.1" means that there are some new functions that weren't there at
2089 some point in the past.
2091 PH/51 Error processing for expansion failure of helo_data from an smtp
2092 transport during callout processing was broken.
2094 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2095 tested/used via the -bh/-bhc/-bs options.
2097 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2098 bug, fixed in subsequent PCRE releases).
2100 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2101 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2103 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2109 SC/01 Use a glob alias rather than an array ref in eximstats generated
2110 parser. This improves both readability and performance.
2112 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2113 Don't display local sender or destination tables in eximstats unless
2114 there is data to show.
2115 Added average volumes into the eximstats top table text output.
2117 SC/03 Collect data on the number of addresses (recipients) as well
2118 as the number of messages in eximstats.
2120 TF/01 Correct an error in the documentation for the redirect router. Exim
2121 does (usually) call initgroups() when daemonizing.
2123 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2124 with consistent privilege compared to when running as a daemon.
2126 TF/03 Note in the spec that $authenticated_id is not set for local
2127 submissions from trusted users.
2129 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2130 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2132 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2133 by adding some example configuration directives to the default
2134 configuration file. A little bit of work is required to uncomment the
2135 directives and define how usernames and passwords are checked, but
2136 there is now a framework to start from.
2138 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2139 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2140 without this. I don't know how relevant this is to other LDAP libraries.
2142 PH/02 Add the verb name to the "unknown ACL verb" error.
2144 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2146 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2148 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2149 directory not expanded when it should be if an expanded home directory
2150 was set for the address (which is overridden by the transport).
2152 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2155 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2156 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2157 because it is too late at that time, and has no effect.
2159 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2160 security issue with \' (bugzilla #107). I could not use the
2161 PQescapeStringConn() function, because it needs a PGconn value as one of
2164 PH/08 When testing addresses using -bt, indicate those final addresses that
2165 are duplicates that would not cause an additional delivery. At least one
2166 person was confused, thinking that -bt output corresponded to deliveries.
2167 (Suppressing duplicates isn't a good idea as you lose the information
2168 about possibly different redirections that led to the duplicates.)
2170 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2171 systems where poll() doesn't work, in particular OS X.
2173 PH/10 Added more information to debugging output for retry time not reached.
2175 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2176 operations in malware.c.
2178 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2181 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2182 syslog_duplication set false, log lines that would normally be written
2183 both the the main log and to the reject log were not written to syslog at
2186 PH/14 In the default configuration, change the use of "message" in ACL warn
2187 statements to "add_header".
2189 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2190 not followed by a command (e.g. "seen endif").
2192 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2193 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2196 PH/17 Added extra conditions to the default value of delay_warning_condition
2200 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2201 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2202 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2205 The Auto-Submitted: and various List- headers are standardised, whereas I
2206 don't think Precedence: ever was.
2208 PH/18 Refactored debugging code in route_finduser() to show more information,
2209 in particular, the error code if getpwnam() issues one.
2211 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2212 This is apparently needed in addition to the PH/07 change above to avoid
2213 any possible encoding problems.
2215 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2216 but not after initializing Perl.
2218 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2219 output them only if debugging. By default they are written stderr,
2220 apparently, which is not desirable.
2222 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2225 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2228 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2230 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2231 authenticated or an ident call has been made. Suppress the default
2232 values for $authenticated_id and $authenticated_sender (but permit -oMai
2233 and -oMas) when testing with -bh.
2235 PH/24 Re-jigged the order of the tests in the default configuration so that the
2236 tests for valid domains and recipients precede the DNS black list and CSA
2237 tests, on the grounds that those ones are more expensive.
2239 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2240 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2241 This bug exists in every version of Exim that I still have, right back to
2244 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2245 However, an attempt to turn on case-sensitivity in a regex key by
2246 including (?-i) didn't work because the subject string was already
2247 lowercased, and the effects were non-intuitive. It turns out that a
2248 one-line patch can be used to allow (?-i) to work as expected.
2254 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2255 other effects) broke the use of negated acl sub-conditions.
2257 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2260 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2261 "Deny" causes Exim to reject the incoming connection with a 554 error.
2262 Unfortunately, if there is a major crisis, such as a disk failure,
2263 tcp-wrappers gives "deny", whereas what one would like would be some
2264 kind of temporary error. A kludge has been added to help with this.
2265 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2266 554 error is used if errno is still zero or contains ENOENT (which occurs
2267 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2270 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2272 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2273 errors. Otherwise a message that provokes a temporary error (when other
2274 messages do not) can cause a whole host to time out.
2276 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2277 addresses were routed directly to files or pipes from a redirect router.
2278 File deliveries just didn't batch; pipe deliveries might have suffered
2281 PH/06 A failure to get a lock for a hints database would erroneously always say
2282 "Failed to get write lock", even when it was really a read lock.
2284 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2285 of 0600. This has been changed to use the value of the lockfile_mode
2286 option (which defaults to 0600).
2288 PH/08 Applied small patch from the Sieve maintainer.
2290 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2291 folder from quota calculations, a direct delivery into this folder messed
2292 up the contents of the maildirsize file. This was because the regex was
2293 used only to exclude .Trash (or whatever) when the size of the mailbox
2294 was calculated. There was no check that a delivery was happening into an
2295 excluded directory. This bug has been fixed by ignoring all quota
2296 processing for deliveries into excluded directories.
2298 PH/10 Added the maildirfolder_create_regex option to appendfile.
2304 PH/01 The code for finding all the local interface addresses on a FreeBSD
2305 system running IPv6 was broken. This may well have applied to all BSD
2306 systems, as well as to others that have similar system calls. The broken
2307 code found IPv4 interfaces correctly, but gave incorrect values for the
2308 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2309 that it would not match correctly against @[] and not recognize the IPv6
2312 PH/02 The ipliteral router was not recognizing addresses of the form user@
2313 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2315 PH/03 Added disable_ipv6.
2317 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2318 decoded form, because it is most often used to construct To: headers
2319 lines in autoreplies, and the decoded form may well be syntactically
2320 invalid. However, $reply_address has leading white space removed, and all
2321 newlines turned into spaces so that the autoreply transport does not
2324 PH/05 If group was specified without a user on a router, and no group or user
2325 was specified on a transport, the group from the router was ignored.
2327 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2328 for visible compile-time settings that can be used to change these
2329 numbers, for those that want even more. Backwards compatibility with old
2330 spool files has been maintained. However, going back to a previous Exim
2331 release will lost any variables that are in spool files.
2333 PH/07 Two small changes when running in the test harness: increase delay when
2334 passing a TCP/IP connection to a new process, in case the original
2335 process has to generate a bounce, and remove special handling of
2336 127.0.0.2 (sic), which is no longer necessary.
2338 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2339 be the same on different OS.
2341 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2344 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2345 whether --show-vars was specified or not
2347 JJ/02 exipick: Added support for new ACL variable spool format introduced
2350 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2351 syntactically invalid From: or Reply-to: line, and a filter used this to
2352 generate an autoreply, and therefore failed to obtain an address for the
2353 autoreply, Exim could try to deliver to a non-existent relative file
2354 name, causing unrelated and misleading errors. What now happens is that
2355 it logs this as a hard delivery error, but does not attempt to create a
2358 PH/11 The exinext utility has a -C option for testing purposes, but although
2359 the given file was scanned by exinext itself; it wasn't being passed on
2360 when Exim was called.
2362 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2363 an end-of-file indication when reading a command response.
2365 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2366 compiled. In many other places in Exim, IPv6 addresses are always
2367 recognized, so I have changed this. It also means that IPv4 domain
2368 literals of the form [IPV4:n.n.n.n] are now always recognized.
2370 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2371 used if the router is not running as root, for example, when verifying at
2372 ACL time, or when using -bh. The debugging output from this situation was
2373 non-existent - all you got was a failure to exec. I have made two
2376 (a) Failures to set uid/gid, the current directory, or a process leader
2377 in a subprocess such as that created by queryprogram now generate
2378 suitable debugging ouput when -d is set.
2380 (b) The queryprogram router detects when it is not running as root,
2381 outputs suitable debugging information if -d is set, and then runs
2382 the subprocess without attempting to change uid/gid.
2384 PH/15 Minor change to Makefile for building test_host (undocumented testing
2387 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2388 additional section of a DNS packet that returns MX or SRV records.
2389 Instead, it always explicitly searches for A/AAAA records. This avoids
2390 major problems that occur when a DNS server includes only records of one
2391 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2392 fixed another bug: if SRV records were looked up and the corresponding
2393 address records were *not* found in the additional section, the port
2394 values from the SRV records were lost.
2396 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2397 using the correct key (the original address) when searching the retry
2398 rules in order to find which one to use for generating the retry hint.
2400 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2401 adding the default one. Similarly, if it contains a Reply-To: header, the
2402 errors_reply_to option, if set, is not used.
2404 PH/19 When calculating a retry time, Exim used to measure the "time since
2405 failure" by looking at the "first failed" field in the retry record. Now
2406 it does not use this if it is later than than the arrival time of the
2407 message. Instead it uses the arrival time. This makes for better
2408 behaviour in cases where some deliveries succeed, thus re-setting the
2409 "first failed" field. An example is a quota failure for a huge message
2410 when small messages continue to be delivered. Without this change, the
2411 "time since failure" will always be short, possible causing more frequent
2412 delivery attempts for the huge message than are intended.
2413 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2415 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2416 $1, $2, $3) because the numerical variables can be reset during some
2417 expansion items (e.g. "match"), thereby losing the authentication data.
2419 PH/21 Make -bV show the size of off_t variables so that the test suite can
2420 decide whether to run tests for quotas > 2G.
2422 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2423 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2424 filecount value is greater than 2G or if a quota value is greater than 2G
2425 on a system where the size of off_t is not greater than 4, a panic error
2428 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2429 never match. The debug and -bh output now contains an explicit error
2430 message indicating a malformed IPv4 address or mask.
2432 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2433 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2434 PH/23 above applies.
2436 PH/25 Do not write to syslog when running in the test harness. The only
2437 occasion when this arises is a failure to open the main or panic logs
2438 (for which there is an explicit test).
2440 PH/26 Added the /no_tell option to "control=freeze".
2442 PH/27 If a host name lookup failed very early in a connection, for example, if
2443 the IP address matched host_lookup and the reverse lookup yielded a name
2444 that did not have a forward lookup, an error message of the form "no IP
2445 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2446 could be logged. Now it outputs the IP address instead of "NULL".
2448 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2449 allows the sender and the authenticated sender to be set when
2450 submitting a message from within Exim. Since child_open_exim() is
2451 documented for local_scan(), the new function should be too.
2453 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2454 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2455 results in an empty string is now treated as unset.
2457 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2459 PH/31 Added sender_verify_fail logging option.
2461 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2462 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2463 tidied the source and removed it altogether.
2465 PH/33 When a queue run was abandoned because the load average was too high, a
2466 log line was always written; now it is written only if the queue_run log
2467 selector is set. In addition, the log line for abandonment now contains
2468 information about the queue run such as the pid. This is always present
2469 in "start" and "stop" lines but was omitted from the "abandon" line.
2471 PH/34 Omit spaces between a header name and the colon in the error message that
2472 is given when verify = headers_syntax fails (if there are lots of them,
2473 the message gets confusing).
2475 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2476 names, as there are now some PTR records that contain slashes. This check
2477 is only to protect against broken name servers that fall over on strange
2478 characters, so the fact that it applies to all lookups doesn't matter.
2480 PH/36 Now that the new test suite is complete, we can remove some of the
2481 special code in Exim that was needed for the old test suite. For example,
2482 sorting DNS records because real resolvers return them in an arbitrary
2483 order. The new test suite's fake resolver always returns records in the
2486 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2487 bounces) except when queue_only is set, to avoid logging races between
2488 the different processes.
2490 PH/38 Panic-die if .include specifies a non-absolute path.
2492 PH/39 A tweak to the "H" retry rule from its user.
2494 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2495 a label. They prevented compilation on older perls.
2497 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2498 a warning to be raised on newish perls.
2500 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2501 on queue. Changes to match documented behaviour of showing count of
2502 messages matching specified criteria.
2504 PH/40 Changed the default ident timeout from 30s to 5s.
2506 PH/41 Added support for the use of login_cap features, on those BSD systems
2507 that have them, for controlling the resources used by pipe deliveries.
2509 PH/42 The content-scanning code uses fopen() to create files in which to put
2510 message data. Previously it was not paying any attention to the mode of
2511 the files. Exim runs with umask(0) because the rest of the code creates
2512 files with open(), and sets the required mode explicitly. Thus, these
2513 files were ending up world-writeable. This was not a big issue, because,
2514 being within the spool directory, they were not world-accessible. I have
2515 created a function called modefopen, which takes an additional mode
2516 argument. It sets umask(777), creates the file, chmods it to the required
2517 mode, then resets the umask. All the relevant calls to fopen() in the
2518 content scanning code have been changed to use this function.
2520 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2521 to 24 hours. This avoids potential overflow problems when processing G
2522 and H retry rules. I suspect nobody ever tinkers with this value.
2524 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2526 PH/45 When the plaintext authenticator is running as a client, the server's
2527 challenges are checked to ensure they are valid base64 strings. By
2528 default, the authentication attempt is cancelled if an invalid string is
2529 received. Setting client_ignore_invalid_base64 true ignores these errors.
2530 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2531 they are received. Thus, the responses can be made to depend on the
2532 challenges. If an invalid string is ignored, an empty string is placed in
2535 PH/46 Messages that are created by the autoreply transport now contains a
2536 References: header, in accordance with RFCs 2822 and 3834.
2538 PH/47 Added authenticated_sender_force to the smtp transport.
2540 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2542 PH/49 Installed latest patch from the Sieve maintainer.
2544 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2545 was also set, the appendfile transport was unnecessarily scanning a
2546 directory of message files (e.g. for maildir delivery) to find the count
2547 of files (along with the size), even though it did not need this
2548 information. It now does the scan only if it needs to find either the
2549 size of the count of files.
2551 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2553 PH/52 Two bugs concerned with error handling when the smtp transport is
2556 (i) Exim was not creating retry information for temporary errors given
2557 for individual recipients after the DATA command when the smtp transport
2558 was used in LMTP mode. This meant that they could be retried too
2559 frequently, and not timed out correctly.
2561 (ii) Exim was setting the flag that allows error details to be returned
2562 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2563 recipients that were returned after the DATA command.
2565 PH/53 This is related to PH/52, but is more general: for any failing address,
2566 when detailed error information was permitted to be returned to the
2567 sender, but the error was temporary, then after the final timeout, only
2568 "retry timeout exceeded" was returned. Now it returns the full error as
2569 well as "retry timeout exceeded".
2571 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2572 do this, and (what is worse) MTAs that accept it.
2574 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2575 will now be deprecated.
2577 PH/56 New os.c-cygwin from the Cygwin maintainer.
2579 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2580 formats (previously only available in exim formats via -bpr, -bpru,
2581 and -bpra. Now also available in native and exiqgrep formats)
2583 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2584 with very large, slow to parse queues
2586 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2588 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2590 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2591 responses to authentication challenges, though it was showing the
2592 challenges; (ii) I've removed the CR characters from the debug output for
2595 PH/58 Allow for the insertion of a newline as well as a space when a string
2596 is turned into more than one encoded-word during RFC 2047 encoding. The
2597 Sieve code now uses this.
2599 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2600 data_4xx, lost_connection, tls_required.
2602 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2603 message was being sent as an SMTP response.
2605 PH/61 Add -l and -k options to exicyclog.
2607 PH/62 When verifying, if an address was redirected to one new address, so that
2608 verification continued, and the new address failed or deferred after
2609 having set something in $address_data, the value of $address_data was not
2610 passed back to the ACL. This was different to the case when no
2611 redirection occurred. The value is now passed back in both cases.
2613 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2614 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2615 use login_cap.h, so on its own it isn't the distinguishing feature. The
2616 new name refers directly to the setclassresources() function.
2618 PH/65 Added configuration files for NetBSD3.
2620 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2622 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2623 is preferred over IPv4.
2625 PH/68 The bounce_return_message and bounce_return_body options were not being
2626 honoured for bounces generated during the reception of non-SMTP messages.
2627 In particular, this applied to messages rejected by the ACL. This bug has
2628 been fixed. However, if bounce_return_message is true and bounce_return_
2629 body is false, the headers that are returned for a non-SMTP message
2630 include only those that have been read before the error was detected.
2631 (In the case of an ACL rejection, they have all been read.)
2633 PH/69 The HTML version of the specification is now built in a directory called
2634 spec_html instead of spec.html, because the latter looks like a path with
2635 a MIME-type, and this confuses some software.
2637 PH/70 Catch two compiler warnings in sieve.c.
2639 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2640 function verify_get_ident() calls ip_connect() to connect a socket, but
2641 if the "connect()" function timed out, ip_connect() used to close the
2642 socket. However, verify_get_ident() also closes the socket later, and in
2643 between Exim writes to the log, which may get opened at this point. When
2644 the socket was closed in ip_connect(), the log could get the same file
2645 descriptor number as the socket. This naturally causes chaos. The fix is
2646 not to close the socket in ip_connect(); the socket should be closed by
2647 the function that creates it. There was only one place in the code where
2648 this was missing, in the iplookup router, which I don't think anybody now
2649 uses, but I've fixed it anyway.
2651 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2652 well as to direct DNS lookups. Otherwise the handling of names in host
2653 lists is inconsistent and therefore confusing.
2659 PH/01 Two changes to the default runtime configuration:
2661 (1) Move the checks for relay_from_hosts and authenticated clients from
2662 after to before the (commented out) DNS black list checks.
2664 (2) Add control=submission to the relay_from_hosts and authenticated
2665 clients checks, on the grounds that messages accepted by these
2666 statements are most likely to be submissions.
2668 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2670 (1) Generate an error if the third argument for the ${prvs expansion is
2673 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2676 (3) Reset the variables that are obtained from the first argument of
2677 ${prvscheck and used in the second argument before leaving the code,
2678 because their memory is reclaimed, so using them afterwards may do
2681 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2682 one (it's much easier than Tom thought :-).
2684 (5) Because of (4), we can now allow for the use of $prvscheck_result
2685 inside the third argument.
2687 PH/03 For some reason, the default setting of PATH when running a command from
2688 a pipe transport was just "/usr/bin". I have changed it to
2691 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2692 anything to be listed in the output from -bV.
2694 PH/05 When a filter generated an autoreply, the entire To: header line was
2695 quoted in the delivery log line, like this:
2697 => >A.N.Other <ano@some.domain> <original@ddress> ...
2699 This has been changed so that it extracts the operative address. There
2700 may be more than one such address. If so, they are comma-separated, like
2703 => >ano@some.domain,ona@other.domain <original@ddress> ...
2705 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2706 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2707 not being looked up in the rDNS to get a host name, Exim was showing the
2708 IP address twice in Received: lines, even though the IP addresses were
2709 identical. For example:
2711 Received: from [1.2.3.4] (helo=[1.2.3.4])
2713 However, if the real host name was known, it was omitting the HELO data
2714 if it matched the actual IP address. This has been tidied up so that it
2715 doesn't show the same IP address twice.
2717 PH/07 When both +timestamp and +memory debugging was on, the value given by
2718 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2719 called by the debug printing, thereby overwriting the timestamp buffer.
2720 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2722 PH/08 When the original message was included in an autoreply transport, it
2723 always said "this is a copy of the message, including all the headers",
2724 even if body_only or headers_only was set. It now gives an appropriate
2727 PH/09 Applied a patch from the Sieve maintainer which:
2729 o fixes some comments
2730 o adds the (disabled) notify extension core
2731 o adds some debug output for the result of if/elsif tests
2732 o points to the current vacation draft in the documentation
2733 and documents the missing references header update
2737 o fixes a bug in processing the envelope test (when testing
2738 multiple envelope elements, the last element determinted the
2741 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2742 Electronic Mail") by including:
2744 Auto-submitted: auto-generated
2746 in the messages that it generates (bounce messages and others, such as
2747 warnings). In the case of bounce messages for non-SMTP mesages, there was
2748 also a typo: it was using "Auto_submitted" (underscore instead of
2749 hyphen). Since every message generated by Exim is necessarily in response
2750 to another message, thes have all been changed to:
2752 Auto-Submitted: auto-replied
2754 in accordance with these statements in the RFC:
2756 The auto-replied keyword:
2758 - SHOULD be used on messages sent in direct response to another
2759 message by an automatic process,
2761 - MUST NOT be used on manually-generated messages,
2763 - MAY be used on Delivery Status Notifications (DSNs) and Message
2764 Disposition Notifications (MDNs),
2766 - MUST NOT be used on messages generated by automatic or periodic
2767 processes, except for messages which are automatic responses to
2770 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2771 to the default Received: header definition.
2773 PH/12 Added log selector acl_warn_skipped (default on).
2775 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2776 variables because (a) they are in the wrong storage pool and (b) even if
2777 they were copied, it wouldn't work properly because of the caching.
2779 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2780 checking when decoding. Apparently there are clients that generate
2781 overlong encoded strings. Why am I not surprised?
2783 PH/15 If the first argument of "${if match_address" was not empty, but did not
2784 contain an "@" character, Exim crashed. Now it writes a panic log message
2785 and treats the condition as false.
2787 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2789 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2790 extension "envelope-auth". The code is finished and in agreement with
2791 other implementations, but there is no documentation so far and in fact,
2792 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2793 not changing the active code.
2795 Print executed "if" and "elsif" statements when debugging is used. This
2796 helps a great deal to understand what a filter does.
2798 Document more things not specified clearly in RFC3028. I had all this
2799 sorted out, when out of a sudden new issues came to my mind. Oops."
2801 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2804 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2805 canonical form (as documented). However, after a host name lookup from
2806 the IP address, check_host() was doing a simple string comparison with
2807 addresses acquired from the DNS when checking that the found name did
2808 have the original IP as one of its addresses. Since any found IPv6
2809 addresses are likely to be in abbreviated form, the comparison could
2810 fail. Luckily, there already exists a function for doing the comparison
2811 by converting both addresses to binary, so now that is used instead of
2812 the text comparison.
2814 PH/20 There was another similar case to PH/19, when a complete host name was
2815 given in a host list; looking up its IP address could give an abbreviated
2816 form, whereas the current host's name might or might not be abbreviated.
2817 The same fix has been applied.
2823 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2824 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2827 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2828 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2830 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2832 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2833 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2834 Also, for "Auto-Submitted", treat anything other than "no" as
2835 non-personal, in accordance with RFC 3834. (Previously it treated
2836 anything starting "auto-" as non-personal.)
2838 TF/01 The control=submission/name=... option had a problem with syntax
2839 errors if the name included a slash character. The /name= option
2840 now slurps the rest of the string, so it can include any characters
2841 but it must come last in the list of options (after /sender_retain
2844 PH/05 Some modifications to the interface to the fake nameserver for the new
2852 TK/01 Added the "success_on_redirect" address verification option. See
2853 NewStuff for rationale and an example.
2855 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2857 PH/02 Patch to exigrep to allow it to work on syslog lines.
2859 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2860 fread() to skip over the body file's header line, because in Cygwin the
2861 header line is locked and is inaccessible.
2863 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2864 co-exist for some time) to make it clear that it is the Exim ID that is
2865 referenced, not the Message-ID: header line.
2867 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2868 string_format() function, because snprintf() does not exist on all
2871 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2872 ${stat: expansion item.
2874 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2875 protocol synchronization error", to keep the pedants happy.
2877 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2878 well as for IRIX systems, when gcc is being used. See the host.c source
2881 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2883 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2886 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2887 transport and to the smtp transport in LMTP mode.
2889 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2891 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2892 run a filter in a subprocess. This could lead to confusion in subsequent
2893 lookups in the parent process. There should also be a search_tidyup() at
2894 the end of the subprocess.
2896 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2897 only if the host matched helo_try_verify_hosts, which caused the
2898 verification to occur when the EHLO/HELO command was issued. The ACL just
2899 tested the remembered result. Now, if a previous verification attempt has
2900 not happened, "verify = helo" does it there and then.
2902 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2904 TK/03 Fix log output including CR from clamd.
2906 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2907 exist provoked a memory error which could cause a segfault.
2909 PH/15 Installed PCRE 6.2
2911 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2913 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2914 of the problem. Specifically, suggested +O2 rather than +O1 for the
2917 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2919 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2920 record, the port was not correctly logged when the outgoing_port log
2921 selector was set (it logged the transort's default port).
2923 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2924 fallback_hosts, and "hosts" in the smtp transport.
2926 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2927 host errors such as "Connection refused".
2929 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2930 authentication with radiusclient 0.4.9:
2932 - Error returned from rc_read_config was caught wrongly
2933 - Username/password not passed on to radius server due to wrong length.
2935 The presumption is that some radiusclient API changes for 4.51/PH/17
2936 were not taken care of correctly. The code is still untested by me (my
2937 Linux distribution still has 0.3.2 of radiusclient), but it was
2938 contributed by a Radius user.
2940 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2941 expanding the "port" option of the smtp transport.
2943 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2944 while reading a MIME header. Thanks to Tom Hughes for a patch.
2946 PH/24 Include config.h inside local_scan.h so that configuration settings are
2949 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2950 that in an ACL for RCPT (for example), you can examine exactly what was
2953 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2954 commands, but it was not correctly comparing the address with the actual
2955 client host address. Thus, it would show the EHLO address in Received:
2956 header lines when this was not necessary.
2958 PH/27 Added the % operator to ${eval:}.
2960 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2961 it should be ignoring failures (because with -C, for example, it has lost
2962 privilege). It wasn't ignoring creation failures other than "already
2965 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2968 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2969 because an input line was too long, either on its own, or by virtue of
2970 too many continuations, the temporary file was not being removed, and the
2971 return code was incorrect.
2973 PH/31 Missing "BOOL" in function definition in filtertest.c.
2975 PH/32 Applied Sieve patches from the maintainer.
2977 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2979 PH/33 Added "verify = not_blind".
2981 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2982 Local/Makefile (with some defaults set). These are used in built scripts
2983 such as exicyclog, but they have never been used in the exim_install
2984 script (though there are many overriding facilities there). I have
2985 arranged that the exim_install script now takes note of these two
2988 PH/35 Installed configuration files for Dragonfly.
2990 PH/36 When a locally submitted message by a trusted user did not contain a
2991 From: header, and the sender address was obtained from -f or from an SMTP
2992 MAIL command, and the trusted user did not use -F to supply a sender
2993 name, $originator_name was incorrectly used when constructing a From:
2994 header. Furthermore, $originator_name was used for submission mode
2995 messages from external hosts without From: headers in a similar way,
2996 which is clearly wrong.
2998 PH/37 Added control=suppress_local_fixups.
3000 PH/38 When log_selector = +received_sender was set, and the addition of the
3001 sender made the log line's construction buffer exactly full, or one byte
3002 less than full, an overflow happened when the terminating "\n" was
3005 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3006 when the result of a list match is failure because a DNS lookup failed.
3008 PH/40 RM_COMMAND is now used in the building process.
3010 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3011 the "build-* directories that it finds.
3013 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3014 address in a domain literal was a prefix of an interface address.
3016 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3017 when verifying a sender address, unless rewrite_headers is false.
3019 PH/44 Wrote a long comment about why errors_to addresses are verified as
3020 recipients, not senders.
3022 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3023 the ratelimit ACL was added.
3025 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3027 PH/46 Added extra information about PostgreSQL errors to the error string.
3029 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3030 avoiding the need to install special zones in a real server. This is
3031 backwards compatible; if it can't find the fake resolver, it drops back.
3032 Thus, both old and new test suites can be run.
3034 TF/02 Added util/ratelimit.pl
3036 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3037 clock is set back in time.
3039 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3040 Candler <B.Candler@pobox.com>.
3042 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3043 OK for addresses that are the result of redirections.
3045 PH/48 A number of further additions for the benefit of the new test suite,
3046 including a fake gethostbyname() that interfaces to the fake DNS resolver
3049 TF/06 The fix for widen_domains has also been applied to qualify_single and
3050 search_parents which are the other dnslookup options that can cause
3053 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3056 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3058 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3059 service" if it encountered an MX record with an empty target hostname.
3060 The message is now "an MX or SRV record indicated no SMTP service".
3062 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3063 if the DNS of the sending site is misconfigured. This is quite a
3064 common situation. This change restores the behaviour of treating a
3065 helo verification defer as a failure.
3067 PH/51 If self=fail was set on a router, the bounce message did not include the
3068 actual error message.
3074 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3076 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3077 command itself ended in error, the underlying message about the transport
3078 filter timeout was being overwritten with the pipe command error. Now the
3079 underlying error message should be appended to the second error message.
3081 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3083 PH/02 Reduce the amount of output that "make" produces by default. Full output
3084 can still be requested.
3086 PH/03 The warning log line about a condition test deferring for a "warn" verb
3087 was being output only once per connection, rather than after each
3088 occurrence (because it was using the same function as for successful
3089 "warn" verbs). This seems wrong, so I have changed it.
3091 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3092 it should not have, which might have caused a crash in the right
3093 circumstances, but probably never did.
3095 PH/04 Installed a modified version of Tony Finch's patch to make submission
3096 mode fix the return path as well as the Sender: header line, and to
3097 add a /name= option so that you can make the user's friendly name appear
3100 TF/03 Added the control = fakedefer ACL modifier.
3102 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3103 Mark Lowes for thorough testing.
3105 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3107 TK/03 Merged latest SRS patch from Miles Wilton.
3109 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3110 with the definition in sysexits.h (which is #included earlier).
3111 Fortunately, Exim does not actually use EX_OK. The code used to try to
3112 preserve the sysexits.h value, by assumimg that macro definitions were
3113 scanned for macro replacements. I have been disabused of this notion,
3114 so now the code just undefines EX_OK before #including unistd.h.
3116 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3117 in the smtp transport. When a block could not be written in a single
3118 write() function, the timeout was being re-applied to each part-write.
3119 This seems wrong - if the receiver was accepting one byte at a time it
3120 would take for ever. The timeout is now adjusted when this happens. It
3121 doesn't have to be particularly precise.
3123 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3124 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3126 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3127 without a domain if the check to <postmaster@domain> fails.
3129 SC/01 Eximstats: added -xls and the ability to specify output files
3130 (patch written by Frank Heydlauf).
3132 SC/02 Eximstats: use FileHandles for outputing results.
3134 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3136 SC/04 Eximstats: fixed display of large numbers with -nvr option
3138 SC/05 Eximstats: fixed merging of reports with empty tables.
3140 SC/06 Eximstats: added the -include_original_destination flag
3142 SC/07 Eximstats: removed tabs and trailing whitespace.
3144 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3146 TK/06 MBOX spool code: Add real "From " MBOX separator line
3147 so the .eml file is really in mbox format (even though
3148 most programs do not really care). Patch from Alex Miller.
3150 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3151 The latter is generated from $received_to and is only set if the
3152 message has one envelope recipient. SA can use these headers,
3153 obviously out-of-the-box. Patch from Alex Miller.
3155 PH/08 The ${def test on a variable was returning false if the variable's
3156 value was "0", contrary to what the specification has always said!
3157 The result should be true unless the variable is empty.
3159 PH/09 The syntax error of a character other than { following "${if
3160 def:variable_name" (after optional whitespace) was not being diagnosed.
3161 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3162 accidental colon was present, for example, could give incorrect results.
3164 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3165 result is used (not including appendfile, where other changes are about
3168 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3169 This involved changing a lot of size variables from int to off_t. It
3170 should work with maildirs and everything.
3172 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3173 spamd dying while we are connected to it.
3175 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3178 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3179 function for BDB 4.3.
3181 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3183 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3184 the sources for PCRE have been re-arranged and more files are now
3187 PH/15 The code I had for printing potentially long long variables in PH/11
3188 above was not the best (it lost precision). The length of off_t variables
3189 is now inspected at build time, and an appropriate printing format (%ld
3190 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3191 to be "long long int" or "long int". This is needed for the internal
3192 formatting function string_vformat().
3194 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3195 the configuration file to be ":syslog", then the script "guesses" where
3196 the logs files are, rather than using the compiled in default. In our
3197 case the guess is not the same as the compiled default, so the script
3198 suddenly stopped working when I started to use syslog. The patch checks
3199 to see if log_file_path is "". If so, it attempts to read it from exim
3200 with no configuration file to get the compiled in version, before it
3201 falls back to the previous guessing code."
3203 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3204 implementing BATV in an Exim configuration. See NewStuff for the gory
3207 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3208 Makefile that are specific to HP-UX.
3210 PH/18 If the "use_postmaster" option was set for a recipient callout together
3211 with the "random" option, the postmaster address was used as the MAIL
3212 FROM address for the random test, but not for the subsequent recipient
3213 test. It is now used for both.
3215 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3216 patch removes a few documentation additions to RFC 3028, because the
3217 latest draft now contains them. It adds the new en;ascii-case comparator
3218 and a new error check for 8bit text in MIME parts. Comparator and
3219 require names are now matched exactly. I enabled the subaddress
3220 extension, but it is not well tested yet (read: it works for me)."
3222 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3223 rework some of the code of TK/09 above to avoid the hardwired use of
3224 "%lld" and "long long". Replaced the call to snprintf() with a call to
3227 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3228 records point to non-existent hosts", "retry timeout exceeded", and
3229 "retry time not reached for any host after a long failure period".
3231 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3232 experimental DomainKeys support:
3234 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3235 (2) On an error such as an illegally used "control", the wrong name for
3236 the control was given.
3238 These problems did NOT occur unless DomainKeys support was compiled.
3240 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3242 PH/24 Added ${if match_ip condition.
3244 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3245 databases so that it will be absolutely obvious if a crash occurs in the
3246 DB library. This is a regular occurrence (often caused by mis-matched
3249 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3250 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3251 picked up on a user's system that detects such things. There doesn't seem
3252 to be a gcc warning option for this - only an attribute that has to be
3253 put on the function's prototype. It seems that in Fedora Core 4 they have
3254 set this on a number of new functions. No doubt there will be more in due
3257 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3258 specify a transport. However, if an address that was verified by such a
3259 router was the subject of a callout, Exim crashed because it tried to
3260 read the rcpt_include_affixes from the non-existent transport. Now it
3261 just assumes that the setting of that option is false. This bug was
3262 introduced by 4.51/PH/31.
3264 PH/28 Changed -d+all to exclude +memory, because that information is very
3265 rarely of interest, but it makes the output a lot bigger. People tend to
3266 do -d+all out of habit.
3268 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3269 code in os-type was giving problems when libc.so lives in lib64, like on
3272 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3273 aren't the modern standard, and it seems that some systems' include files
3274 don't always have them. Exim was already checking for some of the newer
3275 ones like T_AAAA, and defining it itself. I've added checks for all the
3276 record types that Exim uses.
3278 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3279 not automatically generating a new one, as it is supposed to. This
3280 prevented TLS from working. If the file did exist, but contained invalid
3281 data, a new version was generated, as expected. It was only the case of a
3282 non-existent file that was broken.
3284 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3285 with a change in libdomainkeys > 0.64.
3287 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3288 from DNS. If the selector record carries the flag, it now has
3289 precedence over the domain-wide flag.
3291 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3293 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3294 the use of a port name that isn't defined in /etc/services) occurred, the
3295 message was deferred as in a normal delivery, and thus remained on the
3296 spool, instead of being failed because of the mua_wrapper setting. This
3297 is now fixed, and I tidied up some of the mua_wrapper messages at the
3300 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3301 the messages in a hash of arrays rather than using individual hashes.
3302 This is a bit cleaner and results in dramatic memory savings, albeit
3303 at a slight CPU cost.
3305 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3306 as requested by Marc Sherman.
3308 SC/10 Eximstats: added histograms for user specified patterns as requested
3311 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3313 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3314 fopen() in the content-scanning modules that did not already have it.
3320 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3321 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3323 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3325 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3327 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3328 to test Sieve filters that use "vacation".
3330 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3331 that changes the way the GnuTLS parameters are stored in the cache file.
3332 The new format can be generated externally. For backward compatibility,
3333 if the data in the cache doesn't make sense, Exim assumes it has read an
3334 old-format file, and it generates new data and writes a new file. This
3335 means that you can't go back to an older release without removing the
3338 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3339 work if there are any delivery delays because "one_time" forces the
3340 parent to be marked "delivered", so its unseen clone is never tried
3341 again. For this reason, Exim now forbids the simultaneous setting of
3344 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3345 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3346 entry if you want to know the details. The fix, however, neglected to
3347 consider the case when local delivery batching is involved. The test for
3348 "previously delivered" was not happening when checking to see if an
3349 address could be batched with a previous (undelivered) one; under
3350 certain circumstances this could lead to multiple deliveries to the same
3353 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3354 in its include files, and this causes problems building Exim.
3356 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3357 header_syntax) but Exim was just ignoring anything given after a slash.
3358 In particular, this caused confusion with an attempt to use "verify =
3359 reverse_host_lookup/defer_ok". An error is now given when options are
3360 supplied for verify items that do not have them. (Maybe reverse_host_
3361 lookup should have a defer_ok option, but that's a different point.)
3363 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3364 defined by RFC 821) to 2048, because there were problems with some AUTH
3365 commands, and RFC 1869 says the size should be increased for extended
3366 SMTP commands that take arguments.
3368 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3371 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3372 "unknown" error; now it says that the functionality isn't in the binary.
3374 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3375 an address' error message when a string expansion fails (syntax or
3376 whatever). Otherwise the password may appear in the log. Following change
3377 PH/42 below, there is no longer a chance of it appearing in a bounce
3380 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3382 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3383 of its addresses was ever tried. (Bugzilla bug #2.)
3385 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3386 the result incorrectly in the debug output. (It correctly added a newline
3387 to what was transported.)
3389 TF/01 Added $received_time.
3391 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3392 commented out examples of how to interface to a virus scanner and to
3393 SpamAssassin. Also added commented examples of av_scanner and
3394 spamd_address settings.
3396 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3397 and controls are allowed in which ACLs. There were a couple of minor
3398 errors. Some of the entries in the conditions table (which is a table of
3399 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3400 negation of where the condition IS allowed.
3402 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3404 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3405 header file does not have a version number, so I've had to invent a new
3406 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3407 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3408 radiusclient), but it was contributed by a Radius user.
3410 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3411 files or directories, for OpenSSL.
3413 PH/19 When an Exim process that is running as root has to create an Exim log
3414 file, it does so in a subprocess that runs as exim:exim so as to get the
3415 ownership right at creation (otherwise, other Exim processes might see
3416 the file with the wrong ownership). There was no test for failure of this
3417 fork() call, which would lead to the process getting stuck as it waited
3418 for a non-existent subprocess. Forks do occasionally fail when resources
3419 run out. I reviewed all the other calls to fork(); they all seem to check
3422 PH/20 When checking for unexpected SMTP input at connect time (before writing
3423 the banner), Exim was not dealing correctly with a non-positive return
3424 from the read() function. If the client had disconnected by this time,
3425 the result was a log entry for a synchronization error with an empty
3426 string after "input=" when read() returned zero. If read() returned -1
3427 (an event I could not check), uninitialized data bytes were printed.
3428 There were reports of junk text (parts of files, etc) appearing after
3431 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3433 PH/22 Added support for macro redefinition, and (re)definition in between
3434 driver and ACL definitions.
3436 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3437 forgetting to use the resulting value; it was using the unexpanded value.
3439 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3440 hadn't been configured. The fix is from Juergen Kreileder, who
3441 understands it better than I do:
3443 "Here's what I see happening with three configured cyrus_sasl
3444 authenticators configured (plain, login, cram-md5):
3446 On startup auth_cyrus_sasl_init() gets called for each of these.
3447 This means three calls to sasl_listmech() without a specified mech_list.
3448 => SASL tests which mechs of all available mechs actually work
3449 => three warnings about OTP not working
3450 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3452 With the patch, sasl_listmech() also gets called three times. But now
3453 SASL's mech_list option is set to the server_mech specified in the the
3454 authenticator. Or in other words, the answer from sasl_listmech()
3455 gets limited to just the mech you're testing for (which is different
3457 => the return list contains just 'plain' or 'login', 'cram-md5' or
3458 nothing depending on the value of ob->server_mech.
3460 I've just tested the patch: Authentication still works fine,
3461 unavailable mechs specified in the exim configuration are still
3462 caught, and the auth.log warnings about OTP are gone."
3464 PH/25 When debugging is enabled, the contents of the command line are added
3465 to the debugging output, even when log_selector=+arguments is not
3468 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3469 answer is "GNU", and only if the return is "GNU/something" is the answer
3472 PH/27 $acl_verify_message is now set immediately after the failure of a
3473 verification in an ACL, and so is available in subsequent modifiers. In
3474 particular, the message can be preserved by coding like this:
3476 warn !verify = sender
3477 set acl_m0 = $acl_verify_message
3479 Previously, $acl_verify_message was set only while expanding "message"
3480 and "log_message" when a very denied access.
3482 PH/28 Modified OS/os.c-Linux with
3484 -#ifndef OS_LOAD_AVERAGE
3485 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3487 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3488 nomenclature these days.)
3490 PH/29 Installed patch from the Sieve maintainer that adds the options
3491 sieve_useraddress and sieve_subaddress to the redirect router.
3493 PH/30 In these circumstances:
3494 . Two addresses routed to the same list of hosts;
3495 . First host does not offer TLS;
3496 . First host accepts first address;
3497 . First host gives temporary error to second address;
3498 . Second host offers TLS and a TLS session is established;
3499 . Second host accepts second address.
3500 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3501 and peerdn, if requested) that were in fact used only for the second
3504 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3505 attention to any local part prefix or suffix that was matched by the
3506 router that accepted the address. It now behaves in the same way as it
3507 does for delivery: the affixes are removed from the local part unless
3508 rcpt_include_affixes is set on the transport.
3510 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3511 timeout during the DATA phase of an incoming message.
3513 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3514 applied a patch sanctioned by the Sieve maintainer.
3516 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3517 the uid or gid is negative. A case of a negative gid caused this to be
3518 noticed. The fix allows for either to be negative.
3520 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3521 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3524 PH/36 The change PH/12 above was broken. Fixed it.
3526 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3527 the grounds that routing the same address twice would always produce the
3528 same answer. This might have been true once, but it is certainly no
3529 longer true now. Routing a child address may depend on the previous
3530 routing that produced that child. Some complicated redirection strategies
3531 went wrong when messages had multiple recipients, and made Exim's
3532 behaviour dependent on the order in which the addresses were given.
3534 I have moved the duplicate checking until after the routing is complete.
3535 Exim scans the addresses that are assigned to local and remote
3536 transports, and removes any duplicates. This means that more work will be
3537 done, as duplicates will always all be routed, but duplicates are
3538 presumably rare, so I don't expect this is of any significance.
3540 For deliveries to pipes, files, and autoreplies, the duplicate checking
3541 still happens during the routing process, since they are not going to be
3544 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3545 It corrects a timeout issue with spamd. This is Ian's comment: "The
3546 background is that sometimes spamd either never reads data from a
3547 connection it has accepted, or it never writes response data. The exiscan
3548 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3549 blindly assumes that writes won't block so it may never time out."
3551 PH/39 Allow G after quota size as well as K and M.
3553 PH/40 The value set for $authenticated_id in an authenticator may not contain
3554 binary zeroes or newlines because the value is written to log lines and
3555 to spool files. There was no check on this. Now the value is run through
3556 the string_printing() function so that such characters are converted to
3557 printable escape sequences.
3559 PH/41 $message_linecount is a new variable that contains the total number of
3560 lines in the message. Compare $body_linecount, which is the count for the
3563 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3564 bounce and delay warning messages, except in certain special cases, which
3567 (a) An SMTP error message from a remote host;
3568 (b) A message specified in a :fail: redirection;
3569 (c) A message specified in a "fail" command in a system filter;
3570 (d) A message specified in a FAIL return from the queryprogram router;
3571 (e) A message specified by the cannot_route_message router option.
3573 In these cases only, Exim does include the error details in bounce and
3574 warning messages. There are also a few cases where bland messages such
3575 as "unrouteable address" or "local delivery error" are given.
3577 PH/43 $value is now also set for the "else" part of a ${run expansion.
3579 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3580 being worked on, but at least Exim now implements the latest version to
3583 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3584 process to complete was treated as a delivery failure, a timeout while
3585 writing the message to the pipe was logged, but erroneously treated as a
3586 successful delivery. Such timeouts include transport filter timeouts. For
3587 consistency with the overall process timeout, these timeouts are now
3588 treated as errors, giving rise to delivery failures by default. However,
3589 there is now a new Boolean option for the pipe transport called
3590 timeout_defer, which, if set TRUE, converts the failures into defers for
3591 both kinds of timeout. A transport filter timeout is now identified in
3594 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3595 systems where "make" and "gmake" are different, calling "gmake" at top
3596 level broke things. I've arranged for the value of $(MAKE) to be passed
3597 from the Makefile to this script so that it can call the same version of
3601 A note about Exim versions 4.44 and 4.50
3602 ----------------------------------------
3604 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3605 changes of various kinds. As a consequence, a big documentation update was
3606 needed. This delayed the release for rather longer than seemed good, especially
3607 in the light of a couple of (minor) security issues. Therefore, the changes
3608 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3609 release. So 4.44 and 4.50 are in effect two different branches that both start
3612 I have left the 4.50 change log unchanged; it contains all the changes since
3613 4.43. The change log for 4.44 is below; many of its items are identical to
3614 those for 4.50. This seems to be the most sensible way to preserve the
3615 historical information.
3621 1. Minor wording change to the doc/README.SIEVE file.
3623 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3624 computation of the current number of files was incorrect.
3626 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3627 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3630 4. Give more explanation in the error message when the command for a transport
3631 filter fails to execute.
3633 5. There are several places where Exim runs a non-Exim command in a
3634 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3635 was being done only for the command run by the queryprogram router. It is
3636 now done for all such subprocesses. The other cases are: ${run, transport
3637 filters, and the commands run by the lmtp and pipe transports.
3639 6. Added CONFIGURE_GROUP build-time option.
3641 7. Some older OS have a limit of 256 on the maximum number of file
3642 descriptors. Exim was using setrlimit() to set 1000 as a large value
3643 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3644 systems. I've change it so that if it can't get 1000, it tries for 256.
3646 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3647 was an oversight, and furthermore, ever since the addition of extra
3648 controls (e.g. 4.43/32), the checks on when to allow different forms of
3649 "control" were broken. There should now be diagnostics for all cases when a
3650 control that does not make sense is encountered.
3652 9. Added the /retain_sender option to "control=submission".
3654 10. $recipients is now available in the predata ACL (oversight).
3656 11. Tidy the search cache before the fork to do a delivery from a message
3657 received from the command line. Otherwise the child will trigger a lookup
3658 failure and thereby defer the delivery if it tries to use (for example) a
3659 cached ldap connection that the parent has called unbind on.
3661 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3662 of $address_data from the recipient verification was clobbered by the
3663 sender verification.
3665 13. The value of address_data from a sender verification is now available in
3666 $sender_address_data in subsequent conditions in the ACL statement.
3668 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3670 15. Added a new option "connect=<time>" to callout options, to set a different
3673 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3674 was its contents. (It was OK if the option was not defined at all.)
3676 17. A "Completed" log line is now written for messages that are removed from
3677 the spool by the -Mrm option.
3679 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3680 information about exactly what failed.
3682 19. Added -dd to debug only the daemon process.
3684 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3685 handles timeouts, both on the server side and network timeouts. Renamed the
3686 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3688 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3689 It is now set to "smtps".
3691 22. $host_address is now set to the target address during the checking of
3692 ignore_target_hosts.
3694 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3695 being passed; this would have caused $sender_host_name to have been used if
3696 matching the list had actually called for a host name (not very likely,
3697 since this list is usually IP addresses). A host name is now passed as
3700 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3701 code that specifies a non-restarting handler (typically sigaction() in
3702 modern systems) in an attempt to fix a rare and obscure crash bug.
3704 25. Narrowed the window for a race in the daemon that could cause it to ignore
3705 SIGCHLD signals. This is not a major problem, because they are used only to
3706 wake it up if nothing else does.
3708 26. A malformed maildirsize file could cause Exim to calculate negative values
3709 for the mailbox size or file count. Odd effects could occur as a result.
3710 The maildirsize information is now recalculated if the size or filecount
3713 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3714 support for a long time. Removed HAVE_SYS_VFS_H.
3716 28. Installed the latest version of exipick from John Jetmore.
3718 29. In an address list, if the pattern was not a regular expression, an empty
3719 subject address (from a bounce message) matched only if the pattern was an
3720 empty string. Non-empty patterns were not even tested. This was the wrong
3721 because it is perfectly reasonable to use an empty address as part of a
3722 database query. An empty address is now tested by patterns that are
3723 lookups. However, all the other forms of pattern expect the subject to
3724 contain a local part and a domain, and therefore, for them, an empty
3725 address still always fails if the pattern is not itself empty.
3727 30. Exim went into a mad DNS loop when attempting to do a callout where the
3728 host was specified on an smtp transport, and looking it up yielded more
3729 than one IP address.
3731 31. Re-factored the code for checking spool and log partition space into a
3732 function that finds that data and another that does the check. The former
3733 is then used to implement four new variables: $spool_space, $log_space,
3734 $spool_inodes, and $log_inodes.
3736 32. The RFC2047 encoding function was originally intended for short strings
3737 such as real names; it was not keeping to the 75-character limit for
3738 encoded words that the RFC imposes. It now respects the limit, and
3739 generates multiple encoded words if necessary. To be on the safe side, I
3740 have increased the buffer size for the ${rfc2047: expansion operator from
3743 33. It is now permitted to omit both strings after an "if" condition; if the
3744 condition is true, the result is "true". As before, when the second string
3745 is omitted, a false condition yields an empty string. This makes it less
3746 cumbersome to write custom ACL and router conditions.
3748 34. Failure to deliver a bounce message always caused it to be frozen, even if
3749 there was an errors_to setting on the router. The errors_to setting is now
3752 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3753 canonical form (fully expanded) before being placed in
3754 $sender_host_address.
3756 36. The table in the code that translates DNS record types into text (T_A to
3757 "A" for instance) was missing entries for NS and CNAME. It is just possible
3758 that this could have caused confusion if both these types were looked up
3759 for the same domain, because the text type is used as part of Exim's
3760 per-process caching. But the chance of anyone hitting this buglet seems
3763 37. The dnsdb lookup has been extended in a number of ways.
3765 (1) There is a new type, "zns", which walks up the domain tree until it
3766 finds some nameserver records. It should be used with care.
3768 (2) There is a new type, "mxh", which is like "mx" except that it returns
3769 just the host names, not the priorities.
3771 (3) It is now possible to give a list of domains (or IP addresses) to be
3772 looked up. The behaviour when one of the lookups defers can be
3773 controlled by a keyword.
3775 (4) It is now possible to specify the separator character for use when
3776 multiple records are returned.
3778 38. The dnslists ACL condition has been extended: it is now possible to supply
3779 a list of IP addresses and/or domains to be looked up in a particular DNS
3782 39. Added log_selector=+queue_time_overall.
3784 40. When running the queue in the test harness, wait just a tad after forking a
3785 delivery process, to get repeatability of debugging output.
3787 41. Include certificate and key file names in error message when GnuTLS fails
3788 to set them up, because the GnuTLS error message doesn't include the name
3789 of the failing file when there is a problem reading it.
3791 42. Allow both -bf and -bF in the same test run.
3793 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3795 44. The "Exiscan patch" is now merged into the mainline Exim source.
3797 45. Sometimes the final signoff response after QUIT could fail to get
3798 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3799 < 0 before doing a fflush(). This bug looks as though it goes back to the
3800 introduction of TLS in release 3.20, but "sometimes" must have been rare
3801 because the tests only now provoked it.
3803 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3804 (this can affect the format of dates).
3806 47. exim_tidydb, when checking for the continued existence of a message for
3807 which it has found a message-specific retry record, was not finding
3808 messages that were in split spool directories. Consequently, it was
3809 deleting retry records that should have stayed in existence.
3811 48. Steve fixed some bugs in eximstats.
3813 49. The SPA authentication driver was not abandoning authentication and moving
3814 on to the next authenticator when an expansion was forced to fail,
3815 contradicting the general specification for all authenticators. Instead it
3816 was generating a temporary error. It now behaves as specified.
3818 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3819 (the order specifies the preference for clients). The order is now AES256,
3820 AES128, 3DES, ARCFOUR128.
3822 51. Small patch to Sieve code - explicitly set From: when generating an
3825 52. Exim crashed if a remote delivery caused a very long error message to be
3826 recorded - for instance if somebody sent an entire SpamAssassin report back
3827 as a large number of 550 error lines. This bug was coincidentally fixed by
3828 increasing the size of one of Exim's internal buffers (big_buffer) that
3829 happened as part of the Exiscan merge. However, to be on the safe side, I
3830 have made the code more robust (and fixed the comments that describe what
3833 53. Now that there can be additional text after "Completed" in log lines (if
3834 the queue_time_overall log selector is set), a one-byte patch to exigrep
3835 was needed to allow it to recognize "Completed" as not the last thing in
3838 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3839 patch that reportedly fixes this has been added. I am not expert enough to
3840 create a test for it. This is what the patch creator wrote:
3842 "I found a little strange behaviour of ldap code when working with
3843 Windows 2003 AD Domain, where users was placed in more than one
3844 Organization Units. When I tried to give exim partial DN, the exit code
3845 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3846 But simultaneously result of request was absolutely normal ldap result,
3847 so I produce this patch..."
3849 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3850 REFERENCE, so I have modified the code to exclude the patch when that macro
3853 55. Some experimental protocols are using DNS PTR records for new purposes. The
3854 keys for these records are domain names, not reversed IP addresses. The
3855 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3856 leaves it alone. Component reversal etc. now happens only for IP addresses.
3859 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3861 57. Double the size of the debug message buffer (to 2048) so that more of very
3862 long debug lines gets shown.
3864 58. The exicyclog utility now does better if the number of log files to keep
3865 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3867 59. Two changes related to the smtp_active_hostname option:
3869 (1) $smtp_active_hostname is now available as a variable.
3870 (2) The default for smtp_banner uses $smtp_active_hostname instead
3871 of $primary_hostname.
3873 60. The host_aton() function is supposed to be passed a string that is known
3874 to be a valid IP address. However, in the case of IPv6 addresses, it was
3875 not checking this. This is a hostage to fortune. Exim now panics and dies
3876 if the condition is not met. A case was found where this could be provoked
3877 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3878 components; fortuitously, this particular loophole had already been fixed
3879 by change 4.50/55 above.
3881 If there are any other similar loopholes, the new check in host_aton()
3882 itself should stop them being exploited. The report I received stated that
3883 data on the command line could provoke the exploit when Exim was running as
3884 exim, but did not say which command line option was involved. All I could
3885 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3886 running as the user.
3889 61. There was a buffer overflow vulnerability in the SPA authentication code
3890 (which came originally from the Samba project). I have added a test to the
3891 spa_base64_to_bits() function which I hope fixes it.
3894 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3895 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3897 63. The daemon start-up calls getloadavg() while still root for those OS that
3898 need the first call to be done as root, but it missed one case: when
3899 deliver_queue_load_max is set with deliver_drop_privilege. This is
3900 necessary for the benefit of the queue runner, because there is no re-exec
3901 when deliver_drop_privilege is set.
3903 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3904 This has been fixed.
3906 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3907 was in use, was not putting the data itself into the right store pool;
3908 consequently, it could be overwritten for a subsequent message in the same
3909 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3912 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3914 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3915 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3916 the function were treating the return as a boolean value, which happened to
3917 work because 0=false and not-0=true, but is not correct code.
3919 68. The host_aton() function was not handling scoped IPv6 addresses (those
3920 with, for example, "%eth0" on the end) correctly.
3922 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3923 negated items (that is, ~something) in unsigned ints. Some compilers
3924 apparently mutter when there is no cast.
3926 70. If an address verification called from an ACL failed, and did not produce a
3927 user-specific message (i.e. there was only a "system" message), nothing was
3928 put in $acl_verify_message. In this situation, it now puts the system
3931 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3932 session; change 4.31/43 added the unwanted input to the log line - except
3933 that it did not do this in the start of session case. It now does.
3935 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3936 This is wrong and can cause the other end to generate a synchronization
3937 error if it is another Exim or anything else that does the synchronization
3938 check. A QUIT command is no longer sent after a timeout.
3940 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3941 during host lookups.
3943 74. The defer_ok option of callout verification was not working if it was used
3944 when verifying addresses in header lines, that is, for this case:
3946 verify = header_sender/callout=defer_ok
3948 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3949 those file descriptors could be used for SMTP connections. If anything
3950 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3951 could be sent to the SMTP client, causing chaos. The daemon now opens
3952 stdin, stdout, and stderr to /dev/null when it puts itself into the
3955 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3956 log by default. The user can override this with suitable Perl magic.
3958 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3959 the log message when discard triggers, was not working for the DATA ACL or
3960 for the non-SMTP ACL.
3962 78. Error message wording change in sieve.c.
3964 79. If smtp_accept_max_per_host was set, the number of connections could be
3965 restricted to fewer than expected, because the daemon was trying to set up
3966 a new connection before checking whether the processes handling previous
3967 connections had finished. The check for completed processes is now done
3968 earlier. On busy systems, this bug wouldn't be noticed because something
3969 else would have woken the daemon, and it would have reaped the completed
3972 80. If a message was submitted locally by a user whose login name contained one
3973 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3974 It caused a spool format error. I have fixed the spool reading code. A
3975 related problem was that the "from" clause in the Received: line became
3976 illegal because of the space(s). It is now covered by ${quote_local_part.
3978 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3981 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3983 83. Patch from Sieve maintainer to fix unterminated string problem in
3984 "vacation" handling.
3986 84. Some minor changes to the Linux configuration files to help with other
3987 OS variants using glibc.
3989 85. One more patch for Sieve to update vacation handling to latest spec.
3992 ----------------------------------------------------
3993 See the note above about the 4.44 and 4.50 releases.
3994 ----------------------------------------------------
4000 1. Change 4.43/35 introduced a bug that caused file counts to be
4001 incorrectly computed when quota_filecount was set in an appendfile
4004 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4005 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4008 3. Give more explanation in the error message when the command for a transport
4009 filter fails to execute.
4011 4. There are several places where Exim runs a non-Exim command in a
4012 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4013 was being done only for the command run by the queryprogram router. It is
4014 now done for all such subprocesses. The other cases are: ${run, transport
4015 filters, and the commands run by the lmtp and pipe transports.
4017 5. Some older OS have a limit of 256 on the maximum number of file
4018 descriptors. Exim was using setrlimit() to set 1000 as a large value
4019 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4020 systems. I've change it so that if it can't get 1000, it tries for 256.
4022 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4023 was an oversight, and furthermore, ever since the addition of extra
4024 controls (e.g. 4.43/32), the checks on when to allow different forms of
4025 "control" were broken. There should now be diagnostics for all cases when a
4026 control that does not make sense is encountered.
4028 7. $recipients is now available in the predata ACL (oversight).
4030 8. Tidy the search cache before the fork to do a delivery from a message
4031 received from the command line. Otherwise the child will trigger a lookup
4032 failure and thereby defer the delivery if it tries to use (for example) a
4033 cached ldap connection that the parent has called unbind on.
4035 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4036 of $address_data from the recipient verification was clobbered by the
4037 sender verification.
4039 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4040 was its contents. (It was OK if the option was not defined at all.)
4042 11. A "Completed" log line is now written for messages that are removed from
4043 the spool by the -Mrm option.
4045 12. $host_address is now set to the target address during the checking of
4046 ignore_target_hosts.
4048 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4049 being passed; this would have caused $sender_host_name to have been used if
4050 matching the list had actually called for a host name (not very likely,
4051 since this list is usually IP addresses). A host name is now passed as
4054 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4055 code that specifies a non-restarting handler (typically sigaction() in
4056 modern systems) in an attempt to fix a rare and obscure crash bug.
4058 15. Narrowed the window for a race in the daemon that could cause it to ignore
4059 SIGCHLD signals. This is not a major problem, because they are used only to
4060 wake it up if nothing else does.
4062 16. A malformed maildirsize file could cause Exim to calculate negative values
4063 for the mailbox size or file count. Odd effects could occur as a result.
4064 The maildirsize information is now recalculated if the size or filecount
4067 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4068 support for a long time. Removed HAVE_SYS_VFS_H.
4070 18. Updated exipick to current release from John Jetmore.
4072 19. Allow an empty sender to be matched against a lookup in an address list.
4073 Previously the only cases considered were a regular expression, or an
4076 20. Exim went into a mad DNS lookup loop when doing a callout where the
4077 host was specified on the transport, if the DNS lookup yielded more than
4080 21. The RFC2047 encoding function was originally intended for short strings
4081 such as real names; it was not keeping to the 75-character limit for
4082 encoded words that the RFC imposes. It now respects the limit, and
4083 generates multiple encoded words if necessary. To be on the safe side, I
4084 have increased the buffer size for the ${rfc2047: expansion operator from
4087 22. Failure to deliver a bounce message always caused it to be frozen, even if
4088 there was an errors_to setting on the router. The errors_to setting is now
4091 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4092 canonical form (fully expanded) before being placed in
4093 $sender_host_address.
4095 24. Updated eximstats to version 1.33
4097 25. Include certificate and key file names in error message when GnuTLS fails
4098 to set them up, because the GnuTLS error message doesn't include the name
4099 of the failing file when there is a problem reading it.
4101 26. Expand error message when OpenSSL has problems setting up cert/key files.
4104 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4105 (this can affect the format of dates).
4107 28. exim_tidydb, when checking for the continued existence of a message for
4108 which it has found a message-specific retry record, was not finding
4109 messages that were in split spool directories. Consequently, it was
4110 deleting retry records that should have stayed in existence.
4112 29. eximstats updated to version 1.35
4113 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4114 1.35 - bugfix such that pie charts by volume are generated correctly
4116 30. The SPA authentication driver was not abandoning authentication and moving
4117 on to the next authenticator when an expansion was forced to fail,
4118 contradicting the general specification for all authenticators. Instead it
4119 was generating a temporary error. It now behaves as specified.
4121 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4122 (the order specifies the preference for clients). The order is now AES256,
4123 AES128, 3DES, ARCFOUR128.
4125 31. Small patch to Sieve code - explicitly set From: when generating an
4128 32. Exim crashed if a remote delivery caused a very long error message to be
4129 recorded - for instance if somebody sent an entire SpamAssassin report back
4130 as a large number of 550 error lines. This bug was coincidentally fixed by
4131 increasing the size of one of Exim's internal buffers (big_buffer) that
4132 happened as part of the Exiscan merge. However, to be on the safe side, I
4133 have made the code more robust (and fixed the comments that describe what
4136 33. Some experimental protocols are using DNS PTR records for new purposes. The
4137 keys for these records are domain names, not reversed IP addresses. The
4138 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4139 leaves it alone. Component reversal etc. now happens only for IP addresses.
4142 34. The host_aton() function is supposed to be passed a string that is known
4143 to be a valid IP address. However, in the case of IPv6 addresses, it was
4144 not checking this. This is a hostage to fortune. Exim now panics and dies
4145 if the condition is not met. A case was found where this could be provoked
4146 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4147 components; fortuitously, this particular loophole had already been fixed
4148 by change 4.50/55 or 4.44/33 above.
4150 If there are any other similar loopholes, the new check in host_aton()
4151 itself should stop them being exploited. The report I received stated that
4152 data on the command line could provoke the exploit when Exim was running as
4153 exim, but did not say which command line option was involved. All I could
4154 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4155 running as the user.
4158 35. There was a buffer overflow vulnerability in the SPA authentication code
4159 (which came originally from the Samba project). I have added a test to the
4160 spa_base64_to_bits() function which I hope fixes it.
4163 36. The daemon start-up calls getloadavg() while still root for those OS that
4164 need the first call to be done as root, but it missed one case: when
4165 deliver_queue_load_max is set with deliver_drop_privilege. This is
4166 necessary for the benefit of the queue runner, because there is no re-exec
4167 when deliver_drop_privilege is set.
4169 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4170 was in use, was not putting the data itself into the right store pool;
4171 consequently, it could be overwritten for a subsequent message in the same
4172 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4175 38. Sometimes the final signoff response after QUIT could fail to get
4176 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4177 < 0 before doing a fflush(). This bug looks as though it goes back to the
4178 introduction of TLS in release 3.20, but "sometimes" must have been rare
4179 because the tests only now provoked it.
4185 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4186 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4187 Now it returns an integer. A number of calls were still expecting a T/F
4188 return. Fortuitously, in all cases, the tests worked in OK situations,
4189 which is the norm. However, things would have gone wrong on any write
4190 failures on the smtp file descriptor. This function is used when sending
4191 messages over SMTP and also when doing verify callouts.
4193 2. When Exim is called to do synchronous delivery of a locally submitted
4194 message (the -odf or -odi options), it no longer closes stderr before doing
4197 3. Implemented the mua_wrapper option.
4199 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4201 5. Implemented the functions header_remove(), header_testname(),
4202 header_add_at_position(), and receive_remove_recipient(), and exported them
4205 6. If an ACL "warn" statement specified the addition of headers, Exim already
4206 inserted X-ACL-Warn: at the start if there was no header name. However, it
4207 was not making this test for the second and subsequent header lines if
4208 there were newlines in the string. This meant that an invalid header could
4209 be inserted if Exim was badly configured.
4211 7. Allow an ACL "warn" statement to add header lines at the start or after all
4212 the Received: headers, as well as at the end.
4214 8. Added the rcpt_4xx retry error code.
4216 9. Added postmaster_mailfrom=xxx to callout verification option.
4218 10. Added mailfrom=xxxx to the callout verification option, for verify=
4221 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4222 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4224 12. Inserted some casts to stop certain compilers warning when using pointer
4225 differences as field lengths or precisions in printf-type calls (mostly
4226 affecting debugging statements).
4228 13. Added optional readline() support for -be (dynamically loaded).
4230 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4231 same clock tick as a message's arrival, so that its received time was the
4232 same as the "first fail" time on the retry record, and that message
4233 remained on the queue past the ultimate address timeout, every queue runner
4234 would try a delivery (because it was past the ultimate address timeout) but
4235 after another failure, the ultimate address timeout, which should have then
4236 bounced the address, did not kick in. This was a "< instead of <=" error;
4237 in most cases the first failure would have been in the next clock tick
4238 after the received time, and all would be well.
4240 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4241 being recognized when the domain list was tested by the match_domain
4242 condition in an expansion string.
4244 16. Added the ${str2b64: operator.
4246 17. Exim was always calling setrlimit() to set a large limit for the number of
4247 processes, without checking whether the existing limit was already
4248 adequate. (It did check for the limit on file descriptors.) Furthermore,
4249 errors from getrlimit() and setrlimit() were being ignored. Now they are
4250 logged to the main and panic logs, but Exim does carry on, to try to do its
4251 job under whatever limits there are.
4253 18. Imported PCRE 5.0.
4255 19. Trivial typo in log message " temporarily refused connection" (the leading
4258 20. If the log selector return_path_on_delivery was set and an address was
4259 redirected to /dev/null, the delivery process crashed because it assumed
4260 that a return path would always be set for a "successful" delivery. In this
4261 case, the whole delivery is bypassed as an optimization, and therefore no
4264 21. Internal re-arrangement: the function for sending a challenge and reading
4265 a response while authentication was assuming a zero-terminated challenge
4266 string. It's now changed to take a pointer and a length, to allow for
4267 binary data in such strings.
4269 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4271 23. Exim was not respecting finduser_retries when seeking the login of the
4272 uid under which it was called; it was always trying 10 times. (The default
4273 setting of finduser_retries is zero.) Also, it was sleeping after the final
4274 failure, which is pointless.
4276 24. Implemented tls_on_connect_ports.
4278 25. Implemented acl_smtp_predata.
4280 26. If the domain in control=submission is set empty, Exim assumes that the
4281 authenticated id is a complete email address when it generates From: or
4282 Sender: header lines.
4284 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4285 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4286 chown and chgrp in /bin and hostname in /usr/bin.
4288 28. Exim was keeping the "process log" file open after each use, just as it
4289 does for the main log. This opens the possibility of it remaining open for
4290 long periods when the USR1 signal hits a daemon. Occasional processlog
4291 errors were reported, that could have been caused by this. Anyway, it seems
4292 much more sensible not to leave this file open at all, so that is what now
4295 29. The long-running daemon process does not normally write to the log once it
4296 has entered its main loop, and it closes the log before doing so. This is
4297 so that log files can straightforwardly be renamed and moved. However,
4298 there are a couple of unusual error situations where the daemon does write
4299 log entries, and I had neglected to close the log afterwards.
4301 30. The text of an SMTP error response that was received during a remote
4302 delivery was being truncated at 512 bytes. This is too short for some of
4303 the long messages that one sometimes sees. I've increased the limit to
4306 31. It is now possible to make retry rules that apply only when a message has a
4307 specific sender, in particular, an empty sender.
4309 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4310 it possible to be selective about when SMTP synchronization is enforced.
4312 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4314 32. Implemented hosts_connection_nolog.
4316 33. Added an ACL for QUIT.
4318 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4321 35. Added mailbox_size and mailbox_filecount to appendfile.
4323 36. Added control = no_multiline_responses to ACLs.
4325 37. There was a bug in the logic of the code that waits for the clock to tick
4326 in the case where the clock went backwards by a substantial amount such
4327 that the microsecond fraction of "now" was more than the microsecond
4328 fraction of "then" (but the whole seconds number was less).
4330 38. Added support for the libradius Radius client library this is found on
4331 FreeBSD (previously only the radiusclient library was supported).
4337 1. When certain lookups returned multiple values in the form name=value, the
4338 quoting of the values was not always being done properly. Specifically:
4339 (a) If the value started with a double quote, but contained no whitespace,
4341 (b) If the value contained whitespace other than a space character (i.e.
4342 tabs or newlines or carriage returns) it was not quoted.
4343 This fix has been applied to the mysql and pgsql lookups by writing a
4344 separate quoting function and calling it from the lookup code. The fix
4345 should probably also be applied to nisplus, ibase and oracle lookups, but
4346 since I cannot test any of those, I have not disturbed their existing code.
4348 2. A hit in the callout cache for a specific address caused a log line with no
4349 reason for rejecting RCPT. Now it says "Previous (cached) callout
4350 verification failure".
4352 3. There was an off-by-one bug in the queryprogram router. An over-long
4353 return line was truncated at 256 instead of 255 characters, thereby
4354 overflowing its buffer with the terminating zero. As well as fixing this, I
4355 have increased the buffer size to 1024 (and made a note to document this).
4357 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4358 when Exim is waiting for an SMTP response from a remote server, Exim
4359 restarts its select() call on the socket, thereby resetting its timeout.
4360 This is not a problem when such interrupts are rare. Somebody set up a cron
4361 job to run exiwhat every 2 minutes, which is less than the normal select()
4362 timeout (5 or 10 minutes). This meant that the select() timeout never
4363 kicked in because it was always reset. I have fixed this by comparing the
4364 time when an interrupt arrives with the time at the start of the first call
4365 to select(). If more time than the timeout has elapsed, the interrupt is
4366 treated as a timeout.
4368 5. Some internal re-factoring in preparation for the addition of Sieve
4369 extensions (by MH). In particular, the "personal" test is moved to a
4370 separate function, and given an option for scanning Cc: and Bcc: (which is
4371 not set for Exim filters).
4373 6. When Exim created an email address using the login of the caller as the
4374 local part (e.g. when creating a From: or Sender: header line), it was not
4375 quoting the local part when it contained special characters such as @.
4377 7. Installed new OpenBSD configuration files.
4379 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4380 try to make them clearer.
4382 9. Callout options, other than the timeout value, were being ignored when
4383 verifying sender addresses in header lines. For example, when using
4385 verify = header_sender/callout=no_cache
4387 the cache was (incorrectly) being used.
4389 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4390 only the Cygwin environment.
4392 11. When return_path_on_delivery was set as a log selector, if different remote
4393 addresses in the same message used different return paths and parallel
4394 remote delivery occurred, the wrong values would sometimes be logged.
4395 (Whenever a remote delivery process finished, the return path value from
4396 the most recently started remote delivery process was logged.)
4398 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4399 lines when AUTH and/or TLS are in use. This is the "received protocol"
4400 field. Exim used to use "asmtp" for authenticated SMTP, without any
4401 indication (in the protocol name) for TLS use. Now it follows the RFC and
4402 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4403 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4404 names appear in log lines as well as in Received: header lines.
4406 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4407 extensions, and comparison tests, and to fix some bugs.
4409 14. Changes to the "personal" filter test:
4411 (1) The test was buggy in that it was just doing the equivalent of
4412 "contains" tests on header lines. For example, if a user's address was
4413 anne@some.where, the "personal" test would incorrectly be true for
4415 To: susanne@some.where
4417 This test is now done by extracting each address from the header in turn,
4418 and checking the entire address. Other tests that are part of "personal"
4419 are now done using regular expressions (for example, to check local parts
4420 of addresses in From: header lines).
4422 (2) The list of non-personal local parts in From: addresses has been
4423 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4424 taken from the Sieve specification recommendations.
4426 (3) If the message contains any header line starting with "List-" it is
4427 treated as non-personal.
4429 (4) The test for "circular" in the Subject: header line has been removed
4430 because it now seems ill-conceived.
4432 15. Minor typos in src/EDITME comments corrected.
4434 16. Installed latest exipick from John Jetmore.
4436 17. If headers_add on a router specified a text string that was too long for
4437 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4438 of string_sprintf() is now avoided.
4440 18. $message_body_size was not set (it was always zero) when running the DATA
4441 ACL and the local_scan() function.
4443 19. For the "mail" command in an Exim filter, no default was being set for
4444 the once_repeat time, causing a random time value to be used if "once" was
4445 specified. (If the value happened to be <= 0, no repeat happened.) The
4446 default is now 0s, meaning "never repeat". The "vacation" command was OK
4447 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4448 (I found it when inspecting the code).
4450 20. There is now an overall timeout for performing a callout verification. It
4451 defaults to 4 times the callout timeout, which applies to individual SMTP
4452 commands during the callout. The overall timeout applies when there is more
4453 than one host that can be tried. The timeout is checked before trying the
4454 next host. This prevents very long delays if there are a large number of
4455 hosts and all are timing out (e.g. when the network connections are timing
4456 out). The value of the overall timeout can be changed by specifying an
4457 additional sub-option for "callout", called "maxwait". For example:
4459 verify = sender/callout=5s,maxwait=20s
4461 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4462 to the end before writing, but this should make it even safer).
4464 22. Exim was forgetting that it had advertised PIPELINING for the second and
4465 subsequent messages on an SMTP connection. It was also not resetting its
4466 memory on STARTTLS and an internal HELO.
4468 23. When Exim logs an SMTP synchronization error within a session, it now
4469 records whether PIPELINING has been advertised or not.
4471 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4472 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4473 rather than long int.
4475 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4477 26. Added the never_mail option to autoreply.
4483 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4484 crash if the getsockname() call failed; this can happen if a connection is
4485 closed very soon after it is established. The problem was simply in the
4486 order in which certain operations were done, causing Exim to try to write
4487 to the SMTP stream before it had set up the file descriptor. The bug has
4488 been fixed by making things happen in the correct order.
4494 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4495 before the connection was closed, thus losing the rejection response.
4497 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4498 some early Solaris releases, but causes trouble in current releases where
4499 socklen_t is defined.
4501 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4504 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4507 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4508 that releasing the top of it at the end releases what was used for sub-
4509 expansions (unless the block got too big). However, discard this block if
4510 the first thing is a variable or header, so that we can use its block when
4511 it is dynamic (useful for very large $message_headers, for example).
4513 6. Lookups now cache *every* query, not just the most recent. A new, separate
4514 store pool is used for this. It can be recovered when all lookup caches are
4515 flushed. Lookups now release memory at the end of their result strings.
4516 This has involved some general refactoring of the lookup sources.
4518 7. Some code has been added to the store_xxx() functions to reduce the amount
4519 of flapping under certain conditions.
4521 8. log_incoming_interface used to affect only the <= reception log lines. Now
4522 it causes the local interface and port to be added to several more SMTP log
4523 lines, for example "SMTP connection from", and rejection lines.
4525 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4527 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4529 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4530 could be overwritten at the end of the current message (or the start of a
4531 new message if it was set in a HELO ACL). The value is now preserved for
4532 the duration of the SMTP connection.
4534 12. If a transport had a headers_rewrite setting, and a matching header line
4535 contained an unqualified address, that address was qualified, even if it
4536 did not match any rewriting rules. The underlying bug was that the values
4537 of the flags that permit the existence of unqualified sender and recipient
4538 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4539 non-local messages, and by -bnq for local messages) were not being
4540 preserved with the message after it was received.
4542 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4543 "next input=" as part of the text comprising the host identity instead of
4544 the correct text. The code was using the same buffer for two different
4545 strings. However, depending on which order the printing function evaluated
4546 its arguments, the bug did not always show up. Under Linux, for example, my
4547 test suite worked just fine.
4549 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4550 doesn't work with some older versions of Perl. It has been changed to "my",
4551 which in any case is probably the better facility to use.
4553 15. A really picky compiler found some instances of statements for creating
4554 error messages that either had too many or two few arguments for the format
4557 16. The size of the buffer for calls to the DNS resolver has been increased
4558 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4559 for addresses that have a lot of PTR records. This alleviates a problem; it
4560 does not fully solve it.
4562 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4563 buffer now truncates the list and logs the incident, which is the same
4564 action as happens when Exim is looking up a host name and its aliases.
4565 Previously in this situation something unpredictable would happen;
4566 sometimes it was "internal error: store_reset failed".
4568 18. If a server dropped the connection unexpectedly when an Exim client was
4569 using GnuTLS and trying to read a response, the client delivery process
4570 crashed while trying to generate an error log message.
4572 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4573 string, for example:
4575 warn message = H1: something\nH2: something
4577 the text was added as a single header line from Exim's point of view
4578 though it ended up OK in the delivered message. However, searching for the
4579 second and subsequent header lines using $h_h2: did not work. This has been
4580 fixed. Similarly, if a system filter added multiple headers in this way,
4581 the routers could not see them.
4583 20. Expanded the error message when iplsearch is called with an invalid key to
4584 suggest using net-iplsearch in a host list.
4586 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4587 ACLs are no longer actually imposed (and a message to that effect is
4590 22. If a "gecos" field in a passwd entry contained escaped characters, in
4591 particular, if it contained a \" sequence, Exim got it wrong when building
4592 a From: or a Sender: header from that name. A second bug also caused
4593 incorrect handling when an unquoted " was present following a character
4594 that needed quoting.
4596 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4597 was not being matched caselessly.
4599 24. Arranged for all hyphens in the exim.8 source to be escaped with
4602 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4603 properly. Recipient callout cache records were still being keyed to include
4604 the sender, even when use_sender was set false. This led to far more
4605 callouts that were necessary. The sender is no longer included in the key
4606 when use_sender is false.
4608 26. Added "control = submission" modifier to ACLs.
4610 27. Added the ${base62d: operator to decode base 62 numbers.
4612 28. dnsdb lookups can now access SRV records.
4614 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4615 the configuration file.
4617 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4618 (-v) mode. This makes the output for a verbose queue run more intelligible.
4620 31. Added a use_postmaster feature to recipient callouts.
4622 32. Added the $body_zerocount variable, containing the number of binary zero
4623 bytes in the message body.
4625 33. The time of last modification of the "new" subdirectory is now used as the
4626 "mailbox time last read" when there is a quota error for a maildir
4629 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4631 35. Added +ignore_unknown as a special item in host lists.
4633 36. Code for decoding IPv6 addresses in host lists is now included, even if
4634 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4635 address was recognized as an IP address, but was then not correctly decoded
4636 into binary, causing unexpected and incorrect effects when compared with
4643 1. Very minor rewording of debugging text in manualroute to say "list of
4644 hosts" instead of "hostlist".
4646 2. If verify=header_syntax was set, and a header line with an unqualified
4647 address (no domain) and a large number of spaces between the end of the
4648 name and the colon was received, the reception process suffered a buffer
4649 overflow, and (when I tested it) crashed. This was caused by some obsolete
4650 code that should have been removed. The fix is to remove it!
4652 3. When running in the test harness, delay a bit after writing a bounce
4653 message to get a bit more predictability in the log output.
4655 4. Added a call to search_tidyup() just before forking a reception process. In
4656 theory, someone could use a lookup in the expansion of smtp_accept_max_
4657 per_host which, without the tidyup, could leave open a database connection.
4659 5. Added the variables $recipient_data and $sender_data which get set from a
4660 lookup success in an ACL "recipients" or "senders" condition, or a router
4661 "senders" option, similar to $domain_data and $local_part_data.
4663 6. Moved the writing of debug_print from before to after the "senders" test
4666 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4667 problems for message scanning, either using a data ACL, or using
4668 local_scan() because the Received: header was not generated till after they
4669 were called (in order to set the time as the time of reception completion).
4670 I have revised the way this works. The header is now generated after the
4671 body is received, but before the ACL or local_scan() are called. After they
4672 are run, the timestamp in the header is updated.
4678 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4679 before starting a queue runner without re-exec. This happened only when
4680 deliver_drop_privilege was set or when the Exim user was set to root. The
4681 effect of the bug was that timeouts during subsequent deliveries caused
4682 crashes instead of being properly handled. The handler is now left at its
4683 default (and expected) setting.
4685 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4686 message, again when deliver_drop_privilege is set or Exim is run as root.
4687 The bug described in (1) was not present in this case, but the tidying up
4688 of the other signals was missing. I have made the two cases consistent.
4690 3. The ignore_target_hosts setting on a manualroute router was being ignored
4691 for hosts that were looked up using the /MX notation.
4693 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4696 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4697 operated on the sender address. After changing the $sender_address to <>
4698 for the sender address verify, Exim was re-instated it as the original
4699 (before rewriting) address, but remembering that it had rewritten it, so it
4700 wasn't rewriting it again. This bug also had the effect of breaking the
4701 sender address verification caching when the sender address was rewritten.
4703 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4704 This has been changed so that if the ip literal address matches
4705 ignore_target_hosts, the router declines.
4707 7. Added expansion conditions match_domain, match_address, and match_local_
4708 part (NOT match_host).
4710 8. The placeholder for the Received: header didn't have a length field set.
4712 9. Added code to Exim itself and to exim_lock to test for a specific race
4713 condition that could lead to file corruption when using MBX delivery. The
4714 issue is with the lockfile that is created in /tmp. If this file is removed
4715 after a process has opened it but before that process has acquired a lock,
4716 there is the potential for a second process to recreate the file and also
4717 acquire a lock. This could lead to two Exim processes writing to the file
4718 at the same time. The added code performs the same test as UW imapd; it
4719 checks after acquiring the lock that its file descriptor still refers to
4720 the same named file.
4722 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4723 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4726 11. Added the smtp_active_hostname option. If used, this will typically be made
4727 to depend on the incoming interface address. Because $interface_address is
4728 not set up until the daemon has forked a reception process, error responses
4729 that can happen earlier (such as "too many connections") no longer contain
4732 12. If an expansion in a condition on a "warn" statement fails because a lookup
4733 defers, the "warn" statement is abandoned, and the next ACL statement is
4734 processed. Previously this caused the whole ACL to be aborted.
4736 13. Added the iplsearch lookup type.
4738 14. Added ident_timeout as a log selector.
4740 15. Added tls_certificate_verified as a log selector.
4742 16. Added a global option tls_require_ciphers (compare the smtp transport
4743 option of the same name). This controls incoming TLS connections.
4745 17. I finally figured out how to make tls_require_ciphers do a similar thing
4746 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4747 before starting the TLS session.
4749 18. Tabs are now shown as \t in -bP output.
4751 19. If the log selector return_path_on_delivery was set, Exim crashed when
4752 bouncing a message because it had too many Received: header lines.
4754 20. If two routers both had headers_remove settings, and the first one included
4755 a superfluous trailing colon, the final name in the first list and the
4756 first name in the second list were incorrectly joined into one item (with a
4757 colon in the middle).
4763 1. Added -C and -D options to the exinext utility, mainly to make it easier
4764 to include in the automated testing, but these could be helpful when
4765 multiple configurations are in use.
4767 2. The exinext utility was not formatting the output nicely when there was
4768 an alternate port involved in the retry record key, nor when there was a
4769 message id as well (for retries that were specific to a specific message
4770 and a specific host). It was also confused by IPv6 addresses, because of
4771 the additional colons they contain. I have fixed the IPv4 problem, and
4772 patched it up to do a reasonable job for IPv6.
4774 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4775 delivery, the log line now contains "pipelined" if PIPELINING was used.
4777 4. An SMTP transport process used to panic and die if the bind() call to set
4778 an explicit outgoing interface failed. This has been changed; it is now
4779 treated in the same way as a connect() failure.
4781 5. A reference to $sender_host_name in the part of a conditional expansion
4782 that was being skipped was still causing a DNS lookup. This no longer
4785 6. The def: expansion condition was not recognizing references to header lines
4786 that used bh_ and bheader_.
4788 7. Added the _cache feature to named lists.
4790 8. The code for checking quota_filecount in the appendfile transport was
4791 allowing one more file than it should have been.
4793 9. For compatibility with Sendmail, the command line option
4801 and sets the incoming protocol and host name (for trusted callers). The
4802 host name and its colon can be omitted when only the protocol is to be set.
4803 Note the Exim already has two private options, -pd and -ps, that refer to
4804 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4805 "s", but I don't think that's a major issue.
4807 10. A number of refactoring changes to the code, none of which should affect
4810 (a) The number of logging options was getting close to filling up the
4811 32-bit word that was used as a bit map. I have split them into two classes:
4812 those that are passed in the argument to log_write(), and those that are
4813 only ever tested independently outside of that function. These are now in
4814 separate 32-bit words, so there is plenty of room for expansion again.
4815 There is no change in the user interface or the logging behaviour.
4817 (b) When building, for example, log lines, the code previously used a
4818 macro that called string_cat() twice, in order to add two strings. This is
4819 not really sufficiently general. Furthermore, there was one instance where
4820 it was actually wrong because one of the argument was used twice, and in
4821 one call a function was used. (As it happened, calling the function twice
4822 did not affect the overall behaviour.) The macro has been replaced by a
4823 function that can join an arbitrary number of extra strings onto a growing
4826 (c) The code for expansion conditions now uses a table and a binary chop
4827 instead of a serial search (which was left over from when there were very
4828 few conditions). Also, it now recognizes conditions like "pam" even when
4829 the relevant support is not compiled in: a suitably worded error message is
4830 given if an attempt is made to use such a condition.
4832 11. Added ${time_interval:xxxxx}.
4834 12. A bug was causing one of the ddress fields not to be passed back correctly
4835 from remote delivery subprocesses. The field in question was not being
4836 subsequently used, so this caused to problems in practice.
4838 13. Added new log selectors queue_time and deliver_time.
4840 14. Might have fixed a bug in maildirsizefile handling that threw up
4841 "unexpected character" debug warnings, and recalculated the data
4842 unnecessarily. In any case, I expanded the warning message to give more
4845 15. Added the message "Restricted characters in address" to the statements in
4846 the default ACL that block characters like @ and % in local parts.
4848 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4849 Three changes have been made:
4851 (a) There was a serious bug; a negative response to MAIL caused the whole
4852 recipient domain to be cached as invalid, thereby blocking all messages
4853 to all local parts at the same domain, from all senders. This bug has
4854 been fixed. The domain is no longer cached after a negative response to
4855 MAIL if the sender used is not empty.
4857 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4860 (c) A new callout option, "use_sender" has been added for people who want
4861 the modified behaviour.
4867 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4870 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4871 indeed breaks things for older releases.
4873 3. Added additional logging to the case where there is a problem reading data
4874 from a filter that is running in a subprocess using a pipe, in order to
4875 try to track down a specific problem.
4877 4. Testing facility fudge: when running in the test harness and attempting
4878 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4879 getting "No route to host". Convert this to a timeout.
4881 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4884 6. Some OS don't have socklen_t but use size_t instead. This affects the
4885 fifth argument of getsockopt() amongst other things. This is now
4886 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4887 can be set for individual OS. I have set it for SunOS5, OSF1, and
4888 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4889 some earlier ones do not.
4891 7. Change 4.30/15 was not doing the test caselessly.
4893 8. The standard form for an IPv6 address literal was being rejected by address
4894 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4895 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4896 this, as well as the form without the "IPv6" on the front (but only when
4897 address literals are enabled, of course).
4899 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4901 10. Exim crashed if a message with an empty sender address specified by -f
4902 encountered a router with an errors_to setting. This could be provoked only
4903 by a command such as
4907 where an empty string was supplied; "<>" did not hit this bug.
4909 11. Installed PCRE release 4.5.
4911 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4912 remained set. It is now erased.
4914 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4915 times from message ids (which are base 36 rather than the normal 62).
4917 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4918 were being counted as actual protocol errors, and logged if the log
4919 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4920 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4921 and DATA following a set of rejected RCPTs do not count as protocol errors.
4922 In other words, Exim assumes they were pipelined, though this may not
4923 actually be the case. Of course, in all cases the client gets an
4924 appropriate error code.
4926 15. If a lookup fails in an ACL condition, a message about the failure may
4927 be available; it is used if testing the ACL cannot continue, because most
4928 such messages specify what the cause of the deferral is. However, some
4929 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4930 that caused an old message to be retained and used if a later statement
4931 caused a defer, replacing the real cause of the deferral.
4933 16. If an IP address had so many PTR records that the DNS lookup buffer
4934 was not large enough to hold them, Exim could crash while trying to process
4935 the truncated data. It now detects and logs this case.
4937 17. Further to 4.21/58, another change has been made: if (and only if) the
4938 first line of a message (the first header line) ends with CRLF, a bare LF
4939 in a subsequent header line has a space inserted after it, so as not to
4940 terminate the header.
4942 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4943 unnecessarily, used atoi() instead of strtol(), and didn't check the
4944 termination when getting file sizes from file names by regex.
4946 19. Completely re-implemented the support for maildirsize files, in the light
4947 of a number of problems with the previous contributed implementation
4948 (4.30/29). In particular:
4950 . If the quota is zero, the maildirsize file is maintained, but no quota is
4953 . If the maildir directory does not exist, it is created before any attempt
4954 to write a maildirsize file.
4956 . The quota value in the file is just a cache; if the quota is changed in
4957 the transport, the new value overrides.
4959 . A regular expression is available for excluding directories from the
4962 20. The autoreply transport checks the characters in options that define the
4963 message's headers; it allows continued headers, but it was checking with
4964 isspace() after an embedded newline instead of explicitly looking for a
4967 21. If all the "regular" hosts to which an address was routed had passed their
4968 expiry times, and had not reached their retry times, the address was
4969 bounced, even if fallback hosts were defined. Now Exim should go on to try
4972 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4973 equivalent code in the SMTP transport. Some hosts send humungous responses
4974 to HELO/EHLO, more than 1024 it seems.
4976 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4977 gives compiler warnings in some environments. I've now done it "properly",
4980 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4981 (because of problems with the built-in one) was declared to return uschar *
4982 instead of char *, causing compiler failure.
4984 25. Fixed a file descriptor leak when processing alias/forward files.
4986 26. Fixed a minor format string issue in dbfn.c.
4988 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4990 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4991 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4994 29. When a sender address is verified, it is cached, to save repeating the test
4995 when there is more than one recipient in a message. However, when the
4996 verification involves a callout, it is possible for different callout
4997 options to be set for different recipients. It is too complicated to keep
4998 track of this in the cache, so now Exim always runs a verification when a
4999 callout is required, relying on the callout cache for the optimization.
5000 The overhead is duplication of the address routing, but this should not be
5003 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5004 to be verified with callout=postmaster, and the main callout worked but the
5005 postmaster check failed, the verification correctly failed. However, if a
5006 subsequent RCPT command asked for sender verification *without* the
5007 postmaster check, incorrect caching caused this verification also to fail,
5010 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5011 it was not caching the DNS options (qualify_single, search_parents) that
5012 were used when the lookup failed. A subsequent lookup with different
5013 options therefore always gave the same answer, though there were cases
5014 where it should not have. (Example: a "domains = !$mx_any" option on a
5015 dnslookup router: the "domains" option is always processed without any
5016 widening, but the router might have qualify_single set.) Now Exim uses the
5017 cached value only when the same options are set.
5019 32. Added John Jetmore's "exipick" utility to the distribution.
5021 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5022 than a timeout (e.g. a certificate is required, and is not provided), an
5023 Exim server now closes the connection immediately. Previously it waited for
5024 the client to close - but if the client is SSL, it seems that they each
5025 wait for each other, leading to a delay before one of them times out.
5027 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5028 maintained 0.8.x compatibility because I don't think many are using it, and
5029 it is clearly obsolete.
5031 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5034 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5035 client certificate was expired. A simple patch fixes this, though I don't
5036 understand the full logic of why the verify callback is called multiple
5039 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5040 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5041 which causes problems with some clients (such as the Certicom SSL Plus
5042 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5043 disables the coutermeasure allowing Eudora to connect."
5045 38. Exim was not checking that a write() to a log file succeeded. This could
5046 lead to Bad Things if a log got too big, in particular if it hit a file
5047 size limit. Exim now panics and dies if it cannot write to a log file, just
5048 as it does if it cannot open a log file.
5050 39. Modified OS/Makefile-Linux so that it now contains
5052 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5054 The two -D definitions ensure that Exim is compiled with large file
5055 support, which makes it possible to handle log files that are bigger than
5058 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5059 instance) a domain was checked against a named list that involved a lookup,
5060 causing $domain_data to be set, then another domain was checked against the
5061 same list, then the first domain was re-checked, the value of $domain_data
5062 after the final check could be wrong. In particular, if the second check
5063 failed, it could be set empty. This bug probably also applied to
5066 41. The strip_trailing_dot option was not being applied to the address given
5067 with the -f command-line option.
5069 42. The code for reading a message's header from the spool was incrementing
5070 $received_count, but never initializing it. This meant that the value was
5071 incorrect (doubled) while delivering a message in the same process in which
5072 it was received. In the most common configuration of Exim, this never
5073 happens - a fresh exec is done - but it can happen when
5074 deliver_drop_privilege is set.
5076 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5077 it now includes up to 150 characters of the unexpected data in the log
5080 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5081 and building data strings. The size of both of these buffers was 10 000
5082 bytes - far larger than anybody would *ever* want, thought I. Needless to
5083 say, somebody hit the limit. I have increased the maximum line length to
5084 20 000 and the maximum data length of concatenated lines to 100 000. I have
5085 also fixed two bugs, because there was no checking on these buffers. Tsk,
5086 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5087 buffer is too small.
5089 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5090 lsearch lookups. Now it does.
5092 46. When parsing a route_list item in a manualroute router, a fixed-length
5093 buffer was used for the list of hosts. I made this 1024 bytes long,
5094 thinking that nobody would ever have a list of hosts that long. Wrong.
5095 Somebody had a whole pile of complicated expansion conditions, and the
5096 string was silently truncated, leading to an expansion error. It turns out
5097 that it is easier to change to an unlimited length (owing to other changes
5098 that have happened since this code was originally written) than to build
5099 structure for giving a limitation error. The length of the item that
5100 expands into the list of hosts is now unlimited.
5102 47. The lsearch lookup could not handle data where the length of text line was
5103 more than 4095 characters. Such lines were truncated, leading to shortened
5104 data being returned. It should now handle lines of any length.
5106 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5107 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5110 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5111 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5112 no longer happens in comment lines. A list of replacements is now placed
5113 at the head of all of the source files, except those whose only change is
5114 to replace PERL_COMMAND in the very first #! line.
5116 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5117 messages on the queue, with a bottom-up merge sort, using code contributed
5118 by Michael Haardt. This should make operations like -bp somewhat faster on
5119 large queues. It won't affect queue runners, except when queue_run_in_order
5122 51. Installed eximstats 1.31 in the distribution.
5124 52. Added support for SRV lookups to the dnslookup router.
5126 53. If an ACL referred to $message_body or $message_body_end, the value was not
5127 reset for any messages that followed in the same SMTP session.
5129 54. The store-handling optimization for building very long strings was not
5130 differentiating between the different store pools. I don't think this
5131 actually made any difference in practice, but I've tidied it.
5133 55. While running the routers to verify a sender address, $sender_address
5134 was still set to the sender address. This is wrong, because when routing to
5135 send a bounce to the sender, it would be empty. Therefore, I have changed
5136 it so that, while verifying a sender address, $sender_address is set to <>.
5137 (There is no change to what happens when verifying a recipient address.)
5139 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5140 target A or AAAA records (if not already returned) without resetting the
5141 qualify_single or search_parents options of the DNS resolver. These are
5142 inappropriate in this case because the targets of MX and SRV records must
5143 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5144 target that, when qualified, matched something in the local domain. These
5145 two options are now turned off when doing these lookups.
5147 57. It seems that at least some releases of Reiserfs (which does not have the
5148 concept of a fixed number of inodes) returns zero and not -1 for the
5149 number of available inodes. This interacted badly with check_spool_inodes,
5150 which assumed that -1 was the "no such thing" setting. What I have done is
5151 to check that the total number of inodes is greater than zero before doing
5152 the test of how many are available.
5154 58. When a "warn" ACL statement has a log_message modifier, the message is
5155 remembered, and not repeated. This is to avoid a lot of repetition when a
5156 message has many recipients that cause the same warning to be written.
5157 Howewer, Exim was preserving the list of already written lines for an
5158 entire SMTP session, which doesn't seem right. The memory is now reset if a
5159 new message is started.
5161 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5162 debugging output unless log_rewrite was also set.
5164 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5165 of (int)(handle) when we know that handle contains (void *)(-1).
5167 61. The Exim daemon panic-logs an error return when it closes the incoming
5168 connection. However "connection reset by peer" seems to be common, and
5169 isn't really an error worthy of noting specially, so that particular error
5172 62. When Exim is trying to find all the local interfaces, it used to panic and
5173 die if the ioctl to get the interface flags failed. However, it seems that
5174 on at least one OS (Solaris 9) it is possible to have an interface that is
5175 included in the list of interfaces, but for which you get a failure error
5176 for this call. This happens when the interface is not "plumbed" into a
5177 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5178 failure of the "get flags" call assumes that the interface is down.
5180 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5181 makes life easier for people who are doing arithmetic on fields extracted
5182 from dates, where you often get leading zeros that should not be
5183 interpreted as octal.
5185 64. Added qualify_domain to the redirect router, to override the global
5188 65. If a pathologically long header line contained very many addresses (the
5189 report of this problem mentioned 10 000) and each of them was rewritten,
5190 Exim could use up a very large amount of memory. (It kept on making new
5191 copies of the header line as it rewrote, and never released the old ones.)
5192 At the expense of a bit more processing, the header rewriting function has
5193 been changed so that it no longer eats memory in this way.
5195 66. The generation of the Received: header has been moved from the time that a
5196 message starts to be received, to the time that it finishes. The timestamp
5197 in the Received: header should now be very close to that of the <= log
5198 line. There are two side-effects of this change:
5200 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5201 logged header lines no longer include the local Received: line, because
5202 it has not yet been created. The same applies to a copy of the message
5203 that is returned to a non-SMTP sender when a message is rejected.
5205 (b) When a filter file is tested using -bf, no additional Received: header
5206 is added to the test message. After some thought, I decided that this
5209 This change does not affect the value of $received_for. It is still set
5210 after address rewriting, but before local_scan() is called.
5212 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5214 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5215 gave an unhelpful panic error message, and a defer error. I have managed to
5216 change this behaviour so that it now rejects any supplied certificate,
5217 which seems right, as the list of acceptable certificates is empty.
5219 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5220 gave an unhelpful defer error. I have not managed to make this reject any
5221 supplied certificates, but the error message it gives is "no certificate
5222 supplied", which is not helpful.
5224 70. exigrep's output now also includes lines that are not associated with any
5225 message, but which match the given pattern. Implemented by a patch from
5226 Martin Sluka, which also tidied up the Perl a bit.
5228 71. Recipient callout verification, like sender verification, was using <> in
5229 the MAIL FROM command. This isn't really the right thing, since the actual
5230 sender may affect whether the remote host accepts the recipient or not. I
5231 have changed it to use the actual sender in the callout; this means that
5232 the cache record is now keyed on a recipient/sender pair, not just the
5233 recipient address. There doesn't seem to be a real danger of callout loops,
5234 since a callout by the remote host to check the sender would use <>.
5235 [SEE ABOVE: changed after hitting problems.]
5237 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5238 temporary errors. However, in the case of such a code being given after
5239 the end of a data transmission (i.e. after ".") Exim was failing to write
5240 a retry record for the message. (Yes, there was some broken host that was
5241 actually sending 8xx at this point.)
5243 73. An unknown lookup type in a host list could cause Exim to panic-die when
5244 the list was checked. (An example that provoked this was putting <; in the
5245 middle of a list instead of at the start.) If this happened during a DATA
5246 ACL check, a -D file could be left lying around. This kind of configuration
5247 error no longer causes Exim to die; instead it causes a defer errror. The
5248 incident is still logged to the main and panic logs.
5250 74. Buglet left over from Exim 3 conversion. The message "too many messages
5251 in one connection" was written to the rejectlog but not the mainlog, except
5252 when address rewriting (yes!) was being logged.
5254 75. Added write_rejectlog option.
5256 76. When a system filter was run not as root (that is, when system_filter_user
5257 was set), the values of the $n variables were not being returned to the
5258 main process; thus, they were not subsequently available in the $sn
5261 77. Added +return_path_on_delivery log selector.
5263 78. A connection timeout was being treated differently from recipients deferred
5264 when testing hosts_max_try with a message that was older than the host's
5265 retry timeout. (The host should not be counted, thus allowing all hosts to
5266 be tried at least once before bouncing.) This may have been the cause of an
5267 occasionally reported bug whereby a message would remain on the queue
5268 longer than the retry timeout, but would be bounced if a delivery was
5269 forced. I say "may" because I never totally pinned down the problem;
5270 setting up timeout/retry tests is difficult. See also the next item.
5272 79. The ultimate address timeout was not being applied to errors that involved
5273 a combination of host plus message (for example, a timeout on a MAIL
5274 command). When an address resolved to a number of possible hosts, and they
5275 were not all tried for each delivery (e.g. because of hosts_max_try), a
5276 message could remain on the queue longer than the retry timeout.
5278 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5281 81. Fixed an obscure SMTP outgoing bug which required at least the following
5282 conditions: (a) there was another message waiting for the same server;
5283 (b) the server returned 5xx to all RCPT commands in the first message so
5284 that the message was not completed; (c) the server dropped the connection
5285 or gave a negative response to the RSET that Exim sends to abort the
5286 transaction. The observed case was a dropped connection after DATA that had
5287 been sent in pipelining mode. That is, the server had advertised PIPELINING
5288 but was not implementing it correctly. The effect of the bug was incorrect
5289 behaviour, such as trying another host, and this could lead to a crash.
5295 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5296 and daemon.c were passed as pointers to ints; they should have been
5297 pointers to socklen_t variables (which are typically unsigned ints).
5299 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5302 3. Fixed a really odd bug that affected only the testing scheme; patching a
5303 certain fixed string in the binary changed the value of another string that
5304 happened to be identical to the end of the original first string.
5306 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5307 name", it returns that address as the IP address. On some operating
5308 systems (e.g. Solaris), it also passes back the IP address string as the
5309 "host name". However, on others (e.g. Linux), it passes back an empty
5310 string. Exim wasn't checking for this, and was changing the host name to an
5311 empty string, assuming it had been canonicized.
5313 5. Although rare, it is permitted to have more than one PTR record for a given
5314 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5315 all the names associated with an address, because they do in Solaris.
5316 However, it seems that they do not in Linux for data that comes from the
5317 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5318 I found this out when I moved to a new Linux workstation and tried to run
5319 the Exim test suite.
5321 To get round this problem I have changed the code so that it now does its
5322 own call to the DNS to look up PTR records when searching for a host name.
5323 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5324 addresses that are only in /etc/hosts are still found.
5326 This behaviour is, however, controlled by an option called host_lookup_
5327 order, which defaults to "bydns:byaddr". If people want to use the other
5328 order, or indeed, just use one or the other means of lookup, they can
5329 specify it in this variable.
5331 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5332 some operating systems, this comes back from gethostbyaddr() as an empty
5333 string, and this is what Exim used to test for. However, it seems that in
5334 other systems, "." is yielded. Exim now tests for this case too.
5336 7. The values of check_spool_space and check_log_space are now held internally
5337 as a number of kilobytes instead of an absolute number of bytes. If a
5338 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5339 kilobyte. This means that much larger values can be stored.
5341 8. Exim monitor: an attempt to get the action menu when not actually pointing
5342 at a message produces an empty menu entitled "No message selected". This
5343 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5344 no entries in it ("Shell widget menu has zero width and/or height"). So I
5345 have added a single, blank menu entry in this case.
5347 9. Added ${quote_local_part.
5349 10. MIME decoding is now applied to the contents of Subject: header lines when
5352 11. Now that a reference to $sender_host_address automatically causes a reverse
5353 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5354 host lookup before query-style lookups in lists that might use this
5355 variable. This has therefore been abolished, and the "net-" prefix is no
5356 longer necessary for query-style lookups.
5358 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5359 have been a typo for LFLAGS, so it has been changed.
5361 13. The install script calls Exim with "-C /dev/null" in order to find the
5362 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5363 to be output. Howeve, since Exim outputs its version number before the
5364 error, it didn't break the script. It just looked ugly. I fixed this by
5365 always allowing "-C /dev/null" if the caller is root.
5367 14. Ignore overlarge ACL variable number when reading spool file - insurance
5368 against a later release with more variables having written the file.
5370 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5371 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5372 this, as well as the form without the "IPv6" on the front.
5374 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5375 OS/Makefile-Darwin file.
5377 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5378 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5380 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5381 variable. However, some OS use ints for inodes. I've added cast to long int
5382 to get rid of the compiler warning.
5384 19. I had forgotten to lock out "/../" in configuration file names when
5385 ALT_CONFIG_PREFIX was set.
5387 20. Routers used for verification do not need to specify transports. However,
5388 if such a router generated a host list, and callout was configured, Exim
5389 crashed, because it could not find a port number from the (non-existent)
5390 transport. It now assumes port 25 in this circumstance.
5392 21. Added the -t option to exigrep.
5394 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5395 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5396 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5397 them is set, LOOKUP_LSEARCH is forced.
5399 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5401 24. Added sender and host information to the "rejected by local_scan()" log
5402 line; previously there was no indication of these.
5404 25. Added .include_if_exists.
5406 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5407 Linux. It turns out that not all file systems support this. Apparently some
5408 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5409 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5410 not supported on the file descriptor, is now ignored when Exim is trying to
5411 sync a directory. This applies only to Linux.
5413 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5415 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5416 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5419 29. Added support for maildirsize files from supplied patch (modified a bit).
5421 30. The use of :fail: followed by an empty string could lead Exim to respond to
5422 sender verification failures with (e.g.):
5424 550 Verification failed for <xxx>
5425 550 Sender verify failed
5427 where the first response line was missing the '-' that indicates it is not
5428 the final line of the response.
5430 31. The loop for finding the name of the user that called Exim had a hardwired
5431 limit of 10; it now uses the value of finduser_retries, which is used for
5432 all other user lookups.
5434 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5437 33. Exim was neglecting to zero errno before one call of strtol() when
5438 expanding a string and expecting an integer value. On some systems this
5439 resulted in spurious "integer overflow" errors. Also, it was casting the
5440 result into an int without checking.
5442 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5443 did not work. The code looks as if it has *never* worked, though it appears
5444 to have been documented since at least releast 1.62. I have made it work.
5446 35. The "timeout_DNS" error in retry rules, also documented since at least
5447 1.62, also never worked. As it isn't clear exactly what this means, and
5448 clearly it isn't a major issue, I have abolished the feature by treating it
5449 as "timeout", and writing a warning to the main and panic logs.
5451 36. The display of retry rules for -brt wasn't always showing the error code
5454 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5455 timeout_connect_A, timeout_connect_MX.
5457 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5458 to the empty sender.
5460 39. The daemon was not analysing the content of -oX till after it had closed
5461 stderr and disconnected from the controlling terminal. This meant that any
5462 syntax errors were only noted on the panic log, and the return code from
5463 the command was 0. By re-arranging the code a little, I've made the
5464 decoding happen first, so such errors now appear on stderr, and the return
5465 code is 1. However, the actual setting up of the sockets still happens in
5466 the disconnected process, so errors there are still only recorded on the
5469 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5470 connections (as happens on some IP stacks) was logged at start up time as
5471 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5472 it from "IPv6 and IPv4", which means that two separate sockets are being
5475 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5476 says whether AF_INET or AF_INET6 was passed as an argument.
5478 42. Exiwhat output was messed up when time zones were included in log
5481 43. Exiwhat now gives more information about the daemon's listening ports,
5482 and whether -tls-on-connect was used.
5484 44. The "port" option of the smtp transport is now expanded.
5486 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5487 silently ignored. Now an error message is written to the main and panic
5490 46. There's a new ACL modifier called "logwrite" which writes to a log file
5491 as soon as it is encountered.
5493 47. Added $local_user_uid and $local_user_gid at routing time.
5495 48. Exim crashed when trying to verify a sender address that was being
5498 49. Exim was recognizing only a space character after ".include". It now also
5499 recognizes a tab character.
5501 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5502 extracting the relevant information from the specification. The man page no
5503 longer contains scrambled data for the -d option, and I've added a section
5504 at the front about calling Exim under different names.
5506 51. Added "extra_headers" argument to the "mail" command in filter files.
5508 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5511 53. Installed eximstats 1.29.
5513 54. Added transport_filter_timeout as a generic transport option.
5515 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5516 Cc: header lines. This was required by RFC 822, but it not required by RFC
5519 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5520 incoming messages that did not have them. Now it does so only if the
5521 message originates locally, that is, if there is no associated remote host
5522 address. When Resent- header lines are present, this applies to the Resent-
5523 lines rather than the non-Resent- lines.
5525 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5526 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5528 58. Messages for configuration errors now include the name of the main
5529 configuration file - useful now that there may be more than one file in a
5530 list (.included file names were always shown).
5532 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5533 for those rare installations that do not start the daemon as root or run it
5534 setuid root. I've cut out the call to initgroups() if the daemon is not
5537 60. The Exim user and group can now be bound into the binary as text strings
5538 that are looked up at the start of Exim's processing.
5540 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5542 62. Added $mailstore_basename variable.
5544 63. Installed patch to sieve.c from Michael Haardt.
5546 64. When Exim failed to open the panic log after failing to open the main log,
5547 the original message it was trying to log was written to stderr and debug
5548 output, but if they were not available (the usual case in production), it
5549 was lost. Now it is written to syslog before the two lines that record the
5550 failures to open the logs.
5552 65. Users' Exim filters run in subprocesses under the user's uid. It is
5553 possible for a "deliver" command or an alias in a "personal" command to
5554 provoke an address rewrite. If logging of address rewriting is configured,
5555 this fails because the process is not running as root or exim. There may be
5556 a better way of dealing with this, but for the moment (because 4.30 needs
5557 to be released), I have disabled address rewrite logging when running a
5558 filter in a non-root, non-exim process.
5564 1. The buildconfig auxiliary program wasn't quoting the value set for
5565 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5566 not defined. This bug was masked in 4.22 by the effect that was fixed in
5569 2. Some messages that were rejected after a message id was allocated were
5570 shown as "incomplete" by exigrep. It no longer does this for messages that
5571 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5573 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5574 have allow_domain_literals set, the ID did not get logged in the <= line.
5575 Domain literals are now always recognized in Message-ID: header lines.
5577 4. The first argument for a ${extract expansion item is the key name or field
5578 number. Leading and trailing spaces in this item were not being ignored,
5579 causing some misleading effects.
5581 5. When deliver_drop_privilege was set, single queue runner processes started
5582 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5583 same command in the process it spins off) were not dropping privilege.
5585 6. When the daemon running as "exim" started a queue runner, it always
5586 re-executed Exim in the spun-off process. This is a waste of effort when
5587 deliver_drop_privilege is set. The new process now just calls the
5588 queue-runner function directly.
5594 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5597 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5598 ignored. Though the use of -oP was forcing the writing of a pid file, it
5599 was always written to the default place.
5601 3. If the message "no IP address found for host xxxx" is generated during
5602 incoming verification, it is now followed by identification of the incoming
5603 connection (so you can more easily find what provoked it).
5605 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5607 5. Added some features to "harden" Exim a bit more against certain attacks:
5609 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5610 be put in Local/Makefile. This is like the never_users runtime option,
5611 but it cannot be overridden. The default setting is "root".
5613 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5614 prefix string with which any file named in a -C command line option
5617 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5618 is retained for -C and -D only if the caller of Exim is root. Without
5619 it, the exim user may also use -C and -D and retain privilege.
5621 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5622 command line option is disabled.
5624 6. Macro names set by the -D option must start with an upper case letter, just
5625 like macro names defined in the configuration file.
5627 7. Added "dereference=" facility to LDAP.
5629 8. Two instances of the typo "uknown" in the source files are fixed.
5631 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5632 the Configure-Makefile script screwed up while processing it.
5634 10. Incorporated PCRE 4.4.
5636 11. The SMTP synchronization check was not operating right at the start of an
5637 SMTP session. For example, it could not catch a HELO sent before the client
5638 waited for the greeting. There is now a check for outstanding input at the
5639 point when the greeting is written. Because of the duplex, asynchronous
5640 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5641 way, but not yet received, when the check is performed.
5643 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5644 on TCP/IP sockets, because this apparently causes some broken clients to
5647 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5648 unchanged) from the Cygwin maintainer.
5650 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5651 was supported instead of testing for mbx. Effectively a typo.
5653 15. The spa authenticator server code was not checking that the input it
5654 received was valid base64.
5656 16. The debug output line for the "set" modifier in ACLs was not showing the
5657 name of the variable that was being set.
5659 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5661 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5662 lookup on its own. Something else was needed to trigger the lookup. For
5663 example, a match in host_lookup or the need for a host name in a host list.
5664 Now, if $sender_host_name is referenced and the host name has not yet been
5665 looked up, a lookup is performed. If the lookup fails, the variable remains
5666 empty, and $host_lookup_failed is set to "1".
5668 19. Added "eqi" as a case-independent comparison operator.
5670 20. The saslauthd authentication condition could segfault if neither service
5671 nor realm was specified.
5673 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5674 error message that was logged was misleading, and incoming SMTP
5675 connections were dropped. The message is now more accurate, and temporary
5676 errors are given to SMTP connections.
5678 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5679 (except RSET) with a 421 error, until QUIT is received. However, it was
5680 failing to send a response to QUIT.
5682 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5683 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5684 ACL is now run after the helo_try_verify_hosts code.
5686 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5687 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5688 case-independent, so other case variants are also recognized). Apparently
5689 some systems use these upper case variants.
5691 25. If more than two messages were waiting for the same host, and a transport
5692 filter was specified for the transport, Exim sent two messages over the
5693 same TCP/IP connection, and then failed with "socket operation on non-
5694 socket" when it tried to send the third.
5696 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5698 27. The extern definition of crypt16() in expand.c was not being excluded when
5699 the OS had its own crypt16() function.
5701 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5702 as a preferred synonym for return_size_limit, both as an option and as an
5705 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5707 30. Changed the default configuration ACL to relax the local part checking rule
5708 for addresses that are not in any local domains. For these addresses,
5709 slashes and pipe symbols are allowed within local parts, but the sequence
5710 /../ is explicitly forbidden.
5712 31. SPA server authentication was not clearing the challenge buffer before
5715 32. log_message in a "warn" ACL statement was writing to the reject log as
5716 well as to the main log, which contradicts the documentation and doesn't
5717 seem right (because no rejection is happening). So I have stopped it.
5719 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5720 However, I am unable to do any testing of this.
5722 34. Fixed an infelicity in the appendfile transport. When checking directories
5723 for a mailbox, to see if any needed to be created, it was accidentally
5724 using path names with one or more superfluous leading slashes; tracing
5725 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5727 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5728 contents are added to the log line that is written for every discarded
5729 recipient. (Previously a log_message setting was ignored.)
5731 36. The ${quote: operator now quotes the string if it is empty.
5733 37. The install script runs exim in order to find its version number. If for
5734 some reason other than non-existence or emptiness, which it checks, it
5735 could not run './exim', it was installing it with an empty version number,
5736 i.e. as "exim-". This error state is now caught, and the installation is
5739 38. An argument was missing from the function that creates an error message
5740 when Exim fails to connect to the socket for saslauthd authentication.
5741 This could cause Exim to crash, or give a corrupted message.
5743 39. Added isip, isip4, and isip6 to ${if conditions.
5745 40. The ACL variables $acl_xx are now saved with the message, and can be
5746 accessed later in routers, transports, and filters.
5748 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5749 strings in the file are not string-expanded.
5751 42. If a MAIL command specified a SIZE value that was too large to fit into an
5752 int variable, the check against message_size_limit failed. Such values are
5753 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5754 day this will have to be increased, but I don't think I want to be around
5755 when emails are that large.
5762 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5763 iconv() is not standard in FreeBSD.
5765 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5766 IPv6 enabled. The observed symptom was a segmentation fault on return from
5767 the function os_common_find_running_interfaces() in src/os.c.
5769 3. In the check_special_case() function in daemon.c I had used "errno" as an
5770 argument name, which causes warnings on some systems. This was basically a
5771 typo, since it was named "eno" in the comments!
5773 4. The code that waits for the clock to tick (at a resolution of some fraction
5774 of a second) so as to ensure message-id uniqueness was always waiting for
5775 at least one whole tick, when it could have waited for less. [This is
5776 almost certainly not relevant at current processor speeds, where it is
5777 unlikely to ever wait at all. But we try to future-proof.]
5779 5. The function that sleeps for a time interval that includes fractions of a
5780 second contained a race. It did not block SIGALRM between setting the
5781 timer, and suspending (a couple of lines later). If the interval was short
5782 and the sigsuspend() was delayed until after it had expired, the suspension
5783 never ended. On busy systems this could lead to processes getting stuck for
5786 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5787 process, before it forks any delivery processes. The open lookup caching
5788 mechanism meant that the open file or database connection was passed into
5789 the delivery process. The problem was that delivery processes always tidy
5790 up cached lookup data. This could cause a problem for the next delivery
5791 process started by the queue runner, because the external queue runner
5792 process does not know about the closure. So the next delivery process
5793 still has data in the lookup cache. In the case of a file lookup, there was
5794 no problem because closing a file descriptor in a subprocess doesn't affect
5795 the parent. However, if the lookup was caching a connection to a database,
5796 the connection was closed, and the second delivery process was likely to
5797 see errors such as "PGSQL: query failed: server closed the connection
5798 unexpectedly". The problem has been fixed by closing all cached lookups
5799 in a queue runner before running a delivery process.
5801 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5802 seem to have the "const" qualifier which it has on other OS. I've
5805 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5806 *of the same type* (client or server) with the same public name that an
5807 error should be diagnosed.
5809 9. When Exim looked up a host name for an IP address, but failed to find the
5810 original IP address when looking up the host name (a safety check), it
5811 output the message "<ip address> does not match any IP for NULL", which was
5812 confusing, to say the least. The bug was that the host name should have
5813 appeared instead of "NULL".
5815 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5816 user that is built into the binary, and the -C or -D options is used, root
5817 privilege is dropped before the configuration file is read. In addition,
5818 logging is switched to stderr instead of the normal log files. If the
5819 configuration then re-defines the Exim user, the unprivileged environment
5820 is probably not what is expected, so Exim logs a panic warning message (but
5823 However, if deliver_drop_privilege is set, the unprivileged state may well
5824 be exactly what is intended, so the warning has been cut out in that case,
5825 and Exim is allowed to try to write to its normal log files.
5831 1. smtp_return_error_details was not giving details for temporary sender
5832 or receiver verification errors.
5834 2. Diagnose a configuration error if two authenticators have the same public
5837 3. Exim used not to create the message log file for a message until the first
5838 delivery attempt. This could be confusing when incoming messages were held
5839 for policy or load reasons. The message log file is now created at the time
5840 the message is received, and an initial "Received" line is written to it.
5842 4. The automatically generated man page for command line options had a minor
5843 bug that caused no ill effects; however, a more serious problem was that
5844 the procedure for building the man page automatically didn't always
5845 operate. Consequently, release 4.20 contains an out-of-date version. This
5846 shouldn't happen again.
5848 5. When building Exim with embedded Perl support, the script that builds the
5849 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5850 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5852 6. The freeze_tell option was not being used for messages that were frozen on
5853 arrival, either by an ACL or by local_scan().
5855 7. Added the smtp_incomplete_transaction log selector.
5857 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5858 was accepting AUTH without a new EHLO.
5860 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5861 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5864 10. Logging of TCP/IP connections (when configured) now happens in the main
5865 daemon process instead of the child process, so that the TCP/IP connection
5866 count is more accurate (but it can never be perfect).
5868 11. The use of "drop" in a nested ACL was not being handled correctly in the
5869 outer ACL. Now, if condition failure induced by the nested "drop" causes
5870 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5871 or "require"), the connection is dropped.
5873 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5874 that yield "discard" can now be used with an "accept" or a "discard" verb,
5875 but an error is generated for any others (because I can't see a useful way
5876 to define what should happen).
5878 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5879 are now processed in the same way as lines in the Exim configuration file.
5880 In particular, continuation lines are supported.
5882 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5884 15. Added -ti meaning -t -i.
5886 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5887 domains, and warn by logging if others are found.
5889 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5890 returned from GIFCONF when Exim is trying to find the list of interfaces on
5891 a host. The code in os.c has been modified to copy each ifreq to an aligned
5892 structure in all cases.
5894 Also, in some cases, the returned ifreq's were being copied to a 'struct
5895 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5896 means the last couple of bytes of an IPv6 address could be chopped if the
5897 ifreq contained only a normal sockaddr (14 bytes storage).
5899 18. Named domain lists were not supported in the hosts_treat_as_local option.
5900 An entry such as +xxxx was not recognized, and was treated as a literal
5903 19. Ensure that header lines added by a DATA ACL are included in the reject log
5904 if the ACL subsequently rejects the message.
5906 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5907 MD5 (which is deprecated).
5909 21. When testing a filter file using -bf, Exim was writing a message when it
5910 took the sender from a "From " line in the message, but it was not doing so
5911 when it took $return_path from a Return-Path: header line. It now does.
5913 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5914 with a valid header line field name (a series of printing characters
5915 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5917 23. Changed "disc" in the source to "disk" to conform to the documentation and
5918 the book and for uniformity.
5920 24. Ignore Sendmail's -Ooption=value command line item.
5922 25. When execve() failed while trying to run a command in a pipe transport,
5923 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5924 could be confused with a return value of 69 from the command itself. This
5925 has been changed to 127, the value the shell returns if it is asked to run
5926 a non-existent command. The wording for the related log line suggests a
5927 non-existent command as the problem.
5929 26. If received_header_text expands to an empty string, do not add a Received:
5930 header line to the message. (Well, it adds a token one on the spool, but
5931 marks it "old" so that it doesn't get used or transmitted.)
5933 27. Installed eximstats 1.28 (addition of -nt option).
5935 28. There was no check for failure on the call to getsockname() in the daemon
5936 code. This can fail if there is a shortage of resources on the system, with
5937 ENOMEM, for example. A temporary error is now given on failure.
5939 29. Contrary to the C standard, it seems that in some environments, the
5940 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5941 program. Exim now does this explicitly; it affects the formatting of
5942 timestamps using strftime().
5944 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5945 complaints. I've now initialized all the variables, to avoid this.
5947 32. Header lines added by a system filter were not being "seen" during
5948 transport-time rewrites.
5950 33. The info_callback() function passed to OpenSSL is set up with type void
5951 (*)(SSL *, int, int), as described somewhere. However, when calling the
5952 function (actually a macro) that sets it up, the type void(*)() is
5953 expected. I've put in a cast to prevent warnings from picky compilers.
5955 34. If a DNS black list lookup found a CNAME record, but there were no A
5956 records associated with the domain it pointed at, Exim crashed.
5958 35. If a DNS black list lookup returned more than one A record, Exim ignored
5959 all but the first. It now scans all returned addresses if a particular IP
5960 value is being sought. In this situation, the contents of the
5961 $dnslist_value variable are a list of all the addresses, separated by a
5964 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5965 to accept a host name and all its aliases if the forward lookup for any of
5966 them yielded the IP address of the incoming connection. Now it accepts only
5967 those names whose forward lookup yields the correct IP address. Any other
5968 names are discarded. This closes a loophole whereby a rogue DNS
5969 administrator could create reverse DNS records to break through a
5970 wildcarded host restriction in an ACL.
5972 37. If a user filter or a system filter that ran in a subprocess used any of
5973 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5974 the wrong values were passed to the pipe command ($thisaddress had the
5975 value of $0, $0 had the value of $1, etc). This bug was introduced by
5976 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5978 38. Improved the line breaking for long SMTP error messages from ACLs.
5979 Previously, if there was no break point between 40 and 75 characters, Exim
5980 left the rest of the message alone. Two changes have been made: (a) I've
5981 reduced the minimum length to 35 characters; (b) if it can't find a break
5982 point between 35 and 75 characters, it looks ahead and uses the first one
5983 that it finds. This may give the occasional overlong line, but at least the
5984 remaining text gets split now.
5986 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5987 file descriptors might be low, and that setting 1000 would always raise it.
5988 It turns out that in some environments, the limit is already over 1000 and
5989 that lowering it causes trouble. So now Exim takes care not to decrease it.
5991 40. When delivering a message, the value of $return_path is set to $sender_
5992 address at the start of routing (routers may change the value). By an
5993 oversight, this default was not being set up when an address was tested by
5994 -bt or -bv, which affected the outcome if any router or filter referred to
5997 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5998 log only when "message" or "log_message" are set. However, if one of the
5999 conditions was an address verification, or a call to a nested ACL, the
6000 messages generated by the underlying test were being passed through. This
6001 no longer happens. The underlying message is available in $acl_verify_
6002 message for both "message" and "log_message" expansions, so it can be
6003 passed through if needed.
6005 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6006 new expansion $bh_ to give the encoded byte string without charset
6007 translation. Translation happens only if iconv() is available; HAVE_ICONV
6008 indicates this at build time. HEADERS_CHARSET gives the charset to
6009 translate to; headers_charset can change it in the configuration, and
6010 "headers charset" can change it in an individual filter file.
6012 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6013 that creates RFC 2047 encoded "words" labels them as that charset instead
6014 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6015 expansion operator; (ii) when Exim creates a From: line for a local
6016 message; (iii) when a header line is rewritten to include a "phrase" part.
6018 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6019 buggy, causing it to skip the first lines of messages whose message ID
6020 ended in 'D'. This would not have bitten before Exim release 4.14, because
6021 message IDs were unlikely to end in 'D' before then. The effect was to have
6022 incorrect size information for certain domains.
6024 45. #include "config.h" was missing at the start of the crypt16.c module. This
6025 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6028 46. If there was a timeout during a "random" callout check, Exim treated it as
6029 a failure of the random address, and carried on sending RSET and the real
6030 address. If the delay was just some slowness somewhere, the response to the
6031 original RCPT would be taken as a response to RSET and so on, causing
6032 mayhem of various kinds.
6034 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6035 when I implemented it. It didn't allow for the fact that some option values
6036 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6037 the right test for positive values.
6039 48. Domain names in DNS records are case-independent. Exim always looks them up
6040 in lower case. Some resolvers return domain names in exactly the case they
6041 appear in the zone file, that is, they may contain uppercase letters. Not
6042 all resolvers do this - some return always lower case. Exim was treating a
6043 change of case by a resolver as a change of domain, similar to a widening
6044 of a domain abbreviation. This triggered its re-routing code and so it was
6045 trying to route what was effectively the same domain again. This normally
6046 caused routing to fail (because the router wouldn't handle the domain
6047 twice). Now Exim checks for this case specially, and just changes the
6048 casing of the domain that it ultimately uses when it transmits the message
6051 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6054 50. If a filter generated a file delivery with a non-absolute name (possible if
6055 no home directory exists for the router), the forbid_file option was not
6058 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6059 the existing equality matching.
6061 52. Exim was using ints instead of ino_t variables in some places where it was
6062 dealing with inode numbers.
6064 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6065 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6066 called TMPDIR, and if it finds it is different, it changes its value.
6068 54. The smtp_printf() function is now made available to local_scan() so
6069 additional output lines can be written before returning. There is also an
6070 smtp_fflush() function to enable the detection of a dropped connection.
6071 The variables smtp_input and smtp_batched_input are exported to
6074 55. Changed the default runtime configuration: the message "Unknown user"
6075 has been removed from the ACL, and instead placed on the localuser router,
6076 using the cannot_route_message feature. This means that any verification
6077 failures that generate their own messages won't get overridden. Similarly,
6078 the "Unrouteable address" message that was in the ACL for unverifiable
6079 relay addresses has also been removed.
6081 56. Added hosts_avoid_esmtp to the smtp transport.
6083 57. The exicyclog script was not checking for the esoteric option
6084 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6085 will work only if exicyclog is run under the appropriate euid.
6087 58. Following a discussion on the list, the rules by which Exim recognises line
6088 endings on incoming messages have been changed. The -dropcr and drop_cr
6089 options are now no-ops, retained only for backwards compatibility. The
6090 following line terminators are recognized: LF CRLF CR. However, special
6091 processing applies to CR:
6093 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6094 nor a local message in the state where . is a terminator.
6096 (ii) If a bare CR is encountered in a header line, an extra space is added
6097 after the line terminator so as not to end the header. The reasoning
6098 behind this is that bare CRs in header lines are most likely either
6099 to be mistakes, or people trying to play silly games.
6101 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6102 was being incorrectly given as 18 bytes larger than it should have been.
6103 This is a VOB (very old bug).
6105 60. This may never have affected anything current, but just in case it has:
6106 When the local host is found other than at the start of a list of hosts,
6107 the local host, those with the same MX, and any that follow, are discarded.
6108 When the list in question was part of a longer list of hosts, the following
6109 hosts (not currently being processed) were also being discarded. This no
6110 longer happens. I'm not sure if this situation could ever has previously
6113 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6116 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6117 header. This is something that is recommended in a new Internet Draft, and
6118 is something that is documented as being done by Sendmail. There are two
6119 possible values. For messages generated by the autoreply transport, Exim
6122 Auto-Submitted: auto-replied
6124 whereas for all other generated messages (e.g. bounces) it adds
6126 Auto-Submitted: auto-generated
6128 63. The "personal" condition in filters now includes a test for the
6129 Auto-Submitted: header. If it contains the string "auto-" the message it
6130 not considered personal.
6132 64. Added rcpt_include_affixes as a generic transport option.
6134 65. Added queue_only_override (default true).
6136 66. Added the syslog_duplication option.
6138 67. If what should have been the first header line of a message consisted of
6139 a space followed by a colon, Exim was mis-interpreting it as a header line.
6140 It isn't of course - it is syntactically invalid and should therefore be
6141 treated as the start of the message body. The misbehaviour could have
6142 caused a number of strange effects, including loss of data in subsequent
6143 header lines, and spool format errors.
6145 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6146 client host had authenticated. This control can now be exercised by an ACL
6147 for more flexibility.
6149 69. By default, callouts do not happen when testing with -bh. There is now a
6150 variant, -bhc, which does actually run the callout code, including
6151 consulting and updating the callout cache.
6153 70. Added support for saslauthd authentication, courtesy of Alexander
6156 71. If statvfs() failed on the spool or log directories while checking their
6157 size for availability, Exim confusingly gave the error "space shortage".
6158 Furthermore, in debugging mode it crashed with a floating point exception.
6159 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6160 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6161 serious problem, Exim now writes to the main and panic logs when this
6162 happens, with details of the failure. It then refuses to accept the
6163 incoming message, giving the message "spool directory problem" or "log
6164 directory problem" with a 421 code for SMTP messages.
6166 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6167 0, 1, and 2 exist, because some OS complain for execs without them (see
6168 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6169 descriptors. However, the code omitted to check that the open succeeded,
6170 causing mysterious errors if for some reason the permissions on /dev/null
6171 got screwed. Now Exim writes a message to the main and panic logs, and
6172 bombs out if it can't open /dev/null.
6174 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6175 interact so that it is all more flexible. It is supposed to remain
6176 backwards compatible. Also added extra_local_interfaces.
6178 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6179 to bomb out with an assertion failure - to the client this appears as a
6180 connection drop. This problem occurs in the part of the code that was taken
6181 from the Samba project. Fortunately, the assertion is in a very simple
6182 function, so I have fixed this by reproducing the function inline in the
6183 one place where it is called, and arranging for authentication to fail
6184 instead of killing the process with assert().
6186 75. The SPA client code was not working when the server requested OEM rather
6187 than Unicode encoding.
6189 76. Added code to make require_files with a specific uid setting more usable in
6190 the case where statting the file as root fails - usually a non-root-mounted
6191 NFS file system. When this happens and the failure is EACCES, Exim now
6192 forks a subprocess and does the per-uid checking as the relevant uid.
6194 77. Added process_log_path.
6196 78. If log_file_path was not explicitly set, a setting of check_log_space or
6197 check_log_inodes was ignored.
6199 79. If a space check for the spool or log partitions fails, the incident is now
6200 logged. Of course, in the latter case the data may get lost...
6202 80. Added the %p formatting code to string_format() so that it can be used to
6203 print addresses in debug_print(). Adjusted all the address printing in the
6204 debugging in store.c to use %p rather than %d.
6206 81. There was a concern that a line of code in smtp_in.c could overflow a
6207 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6208 initially expressed, the concern was not well-founded, because trailing
6209 spaces are removed early. However, if the trailing spaces were followed by
6210 a NULL, they did not get removed, so the overflow was possible. Two fixes
6213 (a) I re-wrote the offending code in a cleaner fashion.
6214 (b) If an incoming SMTP command contains a NULL character, it is rejected
6217 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6218 runs initgroups(), so that if the Exim user is in any additional groups,
6219 they will be used during message reception.
6225 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob