1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
51 TL/01 When running the test suite, the README says that variables such as
52 no_msglog_check are global and can be placed anywhere in a specific
53 test's script, however it was observed that placement needed to be near
54 the beginning for it to behave that way. Changed the runtest perl
55 script to read through the entire script once to detect and set these
56 variables, reset to the beginning of the script, and then run through
57 the script parsing/test process like normal.
59 TL/02 The BSD's have an arc4random API. One of the functions to induce
60 adding randomness was arc4random_stir(), but it has been removed in
61 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
62 function when detected.
64 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
65 cause callback expansion.
67 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
68 syntax errors in an expansion can be treated as a string instead of
69 logging or causing an error, due to the internal use of bool_lax
70 instead of bool when processing it.
72 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
73 server certificates when making smtp deliveries.
75 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
77 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
79 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
81 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
82 Merged patch from Sebastian Wiedenroth.
84 JH/05 Fix results-pipe from transport process. Several recipients, combined
85 with certificate use, exposed issues where response data items split
86 over buffer boundaries were not parsed properly. This eventually
87 resulted in duplicates being sent. This issue only became common enough
88 to notice due to the introduction of conection certificate information,
89 the item size being so much larger. Found and fixed by Wolfgang Breyha.
91 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
92 size buffer was used, resulting in syntax errors when an expansion
95 JH/07 Add support for directories of certificates when compiled with a GnuTLS
96 version 3.3.6 or later.
98 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
99 is EXPERIMENTAL_EVENT, the main-configuration and transport options
100 both become "event_action", the variables become $event_name, $event_data
101 and $event_defer_errno. There is a new variable $verify_mode, usable in
102 routers, transports and related events. The tls:cert event is now also
103 raised for inbound connections, if the main configuration event_action
106 TL/06 In test suite, disable OCSP for old versions of openssl which contained
107 early OCSP support, but no stapling (appears to be less than 1.0.0).
109 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
110 server certificate names available under the smtp transport option
111 "tls_verify_cert_hostname" now do not permit multi-component wildcard
114 JH/10 Time-related extraction expansions from certificates now use the main
115 option "timezone" setting for output formatting, and are consistent
116 between OpenSSL and GnuTLS compilations. Bug 1541.
118 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
119 encoded parameter in the incoming message. Bug 1558.
121 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
122 include certificate info, eximon was claiming there were spoolfile
125 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
127 JH/14 Log delivery-related information more consistently, using the sequence
128 "H=<name> [<ip>]" wherever possible.
130 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
131 are problematic for Debian distribution, omit them from the release
134 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
136 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
139 JH/17 Fix a null-indirection in certextract expansions when a nondefault
140 output list separator was used.
145 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
146 checkers that were complaining about end of non-void function with no
149 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
150 This was a regression intruduced in 4.83 by another bugfix.
152 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
154 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
155 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
161 TF/01 Correctly close the server side of TLS when forking for delivery.
163 When a message was received over SMTP with TLS, Exim failed to clear up
164 the incoming connection properly after forking off the child process to
165 deliver the message. In some situations the subsequent outgoing
166 delivery connection happened to have the same fd number as the incoming
167 connection previously had. Exim would try to use TLS and fail, logging
168 a "Bad file descriptor" error.
170 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
171 utilities have not been installed.
173 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
174 temporary space as the ACL may create new global variables.
176 TL/01 LDAP support uses per connection or global context settings, depending
177 upon the detected version of the libraries at build time.
179 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
180 to extract and use the src ip:port in logging and expansions as if it
181 were a direct connection from the outside internet. PPv2 support was
182 updated based on HAProxy spec change in May 2014.
184 JH/02 Add ${listextract {number}{list}{success}{fail}}.
186 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
187 Properly escape header and check for NULL return.
189 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
192 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
194 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
195 characters in header names, implemented as a verify condition.
196 Contributed by Michael Fischer v. Mollard.
198 TL/05 Rename SPF condition results err_perm and err_temp to standardized
199 results permerror and temperror. Previous values are deprecated but
200 still accepted. In a future release, err_perm and err_temp will be
201 completely removed, which will be a backward incompatibility if the
202 ACL tests for either of these two old results. Patch contributed by
203 user bes-internal on the mailing list.
205 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
207 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
208 selectors, in both main and reject logs.
210 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
213 JH/07 Add malware type "sock" for talking to simple daemon.
215 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
217 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
218 routers/transports under cutthrough routing.
220 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
221 numbers. Touch up "bool" conditional to keep the same definition.
223 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
225 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
227 JH/12 Expand items in router/transport headers_add or headers_remove lists
228 individually rather than the list as a whole. Bug 1452.
230 Required for reasonable handling of multiple headers_ options when
231 they may be empty; requires that headers_remove items with embedded
232 colons must have them doubled (or the list-separator changed).
234 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
235 view the policy declared in the DMARC record. Currently, $dmarc_status
236 is a combined value of both the record presence and the result of the
239 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
241 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
242 dnslookup router and the smtp transport (applying to the forward
245 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
246 of ldap servers used for a specific lookup. Patch provided by Heiko
249 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
250 New variable $lookup_dnssec_authenticated for observability.
252 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
253 Patch submitted by Lars Timman.
255 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
257 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
258 Requires trusted mode and valid format message id, aborts otherwise.
259 Patch contributed by Heiko Schlichting.
261 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
262 certextract with support for various fields. Bug 1358.
264 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
265 is requested by default, modifiable by smtp transport option
268 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
269 operate on certificate variables to give certificate fingerprints
270 Also new ${sha256:cert_variable}.
272 JH/23 The PRDR feature is moved from being Experimental into the mainline.
274 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
275 Christian Aistleitner.
277 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
279 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
280 file. Patch from Wolfgang Breyha.
282 JH/25 Expand the coverage of the delivery $host and $host_address to
283 client authenticators run in verify callout. Bug 1476.
285 JH/26 Port service names are now accepted for tls_on_connect_ports, to
286 align with daemon_smtp_ports. Bug 72.
288 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
289 support and error reporting did not work properly.
291 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
292 and is readable. Patch from Andrew Colin Kissa.
294 TL/14 Enhance documentation of ${run expansion and how it parses the
295 commandline after expansion, particularly in the case when an
296 unquoted variable expansion results in an empty value.
298 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
300 PP/02 Fix internal collision of T_APL on systems which support RFC3123
301 by renaming away from it. Addresses GH issue 15, reported by
304 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
306 TL/15 SECURITY: prevent double expansion in math comparison functions
307 (can expand unsanitized data). Not remotely exploitable.
314 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
316 PP/02 Make -n do something, by making it not do something.
317 When combined with -bP, the name of an option is not output.
319 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
322 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
323 $sender_host_name and config options to manage this, and basic check
326 PP/05 DSCP support for outbound connections and control modifier for inbound.
328 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
329 (Only plugin which currently uses this is kerberos4, which nobody should
330 be using, but we should make it available and other future plugins might
331 conceivably use it, even though it would break NAT; stuff *should* be
332 using channel bindings instead).
334 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
335 name; added for Sendmail compatibility; requires admin caller.
336 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
337 just ignore it); requires trusted caller.
338 Also parse but ignore: -Ac -Am -X<logfile>
341 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
343 TL/02 Add +smtp_confirmation as a default logging option.
345 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
346 Patch by Magnus Holmgren from 2007-02-20.
348 TL/04 Bugzilla 1281 - Spec typo.
349 Bugzilla 1283 - Spec typo.
350 Bugzilla 1290 - Spec grammar fixes.
352 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
354 TL/06 Add Experimental DMARC support using libopendmarc libraries.
356 TL/07 Fix an out of order global option causing a segfault. Reported to dev
357 mailing list by by Dmitry Isaikin.
359 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
361 JH/02 Support "G" suffix to numbers in ${if comparisons.
363 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
365 NM/01 Bugzilla 1197 - Spec typo
366 Bugzilla 1196 - Spec examples corrections
368 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
370 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
371 gnutls_enable_pkcs11, but renamed to more accurately indicate its
374 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
375 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
377 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
378 "acl {{name}{arg}...}", and optional args on acl condition
381 JH/05 Permit multiple router/transport headers_add/remove lines.
383 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
385 JH/07 Avoid using a waiting database for a single-message-only transport.
386 Performance patch from Paul Fisher. Bugzilla 1262.
388 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
391 JH/09 Add $headers_added variable, with content from use of ACL modifier
392 add_header (but not yet added to the message). Bugzilla 199.
394 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
395 Pulled from Bugzilla 817 by Wolfgang Breyha.
397 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
399 (nb: this is the same fix as in Exim 4.80.1)
401 JH/11 Add A= logging on delivery lines, and a client_set_id option on
404 JH/12 Add optional authenticated_sender logging to A= and a log_selector
407 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
409 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
410 advertise SMTP AUTH mechanism to us, instead of a generic
411 protocol violation error. Also, make Exim more robust to bad
412 data from the Dovecot auth socket.
414 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
416 When a queue runner is handling a message, Exim first routes the
417 recipient addresses, during which it prunes them based on the retry
418 hints database. After that it attempts to deliver the message to
419 any remaining recipients. It then updates the hints database using
422 So if a recipient address works intermittently, it can get repeatedly
423 deferred at routing time. The retry hints record remains fresh so the
424 address never reaches the final cutoff time.
426 This is a fairly common occurrence when a user is bumping up against
427 their storage quota. Exim had some logic in its local delivery code
428 to deal with this. However it did not apply to per-recipient defers
429 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
431 This change adds a proper retry rule check during routing so that the
432 final cutoff time is checked against the message's age. We only do
433 this check if there is an address retry record and there is not a
434 domain retry record; this implies that previous attempts to handle
435 the address had the retry_use_local_parts option turned on. We use
436 this as an approximation for the destination being like a local
437 delivery, as in LMTP.
439 I suspect this new check makes the old local delivery cutoff check
440 redundant, but I have not verified this so I left the code in place.
442 TF/02 Correct gecos expansion when From: is a prefix of the username.
444 Test 0254 submits a message to Exim with the header
448 When I ran the test suite under the user fanf2, Exim expanded
449 the header to contain my full name, whereas it should have added
450 a Resent-Sender: header. It erroneously treats any prefix of the
451 username as equal to the username.
453 This change corrects that bug.
455 GF/01 DCC debug and logging tidyup
456 Error conditions log to paniclog rather than rejectlog.
457 Debug lines prefixed by "DCC: " to remove any ambiguity.
459 TF/03 Avoid unnecessary rebuilds of lookup-related code.
461 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
462 Bug spotted by Jeremy Harris; was flawed since initial commit.
463 Would have resulted in OCSP responses post-SNI triggering an Exim
464 NULL dereference and crash.
466 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
468 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
469 Bug detection, analysis and fix by Samuel Thibault.
470 Bugzilla 1331, Debian bug #698092.
472 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
474 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
475 Server implementation by Todd Lyons, client by JH.
476 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
477 config variable "prdr_enable" controls whether the server
478 advertises the facility. If the client requests PRDR a new
479 acl_data_smtp_prdr ACL is called once for each recipient, after
480 the body content is received and before the acl_smtp_data ACL.
481 The client is controlled by bolth of: a hosts_try_prdr option
482 on the smtp transport, and the server advertisement.
483 Default client logging of deliveries and rejections involving
484 PRDR are flagged with the string "PRDR".
486 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
487 fclose(). Diagnosis by Todd Lyons.
489 PP/17 Update configure.default to handle IPv6 localhost better.
490 Patch by Alain Williams (plus minor tweaks).
493 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
494 This is now consistent with GnuTLS, and is now documented: the
495 previous undocumented portable approach to treating the option as
496 unset was to force an expansion failure. That still works, and
497 an empty string is now equivalent.
499 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
500 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
501 not performing validation itself.
503 PP/20 Added force_command boolean option to pipe transport.
504 Patch from Nick Koston, of cPanel Inc.
506 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
509 TF/04 Added udpsend ACL modifer and hexquote expansion operator
511 PP/21 Fix eximon continuous updating with timestamped log-files.
512 Broken in a format-string cleanup in 4.80, missed when I repaired the
513 other false fix of the same issue.
514 Report and fix from Heiko Schlichting.
517 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
518 Report from Prashanth Katuri.
520 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
521 It's SecureTransport, so affects any MacOS clients which use the
522 system-integrated TLS libraries, including email clients.
524 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
525 using a MIME ACL for non-SMTP local injection.
526 Report and assistance in diagnosis by Warren Baker.
528 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
530 JH/16 Fix comparisons for 64b. Bugzilla 1385.
532 TL/09 Add expansion variable $authenticated_fail_id to keep track of
533 last id that failed so it may be referenced in subsequent ACL's.
535 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
538 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
539 ldap library initialization, allowing self-signed CA's to be
540 used. Also properly sets require_cert option later in code by
541 using NULL (global ldap config) instead of ldap handle (per
542 session). Bug diagnosis and testing by alxgomz.
544 TL/12 Enhanced documentation in the ratelimit.pl script provided in
545 the src/util/ subdirectory.
547 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
548 renamed to Transport Post Delivery Action by Jeremy Harris, as
551 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
552 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
553 redis_servers = needs to be configured which will be used by the redis
554 lookup. Patch from Warren Baker, of The Packet Hub.
556 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
558 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
559 hostname or reverse DNS when processing a host list. Used suggestions
560 from multiple comments on this bug.
562 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
564 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
565 Missed a few lines, added it to make the runtest require no keyboard
568 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
569 contains upper case chars. Make router use caseful_local_part.
571 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
572 support when GnuTLS has been built with p11-kit.
578 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
580 This, or similar/improved, will also be change PP/11 of 4.82.
586 PP/01 Handle short writes when writing local log-files.
587 In practice, only affects FreeBSD (8 onwards).
588 Bugzilla 1053, with thanks to Dmitry Isaikin.
590 NM/01 Bugzilla 949 - Documentation tweak
592 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
595 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
597 PP/02 Implemented gsasl authenticator.
599 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
601 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
602 `pkg-config foo` for cflags/libs.
604 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
605 with rest of GSASL and with heimdal_gssapi.
607 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
608 `pkg-config foo` for cflags/libs for the TLS implementation.
610 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
611 properties get this fed in as external SSF. A number of robustness
612 and debugging improvements to the cyrus_sasl authenticator.
614 PP/08 cyrus_sasl server now expands the server_realm option.
616 PP/09 Bugzilla 1214 - Log authentication information in reject log.
617 Patch by Jeremy Harris.
619 PP/10 Added dbmjz lookup type.
621 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
623 PP/12 MAIL args handles TAB as well as SP, for better interop with
624 non-compliant senders.
625 Analysis and variant patch by Todd Lyons.
627 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
628 Bug report from Lars Müller <lars@samba.org> (via SUSE),
629 Patch from Dirk Mueller <dmueller@suse.com>
631 PP/13 tls_peerdn now print-escaped for spool files.
632 Observed some $tls_peerdn in wild which contained \n, which resulted
633 in spool file corruption.
635 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
636 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
637 or write after TLS renegotiation, which otherwise led to messages
640 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
641 as a tracking header (ie: a signed header comes before the signature).
642 Patch from Wolfgang Breyha.
644 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
645 comma-sep list; embedded commas doubled.
647 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
649 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
651 Report and patch from Dmitry Banschikov.
653 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
654 Removed SSL_clear() after SSL_new() which led to protocol negotiation
655 failures. We appear to now support TLS1.1+ with Exim.
657 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
658 lets Exim select keys and certificates based upon TLS SNI from client.
659 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
660 before an outbound SMTP session. New log_selector, +tls_sni.
662 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
663 NULL dereference. Report and patch from Alun Jones.
665 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
666 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
667 Not seeing resolver debug output on NetBSD, but suspect this is a
668 resolver implementation change.
670 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
671 Left warnings. Added "eximon gdb" invocation mode.
673 PP/21 Defaulting "accept_8bitmime" to true, not false.
675 PP/22 Added -bw for inetd wait mode support.
677 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
678 locate the relevant includes and libraries. Made this the default.
680 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
681 Bugzilla 1246, report and most of solution from Tomasz Kusy.
683 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
684 This may cause build issues on older platforms.
686 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
687 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
688 gnutls_require_mac & gnutls_require_protocols (no longer supported).
689 Added SNI support via GnuTLS too.
690 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
692 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
694 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
695 Applied second patch from Janne, implementing suggestion to default
696 multiple-strings-in-record handling to match SPF spec.
698 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
700 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
701 read-only, out of scope).
702 Patch from Wolfgang Breyha, report from Stuart Northfield.
704 PP/29 Fix three issues highlighted by clang analyser static analysis.
705 Only crash-plausible issue would require the Cambridge-specific
706 iplookup router and a misconfiguration.
707 Report from Marcin Mirosław.
709 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
711 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
712 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
713 As part of this, removing so much warning spew let me fix some minor
714 real issues in debug logging.
716 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
717 assignment on my part. Fixed.
719 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
720 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
721 Janne Snabb (who went above and beyond: thank you).
723 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
724 string otherwise requires a connection and a bunch more work and it's
725 relatively easy to get wrong. Should also expose TLS library linkage
728 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
729 64-bit ${eval} (JH/03).
731 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
732 GNU libc to support some of the 64-bit stuff, should not lead to
733 conflicts. Defined before os.h is pulled in, so if a given platform
734 needs to override this, it can.
736 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
737 protection layer was required, which is not implemented.
738 Bugzilla 1254, patch from Wolfgang Breyha.
740 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
741 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
742 tls_dhparam take prime identifiers. Also unbreak combination of
743 OpenSSL+DH_params+TLSSNI.
745 PP/39 Disable SSLv2 by default in OpenSSL support.
751 PP/01 Solaris build fix for Oracle's LDAP libraries.
752 Bugzilla 1109, patch from Stephen Usher.
754 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
756 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
759 TF/02 Fix a couple more cases where we did not log the error message
760 when unlink() failed. See also change 4.74-TF/03.
762 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
763 lock up or crash if it happened to be inside a call to libc when it
764 got a SIGUSR1 from exiwhat.
766 The SIGUSR1 handler appends the current process status to the process
767 log which is later printed by exiwhat. It used to use the general
768 purpose logging code to do this, but several functions it calls are
769 not safe for signals.
771 The new output code in the SIGUSR1 handler is specific to the process
772 log, and simple enough that it's easy to inspect for signal safety.
773 Removing some special cases also simplifies the general logging code.
774 Removing the spurious timestamps from the process log simplifies
777 TF/04 Improved ratelimit ACL condition.
779 The /noupdate option has been deprecated in favour of /readonly which
780 has clearer semantics. The /leaky, /strict, and /readonly update modes
781 are mutually exclusive. The update mode is no longer included in the
782 database key; it just determines when the database is updated. (This
783 means that when you upgrde Exim will forget old rate measurements.)
785 Exim now checks that the per_* options are used with an update mode that
786 makes sense for the current ACL. For example, when Exim is processing a
787 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
788 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
789 must specify per_mail/readonly. If you omit the update mode it defaults to
790 /leaky where that makes sense (as before) or /readonly where required.
792 The /noupdate option is now undocumented but still supported for
793 backwards compatibility. It is equivalent to /readonly except that in
794 ACLs where /readonly is required you may specify /leaky/noupdate or
795 /strict/noupdate which are treated the same as /readonly.
797 A useful new feature is the /count= option. This is a generalization
798 of the per_byte option, so that you can measure the throughput of other
799 aggregate values. For example, the per_byte option is now equivalent
800 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
802 The per_rcpt option has been generalized using the /count= mechanism
803 (though it's more complicated than the per_byte equivalence). When it is
804 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
805 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
806 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
807 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
808 Note that using per_rcpt with a non-readonly update mode in more than
809 one ACL will cause the recipients to be double-counted. (The per_mail
810 and per_byte options don't have this problem.)
812 The handling of very low rates has changed slightly. If the computed rate
813 is less than the event's count (usually one) then this event is the first
814 after a long gap. In this case the rate is set to the same as this event's
815 count, so that the first message of a spam run is counted properly.
817 The major new feature is a mechanism for counting the rate of unique
818 events. The new per_addr option counts the number of different
819 recipients that someone has sent messages to in the last time period. It
820 behaves like per_rcpt if all the recipient addresses are different, but
821 duplicate recipient addresses do not increase the measured rate. Like
822 the /count= option this is a general mechanism, so the per_addr option
823 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
824 example, measure the rate that a client uses different sender addresses
825 with the options per_mail/unique=$sender_address. There are further
826 details in the main documentation.
828 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
830 TF/06 Removed a few PCRE remnants.
832 TF/07 Automatically extract Exim's version number from tags in the git
833 repository when doing development or release builds.
835 PP/02 Raise smtp_cmd_buffer_size to 16kB.
836 Bugzilla 879. Patch from Paul Fisher.
838 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
839 Heavily based on revision 40f9a89a from Simon Arlott's tree.
842 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
844 PP/05 Variable $av_failed, true if the AV scanner deferred.
845 Bugzilla 1078. Patch from John Horne.
847 PP/06 Stop make process more reliably on build failure.
848 Bugzilla 1087. Patch from Heiko Schlittermann.
850 PP/07 Make maildir_use_size_file an _expandable_ boolean.
851 Bugzilla 1089. Patch from Heiko Schlittermann.
853 PP/08 Handle ${run} returning more data than OS pipe buffer size.
854 Bugzilla 1131. Patch from Holger Weiß.
856 PP/09 Handle IPv6 addresses with SPF.
857 Bugzilla 860. Patch from Wolfgang Breyha.
859 PP/10 GnuTLS: support TLS 1.2 & 1.1.
861 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
864 PP/11 match_* no longer expand right-hand-side by default.
865 New compile-time build option, EXPAND_LISTMATCH_RHS.
866 New expansion conditions, "inlist", "inlisti".
868 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
870 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
872 PP/14 fix log_write() format string regression from TF/03.
873 Bugzilla 1152. Patch from Dmitry Isaikin.
879 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
881 PP/02 Harmonised TLS library version reporting; only show if debugging.
882 Layout now matches that introduced for other libraries in 4.74 PP/03.
884 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
886 PP/04 New "dns_use_edns0" global option.
888 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
891 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
892 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
894 TK/01 Updated PolarSSL code to 0.14.2.
895 Bugzilla 1097. Patch from Andreas Metzler.
897 PP/07 Catch divide-by-zero in ${eval:...}.
900 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
903 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
904 format-string attack -- SECURITY: remote arbitrary code execution.
906 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
907 time unintentionally subject to list matching rules, letting the header
908 cause arbitrary Exim lookups (of items which can occur in lists, *not*
909 arbitrary string expansion). This allowed for information disclosure.
911 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
912 INT_MIN/-1 -- value coerced to INT_MAX.
918 NM/01 Workround for PCRE version dependency in version reporting
921 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
922 This fixes portability to compilers other than gcc, notably
923 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
925 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
926 makefiles for portability to HP-UX and POSIX correctness.
928 PP/01 Permit LOOKUP_foo enabling on the make command-line.
929 Also via indented variable definition in the Makefile.
930 (Debugging by Oliver Heesakkers).
932 PP/02 Restore caching of spamd results with expanded spamd_address.
933 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
935 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
936 Improves build reliability. Fix from: Frank Elsner
938 NM/02 Fix wide character breakage in the rfc2047 coding
939 Fixes bug 1064. Patch from Andrey N. Oktyabrski
941 NM/03 Allow underscore in dnslist lookups
942 Fixes bug 1026. Patch from Graeme Fowler
944 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
945 Code patches from Adam Ciarcinski of NetBSD.
947 NM/04 Fixed exiqgrep to cope with mailq missing size issue
950 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
951 is logged, to avoid truncation. Patch from John Horne.
953 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
954 Patch from Jakob Hirsch.
956 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
957 SQL string expansion failure details.
958 Patch from Andrey Oktyabrski.
960 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
961 Patch from Simon Arlott.
963 PP/09 New lookups functionality failed to compile on old gcc which rejects
964 extern declarations in function scope.
965 Patch from Oliver Fleischmann
967 PP/10 Use sig_atomic_t for flags set from signal handlers.
968 Check getgroups() return and improve debugging.
969 Fixed developed for diagnosis in bug 927 (which turned out to be
972 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
973 Patch from Mark Zealey.
975 PP/12 Bugzilla 1056: Improved spamd server selection.
976 Patch from Mark Zealey.
978 PP/13 Bugzilla 1086: Deal with maildir quota file races.
979 Based on patch from Heiko Schlittermann.
981 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
982 Patch from Uwe Doering, sign-off by Michael Haardt.
984 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
985 variable declaration deep within a block. Bug and patch from
988 PP/15 lookups-Makefile IRIX compatibilty coercion.
990 PP/16 Make DISABLE_DKIM build knob functional.
992 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
993 Patch by Simon Arlott
995 TF/03 Fix valgrind.h portability to C89 compilers that do not support
996 variable argument macros. Our copy now differs from upstream.
1002 TF/01 Failure to get a lock on a hints database can have serious
1003 consequences so log it to the panic log.
1005 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1006 controlled using the smtp_confirmation log selector.
1008 TF/03 Include the error message when we fail to unlink a spool file.
1010 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1011 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1012 for maintaining out-of-tree patches for some time.
1014 PP/01 Bugzilla 139: Documentation and portability issues.
1015 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1016 Handle per-OS dynamic-module compilation flags.
1018 PP/02 Let /dev/null have normal permissions.
1019 The 4.73 fixes were a little too stringent and complained about the
1020 permissions on /dev/null. Exempt it from some checks.
1021 Reported by Andreas M. Kirchwitz.
1023 PP/03 Report version information for many libraries, including
1024 Exim version information for dynamically loaded libraries. Created
1025 version.h, now support a version extension string for distributors
1026 who patch heavily. Dynamic module ABI change.
1028 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1029 privilege escalation vulnerability whereby the Exim run-time user
1030 can cause root to append content of the attacker's choosing to
1033 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1036 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1037 If dropping privileges for untrusted macros, we disabled normal logging
1038 on the basis that it would fail; for the Exim run-time user, this is not
1039 the case, and it resulted in successful deliveries going unlogged.
1040 Fixed. Reported by Andreas Metzler.
1046 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1047 only prepend for the Resent-* case. Fixes regression introduced in
1048 Exim 4.70 by NM/22 for Bugzilla 607.
1050 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1051 increasing numbers of administrators be bitten by this.
1053 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1055 PP/03 Bugzilla 994: added openssl_options main configuration option.
1057 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1059 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1061 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1063 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1064 without a peer certificate, leading to a segfault because of an
1065 assumption that peers always have certificates. Be a little more
1066 paranoid. Problem reported by Martin Tscholak.
1068 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1069 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1070 NB: ClamAV planning to remove STREAM in "middle of 2010".
1071 CL also introduces -bmalware, various -d+acl logging additions and
1072 more caution in buffer sizes.
1074 PP/09 Implemented reverse_ip expansion operator.
1076 PP/10 Bugzilla 937: provide a "debug" ACL control.
1078 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1080 PP/12 Bugzilla 973: Implement --version.
1082 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1084 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1086 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1088 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1089 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1090 ignore trailing whitespace.
1092 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1094 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1097 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1098 Notification from Dr Andrew Aitchison.
1100 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1101 ExtendedDetectionInfo response format.
1102 Notification from John Horne.
1104 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1107 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1108 XSL and documented dependency on system catalogs, with examples of how
1111 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1114 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1115 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1116 configuration file which is writeable by the Exim user or group.
1118 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1119 of configuration files to cover files specified with the -C option if
1120 they are going to be used with root privileges, not just the default
1123 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1124 option (effectively making it always true).
1126 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1127 files to be used while preserving root privileges.
1129 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1130 that rogue child processes cannot use them.
1132 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1133 run-time user, instead of root.
1135 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1136 Exim run-time user without dropping privileges.
1138 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1139 result string, instead of calling string_vformat() twice with the same
1142 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1143 for other users. Others should always drop root privileges if they use
1144 -C on the command line, even for a whitelisted configure file.
1146 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1148 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1154 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1155 $data_path, and $header_path variables; fixed documentation bugs and
1158 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1159 exipick to access non-standard spools, including the "frozen" queue
1162 NM/01 Bugzilla 965: Support mysql stored procedures.
1163 Patch from Alain Williams
1165 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1167 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1168 Patch from Andreas Metzler
1170 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1171 Patch from Kirill Miazine
1173 NM/05 Bugzilla 671: Added umask to procmail example.
1175 JJ/03 installed exipick 20100323.0, fixing doc bug
1177 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1178 directory. Notification and patch from Dan Rosenberg.
1180 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1182 TK/02 Improve log output when DKIM signing operation fails.
1184 MH/01 Treat the transport option dkim_domain as a colon separated
1185 list, not as a single string, and sign the message with each element,
1186 omitting multiple occurences of the same signer.
1188 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1189 Bugzilla 985, 986. Patch by Simon Arlott
1191 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1192 Patch by Simon Arlott
1194 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1195 MBX locking. Notification from Dan Rosenberg.
1201 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1203 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1205 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1207 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1209 NM/04 Bugzilla 915: Flag broken perl installation during build.
1215 TK/01 Added patch by Johannes Berg that expands the main option
1216 "spamd_address" if it starts with a dollar sign.
1218 TK/02 Write list of recipients to X-Envelope-Sender header when building
1219 the mbox-format spool file for content scanning (suggested by Jakob
1222 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1223 (http://www.dcc-servers.net/) support via dccifd. Activated by
1224 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1226 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1227 by Mark Daniel Reidel <mr@df.eu>.
1229 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1230 When building exim an external PCRE library is now needed -
1231 PCRE is a system library on the majority of modern systems.
1232 See entry on PCRE_LIBS in EDITME file.
1234 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1235 conversation. Added nologin parameter to request.
1236 Patch contributed by Kirill Miazine.
1238 TF/01 Do not log submission mode rewrites if they do not change the address.
1240 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1242 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1243 log files in place. Contributed by Roberto Lima.
1245 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1247 TF/03 Bugzilla 615: When checking the local_parts router precondition
1248 after a local_part_suffix or local_part_prefix option, Exim now
1249 does not use the address's named list lookup cache, since this
1250 contains cached lookups for the whole local part.
1252 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1253 Robert Millan. Documentation is in experimental-spec.txt.
1255 TF/04 Bugzilla 668: Fix parallel build (make -j).
1257 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1259 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1260 Patch provided by Jan Srzednicki.
1262 TF/05 Leading white space used to be stripped from $spam_report which
1263 wrecked the formatting. Now it is preserved.
1265 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1266 that they are available at delivery time.
1268 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1270 TF/08 TLS error reporting now respects the incoming_interface and
1271 incoming_port log selectors.
1273 TF/09 Produce a more useful error message if an SMTP transport's hosts
1274 setting expands to an empty string.
1276 NM/06 Bugzilla 744: EXPN did not work under TLS.
1277 Patch provided by Phil Pennock.
1279 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1280 Patch provided by Richard Godbee.
1282 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1283 acl_smtp_notquit, added index entry.
1285 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1286 Patch provided by Eugene Bujak.
1288 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1289 accept(). Patch provided by Maxim Dounin.
1291 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1292 Patch provided by Phil Pennock.
1294 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1296 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1297 Patch provided by Brad "anomie" Jorsch.
1299 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1300 Patch provided by Dean Brooks.
1302 TK/05 Add native DKIM support (does not depend on external libraries).
1304 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1305 Patch provided by Graeme Fowler.
1307 NM/16 Bugzilla 851: Documentation example syntax fix.
1309 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1311 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1314 NM/19 Bugzilla 745: TLS version reporting.
1315 Patch provided by Phil Pennock.
1317 NM/20 Bugzilla 167: bool: condition support.
1318 Patch provided by Phil Pennock.
1320 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1321 clients. Patch provided by Phil Pennock.
1323 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1324 Patch provided by Brad "anomie" Jorsch.
1326 NM/23 Bugzilla 687: Fix misparses in eximstats.
1327 Patch provided by Heiko Schlittermann.
1329 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1330 Patch provided by Heiko Schlittermann.
1332 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1333 plus update to original patch.
1335 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1337 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1338 Patch provided by David Brownlee.
1340 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1342 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1344 NM/30 Bugzilla 888: TLS documentation bugfixes.
1346 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1348 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1349 Unlike the original bugzilla I have changed all shell scripts in src tree.
1351 NM/33 Bugzilla 898: Transport filter timeout fix.
1352 Patch by Todd Rinaldo.
1354 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1355 Patch by Serge Demonchaux.
1357 NM/35 Bugzilla 39: Base64 decode bug fixes.
1358 Patch by Jakob Hirsch.
1360 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1362 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1364 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1366 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1372 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1373 ALT-N's libdkim that I have put here:
1374 http://duncanthrax.net/exim-experimental/
1376 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1377 were called 'true' and it seems that C99 defines that as a reserved
1378 keyword to be used with 'bool' variable types. That means you could
1379 not include C99-style headers which use bools without triggering
1380 build errors in sieve.c.
1382 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1383 as mailq or other aliases. Changed the --help handling significantly
1384 to do whats expected. exim_usage() emits usage/help information.
1386 SC/01 Added the -bylocaldomain option to eximstats.
1388 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1390 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1392 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1398 PH/01 Another patch from the Sieve maintainer.
1400 PH/02 When an IPv6 address is converted to a string for single-key lookup
1401 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1402 dots are used instead of colons so that keys in lsearch files need not
1403 contain colons. This was done some time before quoting was made available
1404 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1405 (notated using the quote facility) so as to distinguish them from IPv4
1406 keys. This meant that lookups for IP addresses in host lists did not work
1407 for iplsearch lookups.
1409 This has been fixed by arranging for IPv6 addresses to be expressed with
1410 colons if the lookup type is iplsearch. This is not incompatible, because
1411 previously such lookups could never work.
1413 The situation is now rather anomolous, since one *can* have colons in
1414 ordinary lsearch keys. However, making the change in all cases is
1415 incompatible and would probably break a number of configurations.
1417 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1420 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1421 conversion specification without a maximum field width, thereby enabling
1422 a rogue spamd server to cause a buffer overflow. While nobody in their
1423 right mind would setup Exim to query an untrusted spamd server, an
1424 attacker that gains access to a server running spamd could potentially
1425 exploit this vulnerability to run arbitrary code as the Exim user.
1427 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1428 $primary_hostname instead of what libspf2 thinks the hosts name is.
1430 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1431 a directory entry by the name of the lookup key. Previously, if a
1432 symlink pointed to a non-existing file or a file in a directory that
1433 Exim lacked permissions to read, a lookup for a key matching that
1434 symlink would fail. Now it is enough that a matching directory entry
1435 exists, symlink or not. (Bugzilla 503.)
1437 PH/03 The body_linecount and body_zerocount variables are now exported in the
1440 PH/04 Added the $dnslist_matched variable.
1442 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1443 This means they are set thereafter only if the connection becomes
1446 PH/06 Added the client_condition to authenticators so that some can be skipped
1447 by clients under certain conditions.
1449 PH/07 The error message for a badly-placed control=no_multiline_responses left
1450 "_responses" off the end of the name.
1452 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1454 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1455 (without spaces) instead of just copying the configuration text.
1457 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1459 PH/11 Added $max_received_linelength.
1461 PH/12 Added +ignore_defer and +include_defer to host lists.
1463 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1464 way in which PCRE > 7.0 is built.
1466 PH/14 Implemented queue_only_load_latch.
1468 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1469 MAIL command. The effect was to mangle the value on 64-bit systems.
1471 PH/16 Another patch from the Sieve maintainer.
1473 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1475 PH/18 If a system quota error occurred while trying to create the file for
1476 a maildir delivery, the message "Mailbox is full" was not appended to the
1477 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1478 only to a quota excession during the actual writing of the file.
1480 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1481 characters?) which causes problems in log lines. The DN values are now
1482 passed through string_printing() before being added to log lines.
1484 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1485 and InterBase are left for another time.)
1487 PH/21 Added message_body_newlines option.
1489 PH/22 Guard against possible overflow in moan_check_errorcopy().
1491 PH/23 POSIX allows open() to be a macro; guard against that.
1493 PH/24 If the recipient of an error message contained an @ in the local part
1494 (suitably quoted, of course), incorrect values were put in $domain and
1495 $local_part during the evaluation of errors_copy.
1501 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1502 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1505 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1506 issue a MAIL command.
1508 PH/02 In an ACL statement such as
1510 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1512 if a client was not listed at all, or was listed with a value other than
1513 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1514 the condition was not true (as it should be), so access was not denied.
1515 The bug was that the ! inversion was incorrectly passed on to the second
1516 item. This has been fixed.
1518 PH/03 Added additional dnslists conditions == and =& which are different from
1519 = and & when the dns lookup returns more than one IP address.
1521 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1522 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1524 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1525 FSYNC, which compiles an option called disable_fsync that allows for
1526 bypassing fsync(). The documentation is heavily laced with warnings.
1528 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1530 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1531 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1532 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1533 including adding "make clean"; (3) Added -fPIC when compiling the test
1534 dynamically loaded module, to get rid of a warning.
1536 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1537 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1538 = 0s. The bug is otherwise harmless.
1540 PH/07 There was a bug in the dovecot authenticator such that the value of
1541 $auth1 could be overwritten, and so not correctly preserved, after a
1542 successful authentication. This usually meant that the value preserved by
1543 the server_setid option was incorrect.
1545 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1547 PH/09 Installed PCRE release 7.0.
1549 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1550 run for batched SMTP input. It is now run at the start of every message
1551 in the batch. While fixing this I discovered that the process information
1552 (output by running exiwhat) was not always getting set for -bs and -bS
1553 input. This is fixed, and it now also says "batched" for BSMTP.
1555 PH/11 Added control=no_pipelining.
1557 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1558 patch, slightly modified), and move the expansion of helo_data till after
1559 the connection is made in the smtp transport (so it can use these
1562 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1564 PH/14 Added log_selector = +pid.
1566 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1568 PH/16 Add ${if forany and ${if forall.
1570 PH/17 Added dsn_from option to vary the From: line in DSNs.
1572 PH/18 Flush SMTP output before performing a callout, unless control =
1573 no_callout_flush is set.
1575 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1576 was true (the default) a successful delivery failed to delete the retry
1577 item, thus causing premature timeout of the address. The bug is now
1580 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1582 PH/21 Long custom messages for fakedefer and fakereject are now split up
1583 into multiline reponses in the same way that messages for "deny" and
1584 other ACL rejections are.
1586 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1587 with slight modification.
1589 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1590 draft, changing the syntax and factoring some duplicate code".
1592 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1593 for deliveries of the second and subsequent messages over the same SMTP
1596 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1597 ${reduce, with only minor "tidies".
1599 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1601 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1602 expansion side effects.
1604 PH/27 When a message times out after an over-quota error from an Exim-imposed
1605 quota, the bounce message says "mailbox is full". This message was not
1606 being given when it was a system quota that was exceeded. It now should
1609 MH/03 Made $recipients available in local_scan(). local_scan() already has
1610 better access to the recipient list through recipients_list[], but
1611 $recipients can be useful in postmaster-provided expansion strings.
1613 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1614 in the case of a MAIL command with additional options following the
1615 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1616 were accidentally chopped off.
1618 PH/29 SMTP synchronization checks are implemented when a command is read -
1619 there is a check that no more input is waiting when there shouldn't be
1620 any. However, for some commands, a delay in an ACL can mean that it is
1621 some time before the response is written. In this time, more input might
1622 arrive, invalidly. So now there are extra checks after an ACL has run for
1623 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1624 pipelining has not been advertised.
1626 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1628 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1629 returned in the SMTP response when smtp_return_error_details was false.
1630 This has been fixed.
1632 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1633 instead of the C I/O that was originally supplied, because problems were
1634 reported on Solaris.
1636 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1637 Exim which did not show up earlier: it was assuming that a call to
1638 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1639 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1640 was a macro that became an assignment, so it seemed to work. This has
1641 changed to a proper function call with a void return, hence the compile
1642 error. Exim's code has been fixed.
1644 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1647 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1649 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1651 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1652 because it turns out to be too big or there is a timeout) there is no
1653 "Completed" line in the log. When some messages of this type were
1654 selected by exigrep, they were listed as "not completed". Others were
1655 picked up by some special patterns. I have improved the selection
1656 criteria to be more general.
1658 PH/38 The host_find_failed option in the manualroute router can now be set
1659 to "ignore", to completely ignore a host whose IP address cannot be
1660 found. If all hosts are ignored, the behaviour is controlled by the new
1661 host_all_ignored option.
1663 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1664 homing or because of multiple MX records with /mx) generated more than
1665 one IP address, and the following item turned out to be the local host,
1666 all the secondary addresses of the first item were incorrectly removed
1667 from the list, along with the local host and any following hosts (which
1668 is what is supposed to happen).
1670 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1671 whoever called Exim into the -H file. In the case of the daemon it was
1672 behaving confusingly. When first started, it used values for whoever
1673 started the daemon, but after a SIGHUP it used the Exim user (because it
1674 calls itself on a restart). I have changed the code so that it now always
1677 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1678 message are rejected with the same error (e.g. no authentication or bad
1679 sender address), and a DATA command is nevertheless sent (as can happen
1680 with PIPELINING or a stupid MUA), the error message that was given to the
1681 RCPT commands is included in the rejection of the DATA command. This is
1682 intended to be helpful for MUAs that show only the final error to their
1685 PH/42 Another patch from the Sieve maintainer.
1687 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1688 Eximstats - Fixed some broken HTML links and added missing column headers
1690 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1691 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1693 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1695 PH/43 Yet another patch from the Sieve maintainer.
1697 PH/44 I found a way to check for a TCP/IP connection going away before sending
1698 the response to the final '.' that terminates a message, but only in the
1699 case where the client has not sent further data following the '.'
1700 (unfortunately, this is allowed). However, in many cases there won't be
1701 any further data because there won't be any more messages to send. A call
1702 to select() can be used: if it shows that the input is "ready", there is
1703 either input waiting, or the socket has been closed. An attempt to read
1704 the next input character can distinguish the two cases. Previously, Exim
1705 would have sent an OK response which the client would never have see.
1706 This could lead to message repetition. This fix should cure that, at
1707 least in a lot of common cases.
1709 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1710 advertised in response to EHLO.
1716 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1717 fixed by 4.65/MH/01 (is this a record?) are fixed:
1719 (i) An empty string was always treated as zero by the numeric comparison
1720 operators. This behaviour has been restored.
1722 (ii) It is documented that the numeric comparison operators always treat
1723 their arguments as decimal numbers. This was broken in that numbers
1724 starting with 0 were being interpreted as octal.
1726 While fixing these problems I realized that there was another issue that
1727 hadn't been noticed. Values of message_size_limit (both the global option
1728 and the transport option) were treated as octal if they started with 0.
1729 The documentation was vague. These values are now always treated as
1730 decimal, and I will make that clear in the documentation.
1736 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1737 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1740 MH/01 Don't check that the operands of numeric comparison operators are
1741 integers when their expansion is in "skipping" mode (fixes bug
1742 introduced by 4.64-PH/07).
1744 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1745 child addresses, Exim now panics and dies. Previously, because the count
1746 is held in a short int, deliveries were likely to be lost. As such a
1747 large number of recipients for a single message is ridiculous
1748 (performance will be very, very poor), I have chosen to impose a limit
1749 rather than extend the field.
1755 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1756 leftover -K file (the existence of which was triggered by #402).
1757 While we were at it, introduced process PID as part of the -K
1758 filename. This should rule out race conditions when creating
1761 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1762 processing considerably. Previous code took too long for large mails,
1763 triggering a timeout which in turn triggers #401.
1765 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1766 in the DK code in transports.c. sendfile() is not really portable,
1767 hence the _LINUX specificness.
1769 TF/01 In the add_headers option to the mail command in an Exim filter,
1770 there was a bug that Exim would claim a syntax error in any
1771 header after the first one which had an odd number of characters
1774 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1775 callout verification, Exim cached a "reject" for the entire domain. This
1776 is correct for most verifications, but it is not correct for a recipient
1777 verification with use_sender or use_postmaster set, because in that case
1778 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1779 case of MAIL FROM:<> rejection from other early rejections (e.g.
1780 rejection of HELO). When verifying a recipient using a non-null MAIL
1781 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1782 Whatever the result of the callout, the value of the domain cache is
1783 left unchanged (for any other kind of callout, getting as far as trying
1784 RCPT means that the domain itself is ok).
1786 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1789 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1790 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1791 session, the daemon ignores ECONNECT errors and logs others; it now
1792 ignores EPIPE as well.
1794 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1795 (quoted-printable decoding).
1797 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1798 later the small subsequent patch to fix an introduced bug.
1800 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1802 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1804 PH/08 An error is now given if message_size_limit is specified negative.
1806 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1807 to be given (somewhat) arbitrary names.
1809 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1812 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1813 miscellaneous code fixes
1815 PH/10 Added the log_reject_target ACL modifier to specify where to log
1818 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1819 hostname. This is wrong, because it relates to the incoming message (and
1820 probably the interface on which it is arriving) and not to the outgoing
1821 callout (which could be using a different interface). This has been
1822 changed to use the value of the helo_data option from the smtp transport
1823 instead - this is what is used when a message is actually being sent. If
1824 there is no remote transport (possible with a router that sets up host
1825 addresses), $smtp_active_hostname is used.
1827 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1828 tweaks were necessary in order to get it to work (see also 21 below):
1829 (a) The code assumed that strncpy() returns a negative number on buffer
1830 overflow, which isn't the case. Replaced with Exim's string_format()
1832 (b) There were several signed/unsigned issues. I just did the minimum
1833 hacking in of casts. There is scope for a larger refactoring.
1834 (c) The code used strcasecmp() which is not a standard C function.
1835 Replaced with Exim's strcmpic() function.
1836 (d) The code set only $1; it now sets $auth1 as well.
1837 (e) A simple test gave the error "authentication client didn't specify
1838 service in request". It would seem that Dovecot has changed its
1839 interface. Fortunately there's a specification; I followed it and
1840 changed what the client sends and it appears to be working now.
1842 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1845 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1846 address A is aliased to B and C, where B exists and C does not. Without
1847 -v the output is "A verified" because verification stops after a
1848 successful redirection if more than one address is generated. However,
1849 with -v the child addresses are also verified. Exim was outputting "A
1850 failed to verify" and then showing the successful verification for C,
1851 with its parentage. It now outputs "B failed to verify", showing B's
1852 parentage before showing the successful verification of C.
1854 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1855 look up a TXT record in a specific list after matching in a combined
1858 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1859 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1860 they consult the DNS. I had assumed they would set it the way they
1861 wanted; and indeed my experiments on Linux seem to show that in some
1862 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1863 To be on the safe side, however, I have now made the interface to
1864 host_find_byname() similar to host_find_bydns(), with an argument
1865 containing the DNS resolver options. The host_find_byname() function now
1866 sets these options at its start, just as host_find_bydns() does. The smtp
1867 transport options dns_qualify_single and dns_search_parents are passed to
1868 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1869 of host_find_byname() use the default settings of RES_DEFNAMES
1870 (qualify_single) but not RES_DNSRCH (search_parents).
1872 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1873 spool_read_header() do less string testing, by means of a preliminary
1874 switch on the second character of optional "-foo" lines. (This is
1875 overdue, caused by the large number of possibilities that now exist.
1876 Originally there were few.) While I was there, I also converted the
1877 str(n)cmp tests so they don't re-test the leading "-" and the first
1878 character, in the hope this might squeeze out yet more improvement.
1880 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1881 flag allowing group syntax was set by the header_syntax check but not
1882 turned off, possible causing trouble later; (2) The flag was not being
1883 set at all for the header_verify test, causing "group"-style headers to
1884 be rejected. I have now set it in this case, and also caused header_
1885 verify to ignore an empty address taken from a group. While doing this, I
1886 came across some other cases where the code for allowing group syntax
1887 while scanning a header line wasn't quite right (mostly, not resetting
1888 the flag correctly in the right place). These bugs could have caused
1889 trouble for malformed header lines. I hope it is now all correct.
1891 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1892 with the "reply" argument non-NULL. The code, however (which originally
1893 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1894 but it didn't always do it. This confused somebody who was copying the
1895 code for some other use. I have removed all the tests.
1897 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1898 feature that was used to support insecure browsers during the U.S. crypto
1899 embargo. It requires special client support, and Exim is probably the
1900 only MTA that supported it -- and would never use it because real RSA is
1901 always available. This code has been removed, because it had the bad
1902 effect of slowing Exim down by computing (never used) parameters for the
1903 RSA_EXPORT functionality.
1905 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1906 authenticator to fail if there's a tab character in the incoming data
1907 (there should never be unless someone is messing about, as it's supposed
1908 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1909 if the connection is using TLS or if the remote IP is the same as the
1910 local IP, and the "valid-client-cert option" if a client certificate has
1913 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1914 authenticators. This can be used for authorization after authentication
1915 succeeds. (In the case of plaintext, it servers for both authentication
1918 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1919 if any retry times were supplied.
1921 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1922 connection, where there is no client IP address to check. In this
1923 situation, the verify now always succeeds.
1925 PH/25 Applied John Jetmore's -Mset patch.
1927 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1929 PH/27 In a string expansion for a processed (not raw) header when multiple
1930 headers of the same name were present, leading whitespace was being
1931 removed from all of them, but trailing whitespace was being removed only
1932 from the last one. Now trailing whitespace is removed from each header
1933 before concatenation. Completely empty headers in a concatenation (as
1934 before) are ignored.
1936 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1937 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1939 PH/29 [Removed. This was a change that I later backed out, and forgot to
1940 correct the ChangeLog entry (that I had efficiently created) before
1941 committing the later change.]
1943 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1944 address errors (4xx response to RCPT) over the same connection as other
1945 messages routed to the same hosts. Such deliveries are always "forced",
1946 so retry times are not inspected. This resulted in far too many retries
1947 for the affected addresses. The effect occurred only when there were more
1948 hosts than the hosts_max_try setting in the smtp transport when it had
1949 the 4xx errors. Those hosts that it had tried were not added to the list
1950 of hosts for which the message was waiting, so if all were tried, there
1951 was no problem. Two fixes have been applied:
1953 (i) If there are any address or message errors in an SMTP delivery, none
1954 of the hosts (tried or untried) are now added to the list of hosts
1955 for which the message is waiting, so the message should not be a
1956 candidate for sending over the same connection that was used for a
1957 successful delivery of some other message. This seems entirely
1958 reasonable: after all the message is NOT "waiting for some host".
1959 This is so "obvious" that I'm not sure why it wasn't done
1960 previously. Hope I haven't missed anything, but it can't do any
1961 harm, as the worst effect is to miss an optimization.
1963 (ii) If, despite (i), such a delivery is accidentally attempted, the
1964 routing retry time is respected, so at least it doesn't keep
1965 hammering the server.
1967 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1968 in ${readsocket because some servers need this prod.
1970 PH/32 Added some extra debug output when updating a wait-xxx database.
1972 PH/33 The hint "could be header name not terminated by colon", which has been
1973 given for certain expansion errors for a long time, was not being given
1974 for the ${if def:h_colon_omitted{... case.
1976 PH/34 The spec says: "With one important exception, whenever a domain list is
1977 being scanned, $domain contains the subject domain." There was at least
1978 one case where this was not true.
1980 PH/35 The error "getsockname() failed: connection reset by peer" was being
1981 written to the panic log as well as the main log, but it isn't really
1982 panic-worthy as it just means the connection died rather early on. I have
1983 removed the panic log writing for the ECONNRESET error when getsockname()
1986 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1987 runs only) independently of the message's sender address. This meant
1988 that, if the 4xx error was in fact related to the sender, a different
1989 message to the same recipient with a different sender could confuse
1990 things. In particualar, this can happen when sending to a greylisting
1991 server, but other circumstances could also provoke similar problems.
1992 I have changed the default so that the retry time for these errors is now
1993 based a combination of the sender and recipient addresses. This change
1994 can be overridden by setting address_retry_include_sender=false in the
1997 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1998 remote server are returned as part of bounce messages. This was not
1999 happening for LMTP over a pipe (the lmtp transport), but now it is the
2000 same for both kinds of LMTP.
2002 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2003 in header lines that were in fact CNAMEs. This is no longer the case.
2005 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2006 and queue runs started by the daemon processed all messages. This has
2007 been fixed so that -R and -S can now usefully be given with -q<time>.
2009 PH/40 Import PCRE release 6.7 (fixes some bugs).
2011 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2013 PH/42 Give an error if -q is specified more than once.
2015 PH/43 Renamed the variables $interface_address and $interface_port as
2016 $received_ip_address and $received_port, to make it clear that these
2017 values apply to message reception, and not to the outgoing interface when
2018 a message is delivered. (The old names remain recognized, of course.)
2020 PH/44 There was no timeout on the connect() call when using a Unix domain
2021 socket in the ${readsocket expansion. There now is.
2023 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2024 be meaningful with "accept".
2026 SC/01 Eximstats V1.43
2027 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2029 SC/02 Eximstats V1.44
2030 Use a glob alias rather than an array ref in the generated
2031 parser. This improves both readability and performance.
2033 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2034 Collect SpamAssassin and rejection statistics.
2035 Don't display local sender or destination tables unless
2036 there is data to show.
2037 Added average volumes into the top table text output.
2039 SC/04 Eximstats V1.46
2040 Collect data on the number of addresses (recipients)
2041 as well as the number of messages.
2043 SC/05 Eximstats V1.47
2044 Added 'Message too big' to the list of mail rejection
2045 reasons (thanks to Marco Gaiarin).
2047 SC/06 Eximstats V1.48
2048 Mainlog lines which have GMT offsets and are too short to
2049 have a flag are now skipped.
2051 SC/07 Eximstats V1.49 (Alain Williams)
2052 Added the -emptyok flag.
2054 SC/08 Eximstats V1.50
2055 Fixes for obtaining the IP address from reject messages.
2057 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2058 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2059 whitesspace changes from 4.64-PH/27
2061 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2064 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2065 are found, allow negative numbers in numeric criteria)
2067 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2069 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2070 to match changes made in 4.64-PH/43
2072 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2074 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2075 "domain" is actually an IP address, and force a failure. This locks out
2076 those revolvers/nameservers that support "A-for-A" lookups, in
2077 contravention of the specifications.
2079 PH/48 When a host name was looked up from an IP address, and the subsequent
2080 forward lookup of the name timed out, the host name was left in
2081 $sender_host_name, contrary to the specification.
2083 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2084 restricted to single-key lookups, Exim was not diagnosing an error if
2085 * or *@ was used with a query-style lookup.
2087 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2089 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2090 long ago, but noone interested enough thought of it. Let's just say that
2091 the "1.1" means that there are some new functions that weren't there at
2092 some point in the past.
2094 PH/51 Error processing for expansion failure of helo_data from an smtp
2095 transport during callout processing was broken.
2097 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2098 tested/used via the -bh/-bhc/-bs options.
2100 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2101 bug, fixed in subsequent PCRE releases).
2103 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2104 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2106 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2112 SC/01 Use a glob alias rather than an array ref in eximstats generated
2113 parser. This improves both readability and performance.
2115 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2116 Don't display local sender or destination tables in eximstats unless
2117 there is data to show.
2118 Added average volumes into the eximstats top table text output.
2120 SC/03 Collect data on the number of addresses (recipients) as well
2121 as the number of messages in eximstats.
2123 TF/01 Correct an error in the documentation for the redirect router. Exim
2124 does (usually) call initgroups() when daemonizing.
2126 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2127 with consistent privilege compared to when running as a daemon.
2129 TF/03 Note in the spec that $authenticated_id is not set for local
2130 submissions from trusted users.
2132 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2133 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2135 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2136 by adding some example configuration directives to the default
2137 configuration file. A little bit of work is required to uncomment the
2138 directives and define how usernames and passwords are checked, but
2139 there is now a framework to start from.
2141 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2142 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2143 without this. I don't know how relevant this is to other LDAP libraries.
2145 PH/02 Add the verb name to the "unknown ACL verb" error.
2147 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2149 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2151 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2152 directory not expanded when it should be if an expanded home directory
2153 was set for the address (which is overridden by the transport).
2155 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2158 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2159 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2160 because it is too late at that time, and has no effect.
2162 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2163 security issue with \' (bugzilla #107). I could not use the
2164 PQescapeStringConn() function, because it needs a PGconn value as one of
2167 PH/08 When testing addresses using -bt, indicate those final addresses that
2168 are duplicates that would not cause an additional delivery. At least one
2169 person was confused, thinking that -bt output corresponded to deliveries.
2170 (Suppressing duplicates isn't a good idea as you lose the information
2171 about possibly different redirections that led to the duplicates.)
2173 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2174 systems where poll() doesn't work, in particular OS X.
2176 PH/10 Added more information to debugging output for retry time not reached.
2178 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2179 operations in malware.c.
2181 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2184 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2185 syslog_duplication set false, log lines that would normally be written
2186 both the the main log and to the reject log were not written to syslog at
2189 PH/14 In the default configuration, change the use of "message" in ACL warn
2190 statements to "add_header".
2192 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2193 not followed by a command (e.g. "seen endif").
2195 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2196 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2199 PH/17 Added extra conditions to the default value of delay_warning_condition
2203 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2204 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2205 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2208 The Auto-Submitted: and various List- headers are standardised, whereas I
2209 don't think Precedence: ever was.
2211 PH/18 Refactored debugging code in route_finduser() to show more information,
2212 in particular, the error code if getpwnam() issues one.
2214 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2215 This is apparently needed in addition to the PH/07 change above to avoid
2216 any possible encoding problems.
2218 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2219 but not after initializing Perl.
2221 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2222 output them only if debugging. By default they are written stderr,
2223 apparently, which is not desirable.
2225 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2228 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2231 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2233 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2234 authenticated or an ident call has been made. Suppress the default
2235 values for $authenticated_id and $authenticated_sender (but permit -oMai
2236 and -oMas) when testing with -bh.
2238 PH/24 Re-jigged the order of the tests in the default configuration so that the
2239 tests for valid domains and recipients precede the DNS black list and CSA
2240 tests, on the grounds that those ones are more expensive.
2242 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2243 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2244 This bug exists in every version of Exim that I still have, right back to
2247 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2248 However, an attempt to turn on case-sensitivity in a regex key by
2249 including (?-i) didn't work because the subject string was already
2250 lowercased, and the effects were non-intuitive. It turns out that a
2251 one-line patch can be used to allow (?-i) to work as expected.
2257 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2258 other effects) broke the use of negated acl sub-conditions.
2260 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2263 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2264 "Deny" causes Exim to reject the incoming connection with a 554 error.
2265 Unfortunately, if there is a major crisis, such as a disk failure,
2266 tcp-wrappers gives "deny", whereas what one would like would be some
2267 kind of temporary error. A kludge has been added to help with this.
2268 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2269 554 error is used if errno is still zero or contains ENOENT (which occurs
2270 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2273 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2275 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2276 errors. Otherwise a message that provokes a temporary error (when other
2277 messages do not) can cause a whole host to time out.
2279 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2280 addresses were routed directly to files or pipes from a redirect router.
2281 File deliveries just didn't batch; pipe deliveries might have suffered
2284 PH/06 A failure to get a lock for a hints database would erroneously always say
2285 "Failed to get write lock", even when it was really a read lock.
2287 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2288 of 0600. This has been changed to use the value of the lockfile_mode
2289 option (which defaults to 0600).
2291 PH/08 Applied small patch from the Sieve maintainer.
2293 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2294 folder from quota calculations, a direct delivery into this folder messed
2295 up the contents of the maildirsize file. This was because the regex was
2296 used only to exclude .Trash (or whatever) when the size of the mailbox
2297 was calculated. There was no check that a delivery was happening into an
2298 excluded directory. This bug has been fixed by ignoring all quota
2299 processing for deliveries into excluded directories.
2301 PH/10 Added the maildirfolder_create_regex option to appendfile.
2307 PH/01 The code for finding all the local interface addresses on a FreeBSD
2308 system running IPv6 was broken. This may well have applied to all BSD
2309 systems, as well as to others that have similar system calls. The broken
2310 code found IPv4 interfaces correctly, but gave incorrect values for the
2311 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2312 that it would not match correctly against @[] and not recognize the IPv6
2315 PH/02 The ipliteral router was not recognizing addresses of the form user@
2316 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2318 PH/03 Added disable_ipv6.
2320 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2321 decoded form, because it is most often used to construct To: headers
2322 lines in autoreplies, and the decoded form may well be syntactically
2323 invalid. However, $reply_address has leading white space removed, and all
2324 newlines turned into spaces so that the autoreply transport does not
2327 PH/05 If group was specified without a user on a router, and no group or user
2328 was specified on a transport, the group from the router was ignored.
2330 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2331 for visible compile-time settings that can be used to change these
2332 numbers, for those that want even more. Backwards compatibility with old
2333 spool files has been maintained. However, going back to a previous Exim
2334 release will lost any variables that are in spool files.
2336 PH/07 Two small changes when running in the test harness: increase delay when
2337 passing a TCP/IP connection to a new process, in case the original
2338 process has to generate a bounce, and remove special handling of
2339 127.0.0.2 (sic), which is no longer necessary.
2341 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2342 be the same on different OS.
2344 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2347 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2348 whether --show-vars was specified or not
2350 JJ/02 exipick: Added support for new ACL variable spool format introduced
2353 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2354 syntactically invalid From: or Reply-to: line, and a filter used this to
2355 generate an autoreply, and therefore failed to obtain an address for the
2356 autoreply, Exim could try to deliver to a non-existent relative file
2357 name, causing unrelated and misleading errors. What now happens is that
2358 it logs this as a hard delivery error, but does not attempt to create a
2361 PH/11 The exinext utility has a -C option for testing purposes, but although
2362 the given file was scanned by exinext itself; it wasn't being passed on
2363 when Exim was called.
2365 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2366 an end-of-file indication when reading a command response.
2368 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2369 compiled. In many other places in Exim, IPv6 addresses are always
2370 recognized, so I have changed this. It also means that IPv4 domain
2371 literals of the form [IPV4:n.n.n.n] are now always recognized.
2373 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2374 used if the router is not running as root, for example, when verifying at
2375 ACL time, or when using -bh. The debugging output from this situation was
2376 non-existent - all you got was a failure to exec. I have made two
2379 (a) Failures to set uid/gid, the current directory, or a process leader
2380 in a subprocess such as that created by queryprogram now generate
2381 suitable debugging ouput when -d is set.
2383 (b) The queryprogram router detects when it is not running as root,
2384 outputs suitable debugging information if -d is set, and then runs
2385 the subprocess without attempting to change uid/gid.
2387 PH/15 Minor change to Makefile for building test_host (undocumented testing
2390 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2391 additional section of a DNS packet that returns MX or SRV records.
2392 Instead, it always explicitly searches for A/AAAA records. This avoids
2393 major problems that occur when a DNS server includes only records of one
2394 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2395 fixed another bug: if SRV records were looked up and the corresponding
2396 address records were *not* found in the additional section, the port
2397 values from the SRV records were lost.
2399 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2400 using the correct key (the original address) when searching the retry
2401 rules in order to find which one to use for generating the retry hint.
2403 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2404 adding the default one. Similarly, if it contains a Reply-To: header, the
2405 errors_reply_to option, if set, is not used.
2407 PH/19 When calculating a retry time, Exim used to measure the "time since
2408 failure" by looking at the "first failed" field in the retry record. Now
2409 it does not use this if it is later than than the arrival time of the
2410 message. Instead it uses the arrival time. This makes for better
2411 behaviour in cases where some deliveries succeed, thus re-setting the
2412 "first failed" field. An example is a quota failure for a huge message
2413 when small messages continue to be delivered. Without this change, the
2414 "time since failure" will always be short, possible causing more frequent
2415 delivery attempts for the huge message than are intended.
2416 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2418 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2419 $1, $2, $3) because the numerical variables can be reset during some
2420 expansion items (e.g. "match"), thereby losing the authentication data.
2422 PH/21 Make -bV show the size of off_t variables so that the test suite can
2423 decide whether to run tests for quotas > 2G.
2425 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2426 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2427 filecount value is greater than 2G or if a quota value is greater than 2G
2428 on a system where the size of off_t is not greater than 4, a panic error
2431 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2432 never match. The debug and -bh output now contains an explicit error
2433 message indicating a malformed IPv4 address or mask.
2435 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2436 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2437 PH/23 above applies.
2439 PH/25 Do not write to syslog when running in the test harness. The only
2440 occasion when this arises is a failure to open the main or panic logs
2441 (for which there is an explicit test).
2443 PH/26 Added the /no_tell option to "control=freeze".
2445 PH/27 If a host name lookup failed very early in a connection, for example, if
2446 the IP address matched host_lookup and the reverse lookup yielded a name
2447 that did not have a forward lookup, an error message of the form "no IP
2448 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2449 could be logged. Now it outputs the IP address instead of "NULL".
2451 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2452 allows the sender and the authenticated sender to be set when
2453 submitting a message from within Exim. Since child_open_exim() is
2454 documented for local_scan(), the new function should be too.
2456 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2457 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2458 results in an empty string is now treated as unset.
2460 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2462 PH/31 Added sender_verify_fail logging option.
2464 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2465 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2466 tidied the source and removed it altogether.
2468 PH/33 When a queue run was abandoned because the load average was too high, a
2469 log line was always written; now it is written only if the queue_run log
2470 selector is set. In addition, the log line for abandonment now contains
2471 information about the queue run such as the pid. This is always present
2472 in "start" and "stop" lines but was omitted from the "abandon" line.
2474 PH/34 Omit spaces between a header name and the colon in the error message that
2475 is given when verify = headers_syntax fails (if there are lots of them,
2476 the message gets confusing).
2478 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2479 names, as there are now some PTR records that contain slashes. This check
2480 is only to protect against broken name servers that fall over on strange
2481 characters, so the fact that it applies to all lookups doesn't matter.
2483 PH/36 Now that the new test suite is complete, we can remove some of the
2484 special code in Exim that was needed for the old test suite. For example,
2485 sorting DNS records because real resolvers return them in an arbitrary
2486 order. The new test suite's fake resolver always returns records in the
2489 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2490 bounces) except when queue_only is set, to avoid logging races between
2491 the different processes.
2493 PH/38 Panic-die if .include specifies a non-absolute path.
2495 PH/39 A tweak to the "H" retry rule from its user.
2497 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2498 a label. They prevented compilation on older perls.
2500 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2501 a warning to be raised on newish perls.
2503 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2504 on queue. Changes to match documented behaviour of showing count of
2505 messages matching specified criteria.
2507 PH/40 Changed the default ident timeout from 30s to 5s.
2509 PH/41 Added support for the use of login_cap features, on those BSD systems
2510 that have them, for controlling the resources used by pipe deliveries.
2512 PH/42 The content-scanning code uses fopen() to create files in which to put
2513 message data. Previously it was not paying any attention to the mode of
2514 the files. Exim runs with umask(0) because the rest of the code creates
2515 files with open(), and sets the required mode explicitly. Thus, these
2516 files were ending up world-writeable. This was not a big issue, because,
2517 being within the spool directory, they were not world-accessible. I have
2518 created a function called modefopen, which takes an additional mode
2519 argument. It sets umask(777), creates the file, chmods it to the required
2520 mode, then resets the umask. All the relevant calls to fopen() in the
2521 content scanning code have been changed to use this function.
2523 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2524 to 24 hours. This avoids potential overflow problems when processing G
2525 and H retry rules. I suspect nobody ever tinkers with this value.
2527 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2529 PH/45 When the plaintext authenticator is running as a client, the server's
2530 challenges are checked to ensure they are valid base64 strings. By
2531 default, the authentication attempt is cancelled if an invalid string is
2532 received. Setting client_ignore_invalid_base64 true ignores these errors.
2533 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2534 they are received. Thus, the responses can be made to depend on the
2535 challenges. If an invalid string is ignored, an empty string is placed in
2538 PH/46 Messages that are created by the autoreply transport now contains a
2539 References: header, in accordance with RFCs 2822 and 3834.
2541 PH/47 Added authenticated_sender_force to the smtp transport.
2543 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2545 PH/49 Installed latest patch from the Sieve maintainer.
2547 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2548 was also set, the appendfile transport was unnecessarily scanning a
2549 directory of message files (e.g. for maildir delivery) to find the count
2550 of files (along with the size), even though it did not need this
2551 information. It now does the scan only if it needs to find either the
2552 size of the count of files.
2554 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2556 PH/52 Two bugs concerned with error handling when the smtp transport is
2559 (i) Exim was not creating retry information for temporary errors given
2560 for individual recipients after the DATA command when the smtp transport
2561 was used in LMTP mode. This meant that they could be retried too
2562 frequently, and not timed out correctly.
2564 (ii) Exim was setting the flag that allows error details to be returned
2565 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2566 recipients that were returned after the DATA command.
2568 PH/53 This is related to PH/52, but is more general: for any failing address,
2569 when detailed error information was permitted to be returned to the
2570 sender, but the error was temporary, then after the final timeout, only
2571 "retry timeout exceeded" was returned. Now it returns the full error as
2572 well as "retry timeout exceeded".
2574 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2575 do this, and (what is worse) MTAs that accept it.
2577 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2578 will now be deprecated.
2580 PH/56 New os.c-cygwin from the Cygwin maintainer.
2582 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2583 formats (previously only available in exim formats via -bpr, -bpru,
2584 and -bpra. Now also available in native and exiqgrep formats)
2586 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2587 with very large, slow to parse queues
2589 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2591 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2593 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2594 responses to authentication challenges, though it was showing the
2595 challenges; (ii) I've removed the CR characters from the debug output for
2598 PH/58 Allow for the insertion of a newline as well as a space when a string
2599 is turned into more than one encoded-word during RFC 2047 encoding. The
2600 Sieve code now uses this.
2602 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2603 data_4xx, lost_connection, tls_required.
2605 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2606 message was being sent as an SMTP response.
2608 PH/61 Add -l and -k options to exicyclog.
2610 PH/62 When verifying, if an address was redirected to one new address, so that
2611 verification continued, and the new address failed or deferred after
2612 having set something in $address_data, the value of $address_data was not
2613 passed back to the ACL. This was different to the case when no
2614 redirection occurred. The value is now passed back in both cases.
2616 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2617 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2618 use login_cap.h, so on its own it isn't the distinguishing feature. The
2619 new name refers directly to the setclassresources() function.
2621 PH/65 Added configuration files for NetBSD3.
2623 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2625 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2626 is preferred over IPv4.
2628 PH/68 The bounce_return_message and bounce_return_body options were not being
2629 honoured for bounces generated during the reception of non-SMTP messages.
2630 In particular, this applied to messages rejected by the ACL. This bug has
2631 been fixed. However, if bounce_return_message is true and bounce_return_
2632 body is false, the headers that are returned for a non-SMTP message
2633 include only those that have been read before the error was detected.
2634 (In the case of an ACL rejection, they have all been read.)
2636 PH/69 The HTML version of the specification is now built in a directory called
2637 spec_html instead of spec.html, because the latter looks like a path with
2638 a MIME-type, and this confuses some software.
2640 PH/70 Catch two compiler warnings in sieve.c.
2642 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2643 function verify_get_ident() calls ip_connect() to connect a socket, but
2644 if the "connect()" function timed out, ip_connect() used to close the
2645 socket. However, verify_get_ident() also closes the socket later, and in
2646 between Exim writes to the log, which may get opened at this point. When
2647 the socket was closed in ip_connect(), the log could get the same file
2648 descriptor number as the socket. This naturally causes chaos. The fix is
2649 not to close the socket in ip_connect(); the socket should be closed by
2650 the function that creates it. There was only one place in the code where
2651 this was missing, in the iplookup router, which I don't think anybody now
2652 uses, but I've fixed it anyway.
2654 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2655 well as to direct DNS lookups. Otherwise the handling of names in host
2656 lists is inconsistent and therefore confusing.
2662 PH/01 Two changes to the default runtime configuration:
2664 (1) Move the checks for relay_from_hosts and authenticated clients from
2665 after to before the (commented out) DNS black list checks.
2667 (2) Add control=submission to the relay_from_hosts and authenticated
2668 clients checks, on the grounds that messages accepted by these
2669 statements are most likely to be submissions.
2671 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2673 (1) Generate an error if the third argument for the ${prvs expansion is
2676 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2679 (3) Reset the variables that are obtained from the first argument of
2680 ${prvscheck and used in the second argument before leaving the code,
2681 because their memory is reclaimed, so using them afterwards may do
2684 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2685 one (it's much easier than Tom thought :-).
2687 (5) Because of (4), we can now allow for the use of $prvscheck_result
2688 inside the third argument.
2690 PH/03 For some reason, the default setting of PATH when running a command from
2691 a pipe transport was just "/usr/bin". I have changed it to
2694 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2695 anything to be listed in the output from -bV.
2697 PH/05 When a filter generated an autoreply, the entire To: header line was
2698 quoted in the delivery log line, like this:
2700 => >A.N.Other <ano@some.domain> <original@ddress> ...
2702 This has been changed so that it extracts the operative address. There
2703 may be more than one such address. If so, they are comma-separated, like
2706 => >ano@some.domain,ona@other.domain <original@ddress> ...
2708 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2709 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2710 not being looked up in the rDNS to get a host name, Exim was showing the
2711 IP address twice in Received: lines, even though the IP addresses were
2712 identical. For example:
2714 Received: from [1.2.3.4] (helo=[1.2.3.4])
2716 However, if the real host name was known, it was omitting the HELO data
2717 if it matched the actual IP address. This has been tidied up so that it
2718 doesn't show the same IP address twice.
2720 PH/07 When both +timestamp and +memory debugging was on, the value given by
2721 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2722 called by the debug printing, thereby overwriting the timestamp buffer.
2723 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2725 PH/08 When the original message was included in an autoreply transport, it
2726 always said "this is a copy of the message, including all the headers",
2727 even if body_only or headers_only was set. It now gives an appropriate
2730 PH/09 Applied a patch from the Sieve maintainer which:
2732 o fixes some comments
2733 o adds the (disabled) notify extension core
2734 o adds some debug output for the result of if/elsif tests
2735 o points to the current vacation draft in the documentation
2736 and documents the missing references header update
2740 o fixes a bug in processing the envelope test (when testing
2741 multiple envelope elements, the last element determinted the
2744 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2745 Electronic Mail") by including:
2747 Auto-submitted: auto-generated
2749 in the messages that it generates (bounce messages and others, such as
2750 warnings). In the case of bounce messages for non-SMTP mesages, there was
2751 also a typo: it was using "Auto_submitted" (underscore instead of
2752 hyphen). Since every message generated by Exim is necessarily in response
2753 to another message, thes have all been changed to:
2755 Auto-Submitted: auto-replied
2757 in accordance with these statements in the RFC:
2759 The auto-replied keyword:
2761 - SHOULD be used on messages sent in direct response to another
2762 message by an automatic process,
2764 - MUST NOT be used on manually-generated messages,
2766 - MAY be used on Delivery Status Notifications (DSNs) and Message
2767 Disposition Notifications (MDNs),
2769 - MUST NOT be used on messages generated by automatic or periodic
2770 processes, except for messages which are automatic responses to
2773 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2774 to the default Received: header definition.
2776 PH/12 Added log selector acl_warn_skipped (default on).
2778 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2779 variables because (a) they are in the wrong storage pool and (b) even if
2780 they were copied, it wouldn't work properly because of the caching.
2782 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2783 checking when decoding. Apparently there are clients that generate
2784 overlong encoded strings. Why am I not surprised?
2786 PH/15 If the first argument of "${if match_address" was not empty, but did not
2787 contain an "@" character, Exim crashed. Now it writes a panic log message
2788 and treats the condition as false.
2790 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2792 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2793 extension "envelope-auth". The code is finished and in agreement with
2794 other implementations, but there is no documentation so far and in fact,
2795 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2796 not changing the active code.
2798 Print executed "if" and "elsif" statements when debugging is used. This
2799 helps a great deal to understand what a filter does.
2801 Document more things not specified clearly in RFC3028. I had all this
2802 sorted out, when out of a sudden new issues came to my mind. Oops."
2804 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2807 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2808 canonical form (as documented). However, after a host name lookup from
2809 the IP address, check_host() was doing a simple string comparison with
2810 addresses acquired from the DNS when checking that the found name did
2811 have the original IP as one of its addresses. Since any found IPv6
2812 addresses are likely to be in abbreviated form, the comparison could
2813 fail. Luckily, there already exists a function for doing the comparison
2814 by converting both addresses to binary, so now that is used instead of
2815 the text comparison.
2817 PH/20 There was another similar case to PH/19, when a complete host name was
2818 given in a host list; looking up its IP address could give an abbreviated
2819 form, whereas the current host's name might or might not be abbreviated.
2820 The same fix has been applied.
2826 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2827 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2830 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2831 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2833 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2835 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2836 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2837 Also, for "Auto-Submitted", treat anything other than "no" as
2838 non-personal, in accordance with RFC 3834. (Previously it treated
2839 anything starting "auto-" as non-personal.)
2841 TF/01 The control=submission/name=... option had a problem with syntax
2842 errors if the name included a slash character. The /name= option
2843 now slurps the rest of the string, so it can include any characters
2844 but it must come last in the list of options (after /sender_retain
2847 PH/05 Some modifications to the interface to the fake nameserver for the new
2855 TK/01 Added the "success_on_redirect" address verification option. See
2856 NewStuff for rationale and an example.
2858 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2860 PH/02 Patch to exigrep to allow it to work on syslog lines.
2862 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2863 fread() to skip over the body file's header line, because in Cygwin the
2864 header line is locked and is inaccessible.
2866 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2867 co-exist for some time) to make it clear that it is the Exim ID that is
2868 referenced, not the Message-ID: header line.
2870 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2871 string_format() function, because snprintf() does not exist on all
2874 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2875 ${stat: expansion item.
2877 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2878 protocol synchronization error", to keep the pedants happy.
2880 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2881 well as for IRIX systems, when gcc is being used. See the host.c source
2884 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2886 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2889 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2890 transport and to the smtp transport in LMTP mode.
2892 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2894 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2895 run a filter in a subprocess. This could lead to confusion in subsequent
2896 lookups in the parent process. There should also be a search_tidyup() at
2897 the end of the subprocess.
2899 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2900 only if the host matched helo_try_verify_hosts, which caused the
2901 verification to occur when the EHLO/HELO command was issued. The ACL just
2902 tested the remembered result. Now, if a previous verification attempt has
2903 not happened, "verify = helo" does it there and then.
2905 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2907 TK/03 Fix log output including CR from clamd.
2909 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2910 exist provoked a memory error which could cause a segfault.
2912 PH/15 Installed PCRE 6.2
2914 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2916 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2917 of the problem. Specifically, suggested +O2 rather than +O1 for the
2920 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2922 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2923 record, the port was not correctly logged when the outgoing_port log
2924 selector was set (it logged the transort's default port).
2926 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2927 fallback_hosts, and "hosts" in the smtp transport.
2929 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2930 host errors such as "Connection refused".
2932 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2933 authentication with radiusclient 0.4.9:
2935 - Error returned from rc_read_config was caught wrongly
2936 - Username/password not passed on to radius server due to wrong length.
2938 The presumption is that some radiusclient API changes for 4.51/PH/17
2939 were not taken care of correctly. The code is still untested by me (my
2940 Linux distribution still has 0.3.2 of radiusclient), but it was
2941 contributed by a Radius user.
2943 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2944 expanding the "port" option of the smtp transport.
2946 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2947 while reading a MIME header. Thanks to Tom Hughes for a patch.
2949 PH/24 Include config.h inside local_scan.h so that configuration settings are
2952 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2953 that in an ACL for RCPT (for example), you can examine exactly what was
2956 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2957 commands, but it was not correctly comparing the address with the actual
2958 client host address. Thus, it would show the EHLO address in Received:
2959 header lines when this was not necessary.
2961 PH/27 Added the % operator to ${eval:}.
2963 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2964 it should be ignoring failures (because with -C, for example, it has lost
2965 privilege). It wasn't ignoring creation failures other than "already
2968 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2971 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2972 because an input line was too long, either on its own, or by virtue of
2973 too many continuations, the temporary file was not being removed, and the
2974 return code was incorrect.
2976 PH/31 Missing "BOOL" in function definition in filtertest.c.
2978 PH/32 Applied Sieve patches from the maintainer.
2980 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2982 PH/33 Added "verify = not_blind".
2984 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2985 Local/Makefile (with some defaults set). These are used in built scripts
2986 such as exicyclog, but they have never been used in the exim_install
2987 script (though there are many overriding facilities there). I have
2988 arranged that the exim_install script now takes note of these two
2991 PH/35 Installed configuration files for Dragonfly.
2993 PH/36 When a locally submitted message by a trusted user did not contain a
2994 From: header, and the sender address was obtained from -f or from an SMTP
2995 MAIL command, and the trusted user did not use -F to supply a sender
2996 name, $originator_name was incorrectly used when constructing a From:
2997 header. Furthermore, $originator_name was used for submission mode
2998 messages from external hosts without From: headers in a similar way,
2999 which is clearly wrong.
3001 PH/37 Added control=suppress_local_fixups.
3003 PH/38 When log_selector = +received_sender was set, and the addition of the
3004 sender made the log line's construction buffer exactly full, or one byte
3005 less than full, an overflow happened when the terminating "\n" was
3008 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3009 when the result of a list match is failure because a DNS lookup failed.
3011 PH/40 RM_COMMAND is now used in the building process.
3013 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3014 the "build-* directories that it finds.
3016 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3017 address in a domain literal was a prefix of an interface address.
3019 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3020 when verifying a sender address, unless rewrite_headers is false.
3022 PH/44 Wrote a long comment about why errors_to addresses are verified as
3023 recipients, not senders.
3025 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3026 the ratelimit ACL was added.
3028 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3030 PH/46 Added extra information about PostgreSQL errors to the error string.
3032 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3033 avoiding the need to install special zones in a real server. This is
3034 backwards compatible; if it can't find the fake resolver, it drops back.
3035 Thus, both old and new test suites can be run.
3037 TF/02 Added util/ratelimit.pl
3039 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3040 clock is set back in time.
3042 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3043 Candler <B.Candler@pobox.com>.
3045 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3046 OK for addresses that are the result of redirections.
3048 PH/48 A number of further additions for the benefit of the new test suite,
3049 including a fake gethostbyname() that interfaces to the fake DNS resolver
3052 TF/06 The fix for widen_domains has also been applied to qualify_single and
3053 search_parents which are the other dnslookup options that can cause
3056 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3059 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3061 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3062 service" if it encountered an MX record with an empty target hostname.
3063 The message is now "an MX or SRV record indicated no SMTP service".
3065 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3066 if the DNS of the sending site is misconfigured. This is quite a
3067 common situation. This change restores the behaviour of treating a
3068 helo verification defer as a failure.
3070 PH/51 If self=fail was set on a router, the bounce message did not include the
3071 actual error message.
3077 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3079 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3080 command itself ended in error, the underlying message about the transport
3081 filter timeout was being overwritten with the pipe command error. Now the
3082 underlying error message should be appended to the second error message.
3084 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3086 PH/02 Reduce the amount of output that "make" produces by default. Full output
3087 can still be requested.
3089 PH/03 The warning log line about a condition test deferring for a "warn" verb
3090 was being output only once per connection, rather than after each
3091 occurrence (because it was using the same function as for successful
3092 "warn" verbs). This seems wrong, so I have changed it.
3094 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3095 it should not have, which might have caused a crash in the right
3096 circumstances, but probably never did.
3098 PH/04 Installed a modified version of Tony Finch's patch to make submission
3099 mode fix the return path as well as the Sender: header line, and to
3100 add a /name= option so that you can make the user's friendly name appear
3103 TF/03 Added the control = fakedefer ACL modifier.
3105 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3106 Mark Lowes for thorough testing.
3108 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3110 TK/03 Merged latest SRS patch from Miles Wilton.
3112 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3113 with the definition in sysexits.h (which is #included earlier).
3114 Fortunately, Exim does not actually use EX_OK. The code used to try to
3115 preserve the sysexits.h value, by assumimg that macro definitions were
3116 scanned for macro replacements. I have been disabused of this notion,
3117 so now the code just undefines EX_OK before #including unistd.h.
3119 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3120 in the smtp transport. When a block could not be written in a single
3121 write() function, the timeout was being re-applied to each part-write.
3122 This seems wrong - if the receiver was accepting one byte at a time it
3123 would take for ever. The timeout is now adjusted when this happens. It
3124 doesn't have to be particularly precise.
3126 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3127 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3129 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3130 without a domain if the check to <postmaster@domain> fails.
3132 SC/01 Eximstats: added -xls and the ability to specify output files
3133 (patch written by Frank Heydlauf).
3135 SC/02 Eximstats: use FileHandles for outputing results.
3137 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3139 SC/04 Eximstats: fixed display of large numbers with -nvr option
3141 SC/05 Eximstats: fixed merging of reports with empty tables.
3143 SC/06 Eximstats: added the -include_original_destination flag
3145 SC/07 Eximstats: removed tabs and trailing whitespace.
3147 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3149 TK/06 MBOX spool code: Add real "From " MBOX separator line
3150 so the .eml file is really in mbox format (even though
3151 most programs do not really care). Patch from Alex Miller.
3153 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3154 The latter is generated from $received_to and is only set if the
3155 message has one envelope recipient. SA can use these headers,
3156 obviously out-of-the-box. Patch from Alex Miller.
3158 PH/08 The ${def test on a variable was returning false if the variable's
3159 value was "0", contrary to what the specification has always said!
3160 The result should be true unless the variable is empty.
3162 PH/09 The syntax error of a character other than { following "${if
3163 def:variable_name" (after optional whitespace) was not being diagnosed.
3164 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3165 accidental colon was present, for example, could give incorrect results.
3167 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3168 result is used (not including appendfile, where other changes are about
3171 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3172 This involved changing a lot of size variables from int to off_t. It
3173 should work with maildirs and everything.
3175 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3176 spamd dying while we are connected to it.
3178 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3181 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3182 function for BDB 4.3.
3184 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3186 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3187 the sources for PCRE have been re-arranged and more files are now
3190 PH/15 The code I had for printing potentially long long variables in PH/11
3191 above was not the best (it lost precision). The length of off_t variables
3192 is now inspected at build time, and an appropriate printing format (%ld
3193 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3194 to be "long long int" or "long int". This is needed for the internal
3195 formatting function string_vformat().
3197 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3198 the configuration file to be ":syslog", then the script "guesses" where
3199 the logs files are, rather than using the compiled in default. In our
3200 case the guess is not the same as the compiled default, so the script
3201 suddenly stopped working when I started to use syslog. The patch checks
3202 to see if log_file_path is "". If so, it attempts to read it from exim
3203 with no configuration file to get the compiled in version, before it
3204 falls back to the previous guessing code."
3206 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3207 implementing BATV in an Exim configuration. See NewStuff for the gory
3210 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3211 Makefile that are specific to HP-UX.
3213 PH/18 If the "use_postmaster" option was set for a recipient callout together
3214 with the "random" option, the postmaster address was used as the MAIL
3215 FROM address for the random test, but not for the subsequent recipient
3216 test. It is now used for both.
3218 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3219 patch removes a few documentation additions to RFC 3028, because the
3220 latest draft now contains them. It adds the new en;ascii-case comparator
3221 and a new error check for 8bit text in MIME parts. Comparator and
3222 require names are now matched exactly. I enabled the subaddress
3223 extension, but it is not well tested yet (read: it works for me)."
3225 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3226 rework some of the code of TK/09 above to avoid the hardwired use of
3227 "%lld" and "long long". Replaced the call to snprintf() with a call to
3230 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3231 records point to non-existent hosts", "retry timeout exceeded", and
3232 "retry time not reached for any host after a long failure period".
3234 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3235 experimental DomainKeys support:
3237 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3238 (2) On an error such as an illegally used "control", the wrong name for
3239 the control was given.
3241 These problems did NOT occur unless DomainKeys support was compiled.
3243 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3245 PH/24 Added ${if match_ip condition.
3247 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3248 databases so that it will be absolutely obvious if a crash occurs in the
3249 DB library. This is a regular occurrence (often caused by mis-matched
3252 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3253 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3254 picked up on a user's system that detects such things. There doesn't seem
3255 to be a gcc warning option for this - only an attribute that has to be
3256 put on the function's prototype. It seems that in Fedora Core 4 they have
3257 set this on a number of new functions. No doubt there will be more in due
3260 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3261 specify a transport. However, if an address that was verified by such a
3262 router was the subject of a callout, Exim crashed because it tried to
3263 read the rcpt_include_affixes from the non-existent transport. Now it
3264 just assumes that the setting of that option is false. This bug was
3265 introduced by 4.51/PH/31.
3267 PH/28 Changed -d+all to exclude +memory, because that information is very
3268 rarely of interest, but it makes the output a lot bigger. People tend to
3269 do -d+all out of habit.
3271 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3272 code in os-type was giving problems when libc.so lives in lib64, like on
3275 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3276 aren't the modern standard, and it seems that some systems' include files
3277 don't always have them. Exim was already checking for some of the newer
3278 ones like T_AAAA, and defining it itself. I've added checks for all the
3279 record types that Exim uses.
3281 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3282 not automatically generating a new one, as it is supposed to. This
3283 prevented TLS from working. If the file did exist, but contained invalid
3284 data, a new version was generated, as expected. It was only the case of a
3285 non-existent file that was broken.
3287 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3288 with a change in libdomainkeys > 0.64.
3290 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3291 from DNS. If the selector record carries the flag, it now has
3292 precedence over the domain-wide flag.
3294 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3296 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3297 the use of a port name that isn't defined in /etc/services) occurred, the
3298 message was deferred as in a normal delivery, and thus remained on the
3299 spool, instead of being failed because of the mua_wrapper setting. This
3300 is now fixed, and I tidied up some of the mua_wrapper messages at the
3303 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3304 the messages in a hash of arrays rather than using individual hashes.
3305 This is a bit cleaner and results in dramatic memory savings, albeit
3306 at a slight CPU cost.
3308 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3309 as requested by Marc Sherman.
3311 SC/10 Eximstats: added histograms for user specified patterns as requested
3314 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3316 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3317 fopen() in the content-scanning modules that did not already have it.
3323 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3324 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3326 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3328 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3330 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3331 to test Sieve filters that use "vacation".
3333 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3334 that changes the way the GnuTLS parameters are stored in the cache file.
3335 The new format can be generated externally. For backward compatibility,
3336 if the data in the cache doesn't make sense, Exim assumes it has read an
3337 old-format file, and it generates new data and writes a new file. This
3338 means that you can't go back to an older release without removing the
3341 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3342 work if there are any delivery delays because "one_time" forces the
3343 parent to be marked "delivered", so its unseen clone is never tried
3344 again. For this reason, Exim now forbids the simultaneous setting of
3347 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3348 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3349 entry if you want to know the details. The fix, however, neglected to
3350 consider the case when local delivery batching is involved. The test for
3351 "previously delivered" was not happening when checking to see if an
3352 address could be batched with a previous (undelivered) one; under
3353 certain circumstances this could lead to multiple deliveries to the same
3356 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3357 in its include files, and this causes problems building Exim.
3359 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3360 header_syntax) but Exim was just ignoring anything given after a slash.
3361 In particular, this caused confusion with an attempt to use "verify =
3362 reverse_host_lookup/defer_ok". An error is now given when options are
3363 supplied for verify items that do not have them. (Maybe reverse_host_
3364 lookup should have a defer_ok option, but that's a different point.)
3366 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3367 defined by RFC 821) to 2048, because there were problems with some AUTH
3368 commands, and RFC 1869 says the size should be increased for extended
3369 SMTP commands that take arguments.
3371 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3374 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3375 "unknown" error; now it says that the functionality isn't in the binary.
3377 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3378 an address' error message when a string expansion fails (syntax or
3379 whatever). Otherwise the password may appear in the log. Following change
3380 PH/42 below, there is no longer a chance of it appearing in a bounce
3383 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3385 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3386 of its addresses was ever tried. (Bugzilla bug #2.)
3388 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3389 the result incorrectly in the debug output. (It correctly added a newline
3390 to what was transported.)
3392 TF/01 Added $received_time.
3394 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3395 commented out examples of how to interface to a virus scanner and to
3396 SpamAssassin. Also added commented examples of av_scanner and
3397 spamd_address settings.
3399 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3400 and controls are allowed in which ACLs. There were a couple of minor
3401 errors. Some of the entries in the conditions table (which is a table of
3402 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3403 negation of where the condition IS allowed.
3405 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3407 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3408 header file does not have a version number, so I've had to invent a new
3409 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3410 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3411 radiusclient), but it was contributed by a Radius user.
3413 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3414 files or directories, for OpenSSL.
3416 PH/19 When an Exim process that is running as root has to create an Exim log
3417 file, it does so in a subprocess that runs as exim:exim so as to get the
3418 ownership right at creation (otherwise, other Exim processes might see
3419 the file with the wrong ownership). There was no test for failure of this
3420 fork() call, which would lead to the process getting stuck as it waited
3421 for a non-existent subprocess. Forks do occasionally fail when resources
3422 run out. I reviewed all the other calls to fork(); they all seem to check
3425 PH/20 When checking for unexpected SMTP input at connect time (before writing
3426 the banner), Exim was not dealing correctly with a non-positive return
3427 from the read() function. If the client had disconnected by this time,
3428 the result was a log entry for a synchronization error with an empty
3429 string after "input=" when read() returned zero. If read() returned -1
3430 (an event I could not check), uninitialized data bytes were printed.
3431 There were reports of junk text (parts of files, etc) appearing after
3434 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3436 PH/22 Added support for macro redefinition, and (re)definition in between
3437 driver and ACL definitions.
3439 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3440 forgetting to use the resulting value; it was using the unexpanded value.
3442 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3443 hadn't been configured. The fix is from Juergen Kreileder, who
3444 understands it better than I do:
3446 "Here's what I see happening with three configured cyrus_sasl
3447 authenticators configured (plain, login, cram-md5):
3449 On startup auth_cyrus_sasl_init() gets called for each of these.
3450 This means three calls to sasl_listmech() without a specified mech_list.
3451 => SASL tests which mechs of all available mechs actually work
3452 => three warnings about OTP not working
3453 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3455 With the patch, sasl_listmech() also gets called three times. But now
3456 SASL's mech_list option is set to the server_mech specified in the the
3457 authenticator. Or in other words, the answer from sasl_listmech()
3458 gets limited to just the mech you're testing for (which is different
3460 => the return list contains just 'plain' or 'login', 'cram-md5' or
3461 nothing depending on the value of ob->server_mech.
3463 I've just tested the patch: Authentication still works fine,
3464 unavailable mechs specified in the exim configuration are still
3465 caught, and the auth.log warnings about OTP are gone."
3467 PH/25 When debugging is enabled, the contents of the command line are added
3468 to the debugging output, even when log_selector=+arguments is not
3471 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3472 answer is "GNU", and only if the return is "GNU/something" is the answer
3475 PH/27 $acl_verify_message is now set immediately after the failure of a
3476 verification in an ACL, and so is available in subsequent modifiers. In
3477 particular, the message can be preserved by coding like this:
3479 warn !verify = sender
3480 set acl_m0 = $acl_verify_message
3482 Previously, $acl_verify_message was set only while expanding "message"
3483 and "log_message" when a very denied access.
3485 PH/28 Modified OS/os.c-Linux with
3487 -#ifndef OS_LOAD_AVERAGE
3488 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3490 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3491 nomenclature these days.)
3493 PH/29 Installed patch from the Sieve maintainer that adds the options
3494 sieve_useraddress and sieve_subaddress to the redirect router.
3496 PH/30 In these circumstances:
3497 . Two addresses routed to the same list of hosts;
3498 . First host does not offer TLS;
3499 . First host accepts first address;
3500 . First host gives temporary error to second address;
3501 . Second host offers TLS and a TLS session is established;
3502 . Second host accepts second address.
3503 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3504 and peerdn, if requested) that were in fact used only for the second
3507 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3508 attention to any local part prefix or suffix that was matched by the
3509 router that accepted the address. It now behaves in the same way as it
3510 does for delivery: the affixes are removed from the local part unless
3511 rcpt_include_affixes is set on the transport.
3513 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3514 timeout during the DATA phase of an incoming message.
3516 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3517 applied a patch sanctioned by the Sieve maintainer.
3519 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3520 the uid or gid is negative. A case of a negative gid caused this to be
3521 noticed. The fix allows for either to be negative.
3523 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3524 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3527 PH/36 The change PH/12 above was broken. Fixed it.
3529 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3530 the grounds that routing the same address twice would always produce the
3531 same answer. This might have been true once, but it is certainly no
3532 longer true now. Routing a child address may depend on the previous
3533 routing that produced that child. Some complicated redirection strategies
3534 went wrong when messages had multiple recipients, and made Exim's
3535 behaviour dependent on the order in which the addresses were given.
3537 I have moved the duplicate checking until after the routing is complete.
3538 Exim scans the addresses that are assigned to local and remote
3539 transports, and removes any duplicates. This means that more work will be
3540 done, as duplicates will always all be routed, but duplicates are
3541 presumably rare, so I don't expect this is of any significance.
3543 For deliveries to pipes, files, and autoreplies, the duplicate checking
3544 still happens during the routing process, since they are not going to be
3547 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3548 It corrects a timeout issue with spamd. This is Ian's comment: "The
3549 background is that sometimes spamd either never reads data from a
3550 connection it has accepted, or it never writes response data. The exiscan
3551 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3552 blindly assumes that writes won't block so it may never time out."
3554 PH/39 Allow G after quota size as well as K and M.
3556 PH/40 The value set for $authenticated_id in an authenticator may not contain
3557 binary zeroes or newlines because the value is written to log lines and
3558 to spool files. There was no check on this. Now the value is run through
3559 the string_printing() function so that such characters are converted to
3560 printable escape sequences.
3562 PH/41 $message_linecount is a new variable that contains the total number of
3563 lines in the message. Compare $body_linecount, which is the count for the
3566 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3567 bounce and delay warning messages, except in certain special cases, which
3570 (a) An SMTP error message from a remote host;
3571 (b) A message specified in a :fail: redirection;
3572 (c) A message specified in a "fail" command in a system filter;
3573 (d) A message specified in a FAIL return from the queryprogram router;
3574 (e) A message specified by the cannot_route_message router option.
3576 In these cases only, Exim does include the error details in bounce and
3577 warning messages. There are also a few cases where bland messages such
3578 as "unrouteable address" or "local delivery error" are given.
3580 PH/43 $value is now also set for the "else" part of a ${run expansion.
3582 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3583 being worked on, but at least Exim now implements the latest version to
3586 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3587 process to complete was treated as a delivery failure, a timeout while
3588 writing the message to the pipe was logged, but erroneously treated as a
3589 successful delivery. Such timeouts include transport filter timeouts. For
3590 consistency with the overall process timeout, these timeouts are now
3591 treated as errors, giving rise to delivery failures by default. However,
3592 there is now a new Boolean option for the pipe transport called
3593 timeout_defer, which, if set TRUE, converts the failures into defers for
3594 both kinds of timeout. A transport filter timeout is now identified in
3597 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3598 systems where "make" and "gmake" are different, calling "gmake" at top
3599 level broke things. I've arranged for the value of $(MAKE) to be passed
3600 from the Makefile to this script so that it can call the same version of
3604 A note about Exim versions 4.44 and 4.50
3605 ----------------------------------------
3607 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3608 changes of various kinds. As a consequence, a big documentation update was
3609 needed. This delayed the release for rather longer than seemed good, especially
3610 in the light of a couple of (minor) security issues. Therefore, the changes
3611 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3612 release. So 4.44 and 4.50 are in effect two different branches that both start
3615 I have left the 4.50 change log unchanged; it contains all the changes since
3616 4.43. The change log for 4.44 is below; many of its items are identical to
3617 those for 4.50. This seems to be the most sensible way to preserve the
3618 historical information.
3624 1. Minor wording change to the doc/README.SIEVE file.
3626 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3627 computation of the current number of files was incorrect.
3629 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3630 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3633 4. Give more explanation in the error message when the command for a transport
3634 filter fails to execute.
3636 5. There are several places where Exim runs a non-Exim command in a
3637 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3638 was being done only for the command run by the queryprogram router. It is
3639 now done for all such subprocesses. The other cases are: ${run, transport
3640 filters, and the commands run by the lmtp and pipe transports.
3642 6. Added CONFIGURE_GROUP build-time option.
3644 7. Some older OS have a limit of 256 on the maximum number of file
3645 descriptors. Exim was using setrlimit() to set 1000 as a large value
3646 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3647 systems. I've change it so that if it can't get 1000, it tries for 256.
3649 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3650 was an oversight, and furthermore, ever since the addition of extra
3651 controls (e.g. 4.43/32), the checks on when to allow different forms of
3652 "control" were broken. There should now be diagnostics for all cases when a
3653 control that does not make sense is encountered.
3655 9. Added the /retain_sender option to "control=submission".
3657 10. $recipients is now available in the predata ACL (oversight).
3659 11. Tidy the search cache before the fork to do a delivery from a message
3660 received from the command line. Otherwise the child will trigger a lookup
3661 failure and thereby defer the delivery if it tries to use (for example) a
3662 cached ldap connection that the parent has called unbind on.
3664 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3665 of $address_data from the recipient verification was clobbered by the
3666 sender verification.
3668 13. The value of address_data from a sender verification is now available in
3669 $sender_address_data in subsequent conditions in the ACL statement.
3671 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3673 15. Added a new option "connect=<time>" to callout options, to set a different
3676 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3677 was its contents. (It was OK if the option was not defined at all.)
3679 17. A "Completed" log line is now written for messages that are removed from
3680 the spool by the -Mrm option.
3682 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3683 information about exactly what failed.
3685 19. Added -dd to debug only the daemon process.
3687 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3688 handles timeouts, both on the server side and network timeouts. Renamed the
3689 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3691 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3692 It is now set to "smtps".
3694 22. $host_address is now set to the target address during the checking of
3695 ignore_target_hosts.
3697 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3698 being passed; this would have caused $sender_host_name to have been used if
3699 matching the list had actually called for a host name (not very likely,
3700 since this list is usually IP addresses). A host name is now passed as
3703 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3704 code that specifies a non-restarting handler (typically sigaction() in
3705 modern systems) in an attempt to fix a rare and obscure crash bug.
3707 25. Narrowed the window for a race in the daemon that could cause it to ignore
3708 SIGCHLD signals. This is not a major problem, because they are used only to
3709 wake it up if nothing else does.
3711 26. A malformed maildirsize file could cause Exim to calculate negative values
3712 for the mailbox size or file count. Odd effects could occur as a result.
3713 The maildirsize information is now recalculated if the size or filecount
3716 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3717 support for a long time. Removed HAVE_SYS_VFS_H.
3719 28. Installed the latest version of exipick from John Jetmore.
3721 29. In an address list, if the pattern was not a regular expression, an empty
3722 subject address (from a bounce message) matched only if the pattern was an
3723 empty string. Non-empty patterns were not even tested. This was the wrong
3724 because it is perfectly reasonable to use an empty address as part of a
3725 database query. An empty address is now tested by patterns that are
3726 lookups. However, all the other forms of pattern expect the subject to
3727 contain a local part and a domain, and therefore, for them, an empty
3728 address still always fails if the pattern is not itself empty.
3730 30. Exim went into a mad DNS loop when attempting to do a callout where the
3731 host was specified on an smtp transport, and looking it up yielded more
3732 than one IP address.
3734 31. Re-factored the code for checking spool and log partition space into a
3735 function that finds that data and another that does the check. The former
3736 is then used to implement four new variables: $spool_space, $log_space,
3737 $spool_inodes, and $log_inodes.
3739 32. The RFC2047 encoding function was originally intended for short strings
3740 such as real names; it was not keeping to the 75-character limit for
3741 encoded words that the RFC imposes. It now respects the limit, and
3742 generates multiple encoded words if necessary. To be on the safe side, I
3743 have increased the buffer size for the ${rfc2047: expansion operator from
3746 33. It is now permitted to omit both strings after an "if" condition; if the
3747 condition is true, the result is "true". As before, when the second string
3748 is omitted, a false condition yields an empty string. This makes it less
3749 cumbersome to write custom ACL and router conditions.
3751 34. Failure to deliver a bounce message always caused it to be frozen, even if
3752 there was an errors_to setting on the router. The errors_to setting is now
3755 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3756 canonical form (fully expanded) before being placed in
3757 $sender_host_address.
3759 36. The table in the code that translates DNS record types into text (T_A to
3760 "A" for instance) was missing entries for NS and CNAME. It is just possible
3761 that this could have caused confusion if both these types were looked up
3762 for the same domain, because the text type is used as part of Exim's
3763 per-process caching. But the chance of anyone hitting this buglet seems
3766 37. The dnsdb lookup has been extended in a number of ways.
3768 (1) There is a new type, "zns", which walks up the domain tree until it
3769 finds some nameserver records. It should be used with care.
3771 (2) There is a new type, "mxh", which is like "mx" except that it returns
3772 just the host names, not the priorities.
3774 (3) It is now possible to give a list of domains (or IP addresses) to be
3775 looked up. The behaviour when one of the lookups defers can be
3776 controlled by a keyword.
3778 (4) It is now possible to specify the separator character for use when
3779 multiple records are returned.
3781 38. The dnslists ACL condition has been extended: it is now possible to supply
3782 a list of IP addresses and/or domains to be looked up in a particular DNS
3785 39. Added log_selector=+queue_time_overall.
3787 40. When running the queue in the test harness, wait just a tad after forking a
3788 delivery process, to get repeatability of debugging output.
3790 41. Include certificate and key file names in error message when GnuTLS fails
3791 to set them up, because the GnuTLS error message doesn't include the name
3792 of the failing file when there is a problem reading it.
3794 42. Allow both -bf and -bF in the same test run.
3796 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3798 44. The "Exiscan patch" is now merged into the mainline Exim source.
3800 45. Sometimes the final signoff response after QUIT could fail to get
3801 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3802 < 0 before doing a fflush(). This bug looks as though it goes back to the
3803 introduction of TLS in release 3.20, but "sometimes" must have been rare
3804 because the tests only now provoked it.
3806 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3807 (this can affect the format of dates).
3809 47. exim_tidydb, when checking for the continued existence of a message for
3810 which it has found a message-specific retry record, was not finding
3811 messages that were in split spool directories. Consequently, it was
3812 deleting retry records that should have stayed in existence.
3814 48. Steve fixed some bugs in eximstats.
3816 49. The SPA authentication driver was not abandoning authentication and moving
3817 on to the next authenticator when an expansion was forced to fail,
3818 contradicting the general specification for all authenticators. Instead it
3819 was generating a temporary error. It now behaves as specified.
3821 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3822 (the order specifies the preference for clients). The order is now AES256,
3823 AES128, 3DES, ARCFOUR128.
3825 51. Small patch to Sieve code - explicitly set From: when generating an
3828 52. Exim crashed if a remote delivery caused a very long error message to be
3829 recorded - for instance if somebody sent an entire SpamAssassin report back
3830 as a large number of 550 error lines. This bug was coincidentally fixed by
3831 increasing the size of one of Exim's internal buffers (big_buffer) that
3832 happened as part of the Exiscan merge. However, to be on the safe side, I
3833 have made the code more robust (and fixed the comments that describe what
3836 53. Now that there can be additional text after "Completed" in log lines (if
3837 the queue_time_overall log selector is set), a one-byte patch to exigrep
3838 was needed to allow it to recognize "Completed" as not the last thing in
3841 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3842 patch that reportedly fixes this has been added. I am not expert enough to
3843 create a test for it. This is what the patch creator wrote:
3845 "I found a little strange behaviour of ldap code when working with
3846 Windows 2003 AD Domain, where users was placed in more than one
3847 Organization Units. When I tried to give exim partial DN, the exit code
3848 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3849 But simultaneously result of request was absolutely normal ldap result,
3850 so I produce this patch..."
3852 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3853 REFERENCE, so I have modified the code to exclude the patch when that macro
3856 55. Some experimental protocols are using DNS PTR records for new purposes. The
3857 keys for these records are domain names, not reversed IP addresses. The
3858 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3859 leaves it alone. Component reversal etc. now happens only for IP addresses.
3862 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3864 57. Double the size of the debug message buffer (to 2048) so that more of very
3865 long debug lines gets shown.
3867 58. The exicyclog utility now does better if the number of log files to keep
3868 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3870 59. Two changes related to the smtp_active_hostname option:
3872 (1) $smtp_active_hostname is now available as a variable.
3873 (2) The default for smtp_banner uses $smtp_active_hostname instead
3874 of $primary_hostname.
3876 60. The host_aton() function is supposed to be passed a string that is known
3877 to be a valid IP address. However, in the case of IPv6 addresses, it was
3878 not checking this. This is a hostage to fortune. Exim now panics and dies
3879 if the condition is not met. A case was found where this could be provoked
3880 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3881 components; fortuitously, this particular loophole had already been fixed
3882 by change 4.50/55 above.
3884 If there are any other similar loopholes, the new check in host_aton()
3885 itself should stop them being exploited. The report I received stated that
3886 data on the command line could provoke the exploit when Exim was running as
3887 exim, but did not say which command line option was involved. All I could
3888 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3889 running as the user.
3892 61. There was a buffer overflow vulnerability in the SPA authentication code
3893 (which came originally from the Samba project). I have added a test to the
3894 spa_base64_to_bits() function which I hope fixes it.
3897 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3898 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3900 63. The daemon start-up calls getloadavg() while still root for those OS that
3901 need the first call to be done as root, but it missed one case: when
3902 deliver_queue_load_max is set with deliver_drop_privilege. This is
3903 necessary for the benefit of the queue runner, because there is no re-exec
3904 when deliver_drop_privilege is set.
3906 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3907 This has been fixed.
3909 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3910 was in use, was not putting the data itself into the right store pool;
3911 consequently, it could be overwritten for a subsequent message in the same
3912 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3915 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3917 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3918 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3919 the function were treating the return as a boolean value, which happened to
3920 work because 0=false and not-0=true, but is not correct code.
3922 68. The host_aton() function was not handling scoped IPv6 addresses (those
3923 with, for example, "%eth0" on the end) correctly.
3925 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3926 negated items (that is, ~something) in unsigned ints. Some compilers
3927 apparently mutter when there is no cast.
3929 70. If an address verification called from an ACL failed, and did not produce a
3930 user-specific message (i.e. there was only a "system" message), nothing was
3931 put in $acl_verify_message. In this situation, it now puts the system
3934 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3935 session; change 4.31/43 added the unwanted input to the log line - except
3936 that it did not do this in the start of session case. It now does.
3938 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3939 This is wrong and can cause the other end to generate a synchronization
3940 error if it is another Exim or anything else that does the synchronization
3941 check. A QUIT command is no longer sent after a timeout.
3943 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3944 during host lookups.
3946 74. The defer_ok option of callout verification was not working if it was used
3947 when verifying addresses in header lines, that is, for this case:
3949 verify = header_sender/callout=defer_ok
3951 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3952 those file descriptors could be used for SMTP connections. If anything
3953 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3954 could be sent to the SMTP client, causing chaos. The daemon now opens
3955 stdin, stdout, and stderr to /dev/null when it puts itself into the
3958 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3959 log by default. The user can override this with suitable Perl magic.
3961 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3962 the log message when discard triggers, was not working for the DATA ACL or
3963 for the non-SMTP ACL.
3965 78. Error message wording change in sieve.c.
3967 79. If smtp_accept_max_per_host was set, the number of connections could be
3968 restricted to fewer than expected, because the daemon was trying to set up
3969 a new connection before checking whether the processes handling previous
3970 connections had finished. The check for completed processes is now done
3971 earlier. On busy systems, this bug wouldn't be noticed because something
3972 else would have woken the daemon, and it would have reaped the completed
3975 80. If a message was submitted locally by a user whose login name contained one
3976 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3977 It caused a spool format error. I have fixed the spool reading code. A
3978 related problem was that the "from" clause in the Received: line became
3979 illegal because of the space(s). It is now covered by ${quote_local_part.
3981 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3984 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3986 83. Patch from Sieve maintainer to fix unterminated string problem in
3987 "vacation" handling.
3989 84. Some minor changes to the Linux configuration files to help with other
3990 OS variants using glibc.
3992 85. One more patch for Sieve to update vacation handling to latest spec.
3995 ----------------------------------------------------
3996 See the note above about the 4.44 and 4.50 releases.
3997 ----------------------------------------------------
4003 1. Change 4.43/35 introduced a bug that caused file counts to be
4004 incorrectly computed when quota_filecount was set in an appendfile
4007 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4008 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4011 3. Give more explanation in the error message when the command for a transport
4012 filter fails to execute.
4014 4. There are several places where Exim runs a non-Exim command in a
4015 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4016 was being done only for the command run by the queryprogram router. It is
4017 now done for all such subprocesses. The other cases are: ${run, transport
4018 filters, and the commands run by the lmtp and pipe transports.
4020 5. Some older OS have a limit of 256 on the maximum number of file
4021 descriptors. Exim was using setrlimit() to set 1000 as a large value
4022 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4023 systems. I've change it so that if it can't get 1000, it tries for 256.
4025 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4026 was an oversight, and furthermore, ever since the addition of extra
4027 controls (e.g. 4.43/32), the checks on when to allow different forms of
4028 "control" were broken. There should now be diagnostics for all cases when a
4029 control that does not make sense is encountered.
4031 7. $recipients is now available in the predata ACL (oversight).
4033 8. Tidy the search cache before the fork to do a delivery from a message
4034 received from the command line. Otherwise the child will trigger a lookup
4035 failure and thereby defer the delivery if it tries to use (for example) a
4036 cached ldap connection that the parent has called unbind on.
4038 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4039 of $address_data from the recipient verification was clobbered by the
4040 sender verification.
4042 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4043 was its contents. (It was OK if the option was not defined at all.)
4045 11. A "Completed" log line is now written for messages that are removed from
4046 the spool by the -Mrm option.
4048 12. $host_address is now set to the target address during the checking of
4049 ignore_target_hosts.
4051 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4052 being passed; this would have caused $sender_host_name to have been used if
4053 matching the list had actually called for a host name (not very likely,
4054 since this list is usually IP addresses). A host name is now passed as
4057 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4058 code that specifies a non-restarting handler (typically sigaction() in
4059 modern systems) in an attempt to fix a rare and obscure crash bug.
4061 15. Narrowed the window for a race in the daemon that could cause it to ignore
4062 SIGCHLD signals. This is not a major problem, because they are used only to
4063 wake it up if nothing else does.
4065 16. A malformed maildirsize file could cause Exim to calculate negative values
4066 for the mailbox size or file count. Odd effects could occur as a result.
4067 The maildirsize information is now recalculated if the size or filecount
4070 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4071 support for a long time. Removed HAVE_SYS_VFS_H.
4073 18. Updated exipick to current release from John Jetmore.
4075 19. Allow an empty sender to be matched against a lookup in an address list.
4076 Previously the only cases considered were a regular expression, or an
4079 20. Exim went into a mad DNS lookup loop when doing a callout where the
4080 host was specified on the transport, if the DNS lookup yielded more than
4083 21. The RFC2047 encoding function was originally intended for short strings
4084 such as real names; it was not keeping to the 75-character limit for
4085 encoded words that the RFC imposes. It now respects the limit, and
4086 generates multiple encoded words if necessary. To be on the safe side, I
4087 have increased the buffer size for the ${rfc2047: expansion operator from
4090 22. Failure to deliver a bounce message always caused it to be frozen, even if
4091 there was an errors_to setting on the router. The errors_to setting is now
4094 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4095 canonical form (fully expanded) before being placed in
4096 $sender_host_address.
4098 24. Updated eximstats to version 1.33
4100 25. Include certificate and key file names in error message when GnuTLS fails
4101 to set them up, because the GnuTLS error message doesn't include the name
4102 of the failing file when there is a problem reading it.
4104 26. Expand error message when OpenSSL has problems setting up cert/key files.
4107 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4108 (this can affect the format of dates).
4110 28. exim_tidydb, when checking for the continued existence of a message for
4111 which it has found a message-specific retry record, was not finding
4112 messages that were in split spool directories. Consequently, it was
4113 deleting retry records that should have stayed in existence.
4115 29. eximstats updated to version 1.35
4116 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4117 1.35 - bugfix such that pie charts by volume are generated correctly
4119 30. The SPA authentication driver was not abandoning authentication and moving
4120 on to the next authenticator when an expansion was forced to fail,
4121 contradicting the general specification for all authenticators. Instead it
4122 was generating a temporary error. It now behaves as specified.
4124 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4125 (the order specifies the preference for clients). The order is now AES256,
4126 AES128, 3DES, ARCFOUR128.
4128 31. Small patch to Sieve code - explicitly set From: when generating an
4131 32. Exim crashed if a remote delivery caused a very long error message to be
4132 recorded - for instance if somebody sent an entire SpamAssassin report back
4133 as a large number of 550 error lines. This bug was coincidentally fixed by
4134 increasing the size of one of Exim's internal buffers (big_buffer) that
4135 happened as part of the Exiscan merge. However, to be on the safe side, I
4136 have made the code more robust (and fixed the comments that describe what
4139 33. Some experimental protocols are using DNS PTR records for new purposes. The
4140 keys for these records are domain names, not reversed IP addresses. The
4141 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4142 leaves it alone. Component reversal etc. now happens only for IP addresses.
4145 34. The host_aton() function is supposed to be passed a string that is known
4146 to be a valid IP address. However, in the case of IPv6 addresses, it was
4147 not checking this. This is a hostage to fortune. Exim now panics and dies
4148 if the condition is not met. A case was found where this could be provoked
4149 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4150 components; fortuitously, this particular loophole had already been fixed
4151 by change 4.50/55 or 4.44/33 above.
4153 If there are any other similar loopholes, the new check in host_aton()
4154 itself should stop them being exploited. The report I received stated that
4155 data on the command line could provoke the exploit when Exim was running as
4156 exim, but did not say which command line option was involved. All I could
4157 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4158 running as the user.
4161 35. There was a buffer overflow vulnerability in the SPA authentication code
4162 (which came originally from the Samba project). I have added a test to the
4163 spa_base64_to_bits() function which I hope fixes it.
4166 36. The daemon start-up calls getloadavg() while still root for those OS that
4167 need the first call to be done as root, but it missed one case: when
4168 deliver_queue_load_max is set with deliver_drop_privilege. This is
4169 necessary for the benefit of the queue runner, because there is no re-exec
4170 when deliver_drop_privilege is set.
4172 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4173 was in use, was not putting the data itself into the right store pool;
4174 consequently, it could be overwritten for a subsequent message in the same
4175 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4178 38. Sometimes the final signoff response after QUIT could fail to get
4179 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4180 < 0 before doing a fflush(). This bug looks as though it goes back to the
4181 introduction of TLS in release 3.20, but "sometimes" must have been rare
4182 because the tests only now provoked it.
4188 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4189 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4190 Now it returns an integer. A number of calls were still expecting a T/F
4191 return. Fortuitously, in all cases, the tests worked in OK situations,
4192 which is the norm. However, things would have gone wrong on any write
4193 failures on the smtp file descriptor. This function is used when sending
4194 messages over SMTP and also when doing verify callouts.
4196 2. When Exim is called to do synchronous delivery of a locally submitted
4197 message (the -odf or -odi options), it no longer closes stderr before doing
4200 3. Implemented the mua_wrapper option.
4202 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4204 5. Implemented the functions header_remove(), header_testname(),
4205 header_add_at_position(), and receive_remove_recipient(), and exported them
4208 6. If an ACL "warn" statement specified the addition of headers, Exim already
4209 inserted X-ACL-Warn: at the start if there was no header name. However, it
4210 was not making this test for the second and subsequent header lines if
4211 there were newlines in the string. This meant that an invalid header could
4212 be inserted if Exim was badly configured.
4214 7. Allow an ACL "warn" statement to add header lines at the start or after all
4215 the Received: headers, as well as at the end.
4217 8. Added the rcpt_4xx retry error code.
4219 9. Added postmaster_mailfrom=xxx to callout verification option.
4221 10. Added mailfrom=xxxx to the callout verification option, for verify=
4224 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4225 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4227 12. Inserted some casts to stop certain compilers warning when using pointer
4228 differences as field lengths or precisions in printf-type calls (mostly
4229 affecting debugging statements).
4231 13. Added optional readline() support for -be (dynamically loaded).
4233 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4234 same clock tick as a message's arrival, so that its received time was the
4235 same as the "first fail" time on the retry record, and that message
4236 remained on the queue past the ultimate address timeout, every queue runner
4237 would try a delivery (because it was past the ultimate address timeout) but
4238 after another failure, the ultimate address timeout, which should have then
4239 bounced the address, did not kick in. This was a "< instead of <=" error;
4240 in most cases the first failure would have been in the next clock tick
4241 after the received time, and all would be well.
4243 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4244 being recognized when the domain list was tested by the match_domain
4245 condition in an expansion string.
4247 16. Added the ${str2b64: operator.
4249 17. Exim was always calling setrlimit() to set a large limit for the number of
4250 processes, without checking whether the existing limit was already
4251 adequate. (It did check for the limit on file descriptors.) Furthermore,
4252 errors from getrlimit() and setrlimit() were being ignored. Now they are
4253 logged to the main and panic logs, but Exim does carry on, to try to do its
4254 job under whatever limits there are.
4256 18. Imported PCRE 5.0.
4258 19. Trivial typo in log message " temporarily refused connection" (the leading
4261 20. If the log selector return_path_on_delivery was set and an address was
4262 redirected to /dev/null, the delivery process crashed because it assumed
4263 that a return path would always be set for a "successful" delivery. In this
4264 case, the whole delivery is bypassed as an optimization, and therefore no
4267 21. Internal re-arrangement: the function for sending a challenge and reading
4268 a response while authentication was assuming a zero-terminated challenge
4269 string. It's now changed to take a pointer and a length, to allow for
4270 binary data in such strings.
4272 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4274 23. Exim was not respecting finduser_retries when seeking the login of the
4275 uid under which it was called; it was always trying 10 times. (The default
4276 setting of finduser_retries is zero.) Also, it was sleeping after the final
4277 failure, which is pointless.
4279 24. Implemented tls_on_connect_ports.
4281 25. Implemented acl_smtp_predata.
4283 26. If the domain in control=submission is set empty, Exim assumes that the
4284 authenticated id is a complete email address when it generates From: or
4285 Sender: header lines.
4287 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4288 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4289 chown and chgrp in /bin and hostname in /usr/bin.
4291 28. Exim was keeping the "process log" file open after each use, just as it
4292 does for the main log. This opens the possibility of it remaining open for
4293 long periods when the USR1 signal hits a daemon. Occasional processlog
4294 errors were reported, that could have been caused by this. Anyway, it seems
4295 much more sensible not to leave this file open at all, so that is what now
4298 29. The long-running daemon process does not normally write to the log once it
4299 has entered its main loop, and it closes the log before doing so. This is
4300 so that log files can straightforwardly be renamed and moved. However,
4301 there are a couple of unusual error situations where the daemon does write
4302 log entries, and I had neglected to close the log afterwards.
4304 30. The text of an SMTP error response that was received during a remote
4305 delivery was being truncated at 512 bytes. This is too short for some of
4306 the long messages that one sometimes sees. I've increased the limit to
4309 31. It is now possible to make retry rules that apply only when a message has a
4310 specific sender, in particular, an empty sender.
4312 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4313 it possible to be selective about when SMTP synchronization is enforced.
4315 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4317 32. Implemented hosts_connection_nolog.
4319 33. Added an ACL for QUIT.
4321 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4324 35. Added mailbox_size and mailbox_filecount to appendfile.
4326 36. Added control = no_multiline_responses to ACLs.
4328 37. There was a bug in the logic of the code that waits for the clock to tick
4329 in the case where the clock went backwards by a substantial amount such
4330 that the microsecond fraction of "now" was more than the microsecond
4331 fraction of "then" (but the whole seconds number was less).
4333 38. Added support for the libradius Radius client library this is found on
4334 FreeBSD (previously only the radiusclient library was supported).
4340 1. When certain lookups returned multiple values in the form name=value, the
4341 quoting of the values was not always being done properly. Specifically:
4342 (a) If the value started with a double quote, but contained no whitespace,
4344 (b) If the value contained whitespace other than a space character (i.e.
4345 tabs or newlines or carriage returns) it was not quoted.
4346 This fix has been applied to the mysql and pgsql lookups by writing a
4347 separate quoting function and calling it from the lookup code. The fix
4348 should probably also be applied to nisplus, ibase and oracle lookups, but
4349 since I cannot test any of those, I have not disturbed their existing code.
4351 2. A hit in the callout cache for a specific address caused a log line with no
4352 reason for rejecting RCPT. Now it says "Previous (cached) callout
4353 verification failure".
4355 3. There was an off-by-one bug in the queryprogram router. An over-long
4356 return line was truncated at 256 instead of 255 characters, thereby
4357 overflowing its buffer with the terminating zero. As well as fixing this, I
4358 have increased the buffer size to 1024 (and made a note to document this).
4360 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4361 when Exim is waiting for an SMTP response from a remote server, Exim
4362 restarts its select() call on the socket, thereby resetting its timeout.
4363 This is not a problem when such interrupts are rare. Somebody set up a cron
4364 job to run exiwhat every 2 minutes, which is less than the normal select()
4365 timeout (5 or 10 minutes). This meant that the select() timeout never
4366 kicked in because it was always reset. I have fixed this by comparing the
4367 time when an interrupt arrives with the time at the start of the first call
4368 to select(). If more time than the timeout has elapsed, the interrupt is
4369 treated as a timeout.
4371 5. Some internal re-factoring in preparation for the addition of Sieve
4372 extensions (by MH). In particular, the "personal" test is moved to a
4373 separate function, and given an option for scanning Cc: and Bcc: (which is
4374 not set for Exim filters).
4376 6. When Exim created an email address using the login of the caller as the
4377 local part (e.g. when creating a From: or Sender: header line), it was not
4378 quoting the local part when it contained special characters such as @.
4380 7. Installed new OpenBSD configuration files.
4382 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4383 try to make them clearer.
4385 9. Callout options, other than the timeout value, were being ignored when
4386 verifying sender addresses in header lines. For example, when using
4388 verify = header_sender/callout=no_cache
4390 the cache was (incorrectly) being used.
4392 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4393 only the Cygwin environment.
4395 11. When return_path_on_delivery was set as a log selector, if different remote
4396 addresses in the same message used different return paths and parallel
4397 remote delivery occurred, the wrong values would sometimes be logged.
4398 (Whenever a remote delivery process finished, the return path value from
4399 the most recently started remote delivery process was logged.)
4401 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4402 lines when AUTH and/or TLS are in use. This is the "received protocol"
4403 field. Exim used to use "asmtp" for authenticated SMTP, without any
4404 indication (in the protocol name) for TLS use. Now it follows the RFC and
4405 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4406 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4407 names appear in log lines as well as in Received: header lines.
4409 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4410 extensions, and comparison tests, and to fix some bugs.
4412 14. Changes to the "personal" filter test:
4414 (1) The test was buggy in that it was just doing the equivalent of
4415 "contains" tests on header lines. For example, if a user's address was
4416 anne@some.where, the "personal" test would incorrectly be true for
4418 To: susanne@some.where
4420 This test is now done by extracting each address from the header in turn,
4421 and checking the entire address. Other tests that are part of "personal"
4422 are now done using regular expressions (for example, to check local parts
4423 of addresses in From: header lines).
4425 (2) The list of non-personal local parts in From: addresses has been
4426 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4427 taken from the Sieve specification recommendations.
4429 (3) If the message contains any header line starting with "List-" it is
4430 treated as non-personal.
4432 (4) The test for "circular" in the Subject: header line has been removed
4433 because it now seems ill-conceived.
4435 15. Minor typos in src/EDITME comments corrected.
4437 16. Installed latest exipick from John Jetmore.
4439 17. If headers_add on a router specified a text string that was too long for
4440 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4441 of string_sprintf() is now avoided.
4443 18. $message_body_size was not set (it was always zero) when running the DATA
4444 ACL and the local_scan() function.
4446 19. For the "mail" command in an Exim filter, no default was being set for
4447 the once_repeat time, causing a random time value to be used if "once" was
4448 specified. (If the value happened to be <= 0, no repeat happened.) The
4449 default is now 0s, meaning "never repeat". The "vacation" command was OK
4450 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4451 (I found it when inspecting the code).
4453 20. There is now an overall timeout for performing a callout verification. It
4454 defaults to 4 times the callout timeout, which applies to individual SMTP
4455 commands during the callout. The overall timeout applies when there is more
4456 than one host that can be tried. The timeout is checked before trying the
4457 next host. This prevents very long delays if there are a large number of
4458 hosts and all are timing out (e.g. when the network connections are timing
4459 out). The value of the overall timeout can be changed by specifying an
4460 additional sub-option for "callout", called "maxwait". For example:
4462 verify = sender/callout=5s,maxwait=20s
4464 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4465 to the end before writing, but this should make it even safer).
4467 22. Exim was forgetting that it had advertised PIPELINING for the second and
4468 subsequent messages on an SMTP connection. It was also not resetting its
4469 memory on STARTTLS and an internal HELO.
4471 23. When Exim logs an SMTP synchronization error within a session, it now
4472 records whether PIPELINING has been advertised or not.
4474 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4475 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4476 rather than long int.
4478 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4480 26. Added the never_mail option to autoreply.
4486 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4487 crash if the getsockname() call failed; this can happen if a connection is
4488 closed very soon after it is established. The problem was simply in the
4489 order in which certain operations were done, causing Exim to try to write
4490 to the SMTP stream before it had set up the file descriptor. The bug has
4491 been fixed by making things happen in the correct order.
4497 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4498 before the connection was closed, thus losing the rejection response.
4500 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4501 some early Solaris releases, but causes trouble in current releases where
4502 socklen_t is defined.
4504 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4507 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4510 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4511 that releasing the top of it at the end releases what was used for sub-
4512 expansions (unless the block got too big). However, discard this block if
4513 the first thing is a variable or header, so that we can use its block when
4514 it is dynamic (useful for very large $message_headers, for example).
4516 6. Lookups now cache *every* query, not just the most recent. A new, separate
4517 store pool is used for this. It can be recovered when all lookup caches are
4518 flushed. Lookups now release memory at the end of their result strings.
4519 This has involved some general refactoring of the lookup sources.
4521 7. Some code has been added to the store_xxx() functions to reduce the amount
4522 of flapping under certain conditions.
4524 8. log_incoming_interface used to affect only the <= reception log lines. Now
4525 it causes the local interface and port to be added to several more SMTP log
4526 lines, for example "SMTP connection from", and rejection lines.
4528 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4530 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4532 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4533 could be overwritten at the end of the current message (or the start of a
4534 new message if it was set in a HELO ACL). The value is now preserved for
4535 the duration of the SMTP connection.
4537 12. If a transport had a headers_rewrite setting, and a matching header line
4538 contained an unqualified address, that address was qualified, even if it
4539 did not match any rewriting rules. The underlying bug was that the values
4540 of the flags that permit the existence of unqualified sender and recipient
4541 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4542 non-local messages, and by -bnq for local messages) were not being
4543 preserved with the message after it was received.
4545 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4546 "next input=" as part of the text comprising the host identity instead of
4547 the correct text. The code was using the same buffer for two different
4548 strings. However, depending on which order the printing function evaluated
4549 its arguments, the bug did not always show up. Under Linux, for example, my
4550 test suite worked just fine.
4552 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4553 doesn't work with some older versions of Perl. It has been changed to "my",
4554 which in any case is probably the better facility to use.
4556 15. A really picky compiler found some instances of statements for creating
4557 error messages that either had too many or two few arguments for the format
4560 16. The size of the buffer for calls to the DNS resolver has been increased
4561 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4562 for addresses that have a lot of PTR records. This alleviates a problem; it
4563 does not fully solve it.
4565 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4566 buffer now truncates the list and logs the incident, which is the same
4567 action as happens when Exim is looking up a host name and its aliases.
4568 Previously in this situation something unpredictable would happen;
4569 sometimes it was "internal error: store_reset failed".
4571 18. If a server dropped the connection unexpectedly when an Exim client was
4572 using GnuTLS and trying to read a response, the client delivery process
4573 crashed while trying to generate an error log message.
4575 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4576 string, for example:
4578 warn message = H1: something\nH2: something
4580 the text was added as a single header line from Exim's point of view
4581 though it ended up OK in the delivered message. However, searching for the
4582 second and subsequent header lines using $h_h2: did not work. This has been
4583 fixed. Similarly, if a system filter added multiple headers in this way,
4584 the routers could not see them.
4586 20. Expanded the error message when iplsearch is called with an invalid key to
4587 suggest using net-iplsearch in a host list.
4589 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4590 ACLs are no longer actually imposed (and a message to that effect is
4593 22. If a "gecos" field in a passwd entry contained escaped characters, in
4594 particular, if it contained a \" sequence, Exim got it wrong when building
4595 a From: or a Sender: header from that name. A second bug also caused
4596 incorrect handling when an unquoted " was present following a character
4597 that needed quoting.
4599 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4600 was not being matched caselessly.
4602 24. Arranged for all hyphens in the exim.8 source to be escaped with
4605 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4606 properly. Recipient callout cache records were still being keyed to include
4607 the sender, even when use_sender was set false. This led to far more
4608 callouts that were necessary. The sender is no longer included in the key
4609 when use_sender is false.
4611 26. Added "control = submission" modifier to ACLs.
4613 27. Added the ${base62d: operator to decode base 62 numbers.
4615 28. dnsdb lookups can now access SRV records.
4617 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4618 the configuration file.
4620 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4621 (-v) mode. This makes the output for a verbose queue run more intelligible.
4623 31. Added a use_postmaster feature to recipient callouts.
4625 32. Added the $body_zerocount variable, containing the number of binary zero
4626 bytes in the message body.
4628 33. The time of last modification of the "new" subdirectory is now used as the
4629 "mailbox time last read" when there is a quota error for a maildir
4632 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4634 35. Added +ignore_unknown as a special item in host lists.
4636 36. Code for decoding IPv6 addresses in host lists is now included, even if
4637 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4638 address was recognized as an IP address, but was then not correctly decoded
4639 into binary, causing unexpected and incorrect effects when compared with
4646 1. Very minor rewording of debugging text in manualroute to say "list of
4647 hosts" instead of "hostlist".
4649 2. If verify=header_syntax was set, and a header line with an unqualified
4650 address (no domain) and a large number of spaces between the end of the
4651 name and the colon was received, the reception process suffered a buffer
4652 overflow, and (when I tested it) crashed. This was caused by some obsolete
4653 code that should have been removed. The fix is to remove it!
4655 3. When running in the test harness, delay a bit after writing a bounce
4656 message to get a bit more predictability in the log output.
4658 4. Added a call to search_tidyup() just before forking a reception process. In
4659 theory, someone could use a lookup in the expansion of smtp_accept_max_
4660 per_host which, without the tidyup, could leave open a database connection.
4662 5. Added the variables $recipient_data and $sender_data which get set from a
4663 lookup success in an ACL "recipients" or "senders" condition, or a router
4664 "senders" option, similar to $domain_data and $local_part_data.
4666 6. Moved the writing of debug_print from before to after the "senders" test
4669 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4670 problems for message scanning, either using a data ACL, or using
4671 local_scan() because the Received: header was not generated till after they
4672 were called (in order to set the time as the time of reception completion).
4673 I have revised the way this works. The header is now generated after the
4674 body is received, but before the ACL or local_scan() are called. After they
4675 are run, the timestamp in the header is updated.
4681 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4682 before starting a queue runner without re-exec. This happened only when
4683 deliver_drop_privilege was set or when the Exim user was set to root. The
4684 effect of the bug was that timeouts during subsequent deliveries caused
4685 crashes instead of being properly handled. The handler is now left at its
4686 default (and expected) setting.
4688 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4689 message, again when deliver_drop_privilege is set or Exim is run as root.
4690 The bug described in (1) was not present in this case, but the tidying up
4691 of the other signals was missing. I have made the two cases consistent.
4693 3. The ignore_target_hosts setting on a manualroute router was being ignored
4694 for hosts that were looked up using the /MX notation.
4696 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4699 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4700 operated on the sender address. After changing the $sender_address to <>
4701 for the sender address verify, Exim was re-instated it as the original
4702 (before rewriting) address, but remembering that it had rewritten it, so it
4703 wasn't rewriting it again. This bug also had the effect of breaking the
4704 sender address verification caching when the sender address was rewritten.
4706 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4707 This has been changed so that if the ip literal address matches
4708 ignore_target_hosts, the router declines.
4710 7. Added expansion conditions match_domain, match_address, and match_local_
4711 part (NOT match_host).
4713 8. The placeholder for the Received: header didn't have a length field set.
4715 9. Added code to Exim itself and to exim_lock to test for a specific race
4716 condition that could lead to file corruption when using MBX delivery. The
4717 issue is with the lockfile that is created in /tmp. If this file is removed
4718 after a process has opened it but before that process has acquired a lock,
4719 there is the potential for a second process to recreate the file and also
4720 acquire a lock. This could lead to two Exim processes writing to the file
4721 at the same time. The added code performs the same test as UW imapd; it
4722 checks after acquiring the lock that its file descriptor still refers to
4723 the same named file.
4725 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4726 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4729 11. Added the smtp_active_hostname option. If used, this will typically be made
4730 to depend on the incoming interface address. Because $interface_address is
4731 not set up until the daemon has forked a reception process, error responses
4732 that can happen earlier (such as "too many connections") no longer contain
4735 12. If an expansion in a condition on a "warn" statement fails because a lookup
4736 defers, the "warn" statement is abandoned, and the next ACL statement is
4737 processed. Previously this caused the whole ACL to be aborted.
4739 13. Added the iplsearch lookup type.
4741 14. Added ident_timeout as a log selector.
4743 15. Added tls_certificate_verified as a log selector.
4745 16. Added a global option tls_require_ciphers (compare the smtp transport
4746 option of the same name). This controls incoming TLS connections.
4748 17. I finally figured out how to make tls_require_ciphers do a similar thing
4749 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4750 before starting the TLS session.
4752 18. Tabs are now shown as \t in -bP output.
4754 19. If the log selector return_path_on_delivery was set, Exim crashed when
4755 bouncing a message because it had too many Received: header lines.
4757 20. If two routers both had headers_remove settings, and the first one included
4758 a superfluous trailing colon, the final name in the first list and the
4759 first name in the second list were incorrectly joined into one item (with a
4760 colon in the middle).
4766 1. Added -C and -D options to the exinext utility, mainly to make it easier
4767 to include in the automated testing, but these could be helpful when
4768 multiple configurations are in use.
4770 2. The exinext utility was not formatting the output nicely when there was
4771 an alternate port involved in the retry record key, nor when there was a
4772 message id as well (for retries that were specific to a specific message
4773 and a specific host). It was also confused by IPv6 addresses, because of
4774 the additional colons they contain. I have fixed the IPv4 problem, and
4775 patched it up to do a reasonable job for IPv6.
4777 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4778 delivery, the log line now contains "pipelined" if PIPELINING was used.
4780 4. An SMTP transport process used to panic and die if the bind() call to set
4781 an explicit outgoing interface failed. This has been changed; it is now
4782 treated in the same way as a connect() failure.
4784 5. A reference to $sender_host_name in the part of a conditional expansion
4785 that was being skipped was still causing a DNS lookup. This no longer
4788 6. The def: expansion condition was not recognizing references to header lines
4789 that used bh_ and bheader_.
4791 7. Added the _cache feature to named lists.
4793 8. The code for checking quota_filecount in the appendfile transport was
4794 allowing one more file than it should have been.
4796 9. For compatibility with Sendmail, the command line option
4804 and sets the incoming protocol and host name (for trusted callers). The
4805 host name and its colon can be omitted when only the protocol is to be set.
4806 Note the Exim already has two private options, -pd and -ps, that refer to
4807 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4808 "s", but I don't think that's a major issue.
4810 10. A number of refactoring changes to the code, none of which should affect
4813 (a) The number of logging options was getting close to filling up the
4814 32-bit word that was used as a bit map. I have split them into two classes:
4815 those that are passed in the argument to log_write(), and those that are
4816 only ever tested independently outside of that function. These are now in
4817 separate 32-bit words, so there is plenty of room for expansion again.
4818 There is no change in the user interface or the logging behaviour.
4820 (b) When building, for example, log lines, the code previously used a
4821 macro that called string_cat() twice, in order to add two strings. This is
4822 not really sufficiently general. Furthermore, there was one instance where
4823 it was actually wrong because one of the argument was used twice, and in
4824 one call a function was used. (As it happened, calling the function twice
4825 did not affect the overall behaviour.) The macro has been replaced by a
4826 function that can join an arbitrary number of extra strings onto a growing
4829 (c) The code for expansion conditions now uses a table and a binary chop
4830 instead of a serial search (which was left over from when there were very
4831 few conditions). Also, it now recognizes conditions like "pam" even when
4832 the relevant support is not compiled in: a suitably worded error message is
4833 given if an attempt is made to use such a condition.
4835 11. Added ${time_interval:xxxxx}.
4837 12. A bug was causing one of the ddress fields not to be passed back correctly
4838 from remote delivery subprocesses. The field in question was not being
4839 subsequently used, so this caused to problems in practice.
4841 13. Added new log selectors queue_time and deliver_time.
4843 14. Might have fixed a bug in maildirsizefile handling that threw up
4844 "unexpected character" debug warnings, and recalculated the data
4845 unnecessarily. In any case, I expanded the warning message to give more
4848 15. Added the message "Restricted characters in address" to the statements in
4849 the default ACL that block characters like @ and % in local parts.
4851 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4852 Three changes have been made:
4854 (a) There was a serious bug; a negative response to MAIL caused the whole
4855 recipient domain to be cached as invalid, thereby blocking all messages
4856 to all local parts at the same domain, from all senders. This bug has
4857 been fixed. The domain is no longer cached after a negative response to
4858 MAIL if the sender used is not empty.
4860 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4863 (c) A new callout option, "use_sender" has been added for people who want
4864 the modified behaviour.
4870 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4873 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4874 indeed breaks things for older releases.
4876 3. Added additional logging to the case where there is a problem reading data
4877 from a filter that is running in a subprocess using a pipe, in order to
4878 try to track down a specific problem.
4880 4. Testing facility fudge: when running in the test harness and attempting
4881 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4882 getting "No route to host". Convert this to a timeout.
4884 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4887 6. Some OS don't have socklen_t but use size_t instead. This affects the
4888 fifth argument of getsockopt() amongst other things. This is now
4889 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4890 can be set for individual OS. I have set it for SunOS5, OSF1, and
4891 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4892 some earlier ones do not.
4894 7. Change 4.30/15 was not doing the test caselessly.
4896 8. The standard form for an IPv6 address literal was being rejected by address
4897 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4898 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4899 this, as well as the form without the "IPv6" on the front (but only when
4900 address literals are enabled, of course).
4902 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4904 10. Exim crashed if a message with an empty sender address specified by -f
4905 encountered a router with an errors_to setting. This could be provoked only
4906 by a command such as
4910 where an empty string was supplied; "<>" did not hit this bug.
4912 11. Installed PCRE release 4.5.
4914 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4915 remained set. It is now erased.
4917 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4918 times from message ids (which are base 36 rather than the normal 62).
4920 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4921 were being counted as actual protocol errors, and logged if the log
4922 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4923 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4924 and DATA following a set of rejected RCPTs do not count as protocol errors.
4925 In other words, Exim assumes they were pipelined, though this may not
4926 actually be the case. Of course, in all cases the client gets an
4927 appropriate error code.
4929 15. If a lookup fails in an ACL condition, a message about the failure may
4930 be available; it is used if testing the ACL cannot continue, because most
4931 such messages specify what the cause of the deferral is. However, some
4932 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4933 that caused an old message to be retained and used if a later statement
4934 caused a defer, replacing the real cause of the deferral.
4936 16. If an IP address had so many PTR records that the DNS lookup buffer
4937 was not large enough to hold them, Exim could crash while trying to process
4938 the truncated data. It now detects and logs this case.
4940 17. Further to 4.21/58, another change has been made: if (and only if) the
4941 first line of a message (the first header line) ends with CRLF, a bare LF
4942 in a subsequent header line has a space inserted after it, so as not to
4943 terminate the header.
4945 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4946 unnecessarily, used atoi() instead of strtol(), and didn't check the
4947 termination when getting file sizes from file names by regex.
4949 19. Completely re-implemented the support for maildirsize files, in the light
4950 of a number of problems with the previous contributed implementation
4951 (4.30/29). In particular:
4953 . If the quota is zero, the maildirsize file is maintained, but no quota is
4956 . If the maildir directory does not exist, it is created before any attempt
4957 to write a maildirsize file.
4959 . The quota value in the file is just a cache; if the quota is changed in
4960 the transport, the new value overrides.
4962 . A regular expression is available for excluding directories from the
4965 20. The autoreply transport checks the characters in options that define the
4966 message's headers; it allows continued headers, but it was checking with
4967 isspace() after an embedded newline instead of explicitly looking for a
4970 21. If all the "regular" hosts to which an address was routed had passed their
4971 expiry times, and had not reached their retry times, the address was
4972 bounced, even if fallback hosts were defined. Now Exim should go on to try
4975 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4976 equivalent code in the SMTP transport. Some hosts send humungous responses
4977 to HELO/EHLO, more than 1024 it seems.
4979 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4980 gives compiler warnings in some environments. I've now done it "properly",
4983 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4984 (because of problems with the built-in one) was declared to return uschar *
4985 instead of char *, causing compiler failure.
4987 25. Fixed a file descriptor leak when processing alias/forward files.
4989 26. Fixed a minor format string issue in dbfn.c.
4991 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4993 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4994 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4997 29. When a sender address is verified, it is cached, to save repeating the test
4998 when there is more than one recipient in a message. However, when the
4999 verification involves a callout, it is possible for different callout
5000 options to be set for different recipients. It is too complicated to keep
5001 track of this in the cache, so now Exim always runs a verification when a
5002 callout is required, relying on the callout cache for the optimization.
5003 The overhead is duplication of the address routing, but this should not be
5006 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5007 to be verified with callout=postmaster, and the main callout worked but the
5008 postmaster check failed, the verification correctly failed. However, if a
5009 subsequent RCPT command asked for sender verification *without* the
5010 postmaster check, incorrect caching caused this verification also to fail,
5013 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5014 it was not caching the DNS options (qualify_single, search_parents) that
5015 were used when the lookup failed. A subsequent lookup with different
5016 options therefore always gave the same answer, though there were cases
5017 where it should not have. (Example: a "domains = !$mx_any" option on a
5018 dnslookup router: the "domains" option is always processed without any
5019 widening, but the router might have qualify_single set.) Now Exim uses the
5020 cached value only when the same options are set.
5022 32. Added John Jetmore's "exipick" utility to the distribution.
5024 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5025 than a timeout (e.g. a certificate is required, and is not provided), an
5026 Exim server now closes the connection immediately. Previously it waited for
5027 the client to close - but if the client is SSL, it seems that they each
5028 wait for each other, leading to a delay before one of them times out.
5030 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5031 maintained 0.8.x compatibility because I don't think many are using it, and
5032 it is clearly obsolete.
5034 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5037 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5038 client certificate was expired. A simple patch fixes this, though I don't
5039 understand the full logic of why the verify callback is called multiple
5042 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5043 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5044 which causes problems with some clients (such as the Certicom SSL Plus
5045 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5046 disables the coutermeasure allowing Eudora to connect."
5048 38. Exim was not checking that a write() to a log file succeeded. This could
5049 lead to Bad Things if a log got too big, in particular if it hit a file
5050 size limit. Exim now panics and dies if it cannot write to a log file, just
5051 as it does if it cannot open a log file.
5053 39. Modified OS/Makefile-Linux so that it now contains
5055 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5057 The two -D definitions ensure that Exim is compiled with large file
5058 support, which makes it possible to handle log files that are bigger than
5061 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5062 instance) a domain was checked against a named list that involved a lookup,
5063 causing $domain_data to be set, then another domain was checked against the
5064 same list, then the first domain was re-checked, the value of $domain_data
5065 after the final check could be wrong. In particular, if the second check
5066 failed, it could be set empty. This bug probably also applied to
5069 41. The strip_trailing_dot option was not being applied to the address given
5070 with the -f command-line option.
5072 42. The code for reading a message's header from the spool was incrementing
5073 $received_count, but never initializing it. This meant that the value was
5074 incorrect (doubled) while delivering a message in the same process in which
5075 it was received. In the most common configuration of Exim, this never
5076 happens - a fresh exec is done - but it can happen when
5077 deliver_drop_privilege is set.
5079 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5080 it now includes up to 150 characters of the unexpected data in the log
5083 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5084 and building data strings. The size of both of these buffers was 10 000
5085 bytes - far larger than anybody would *ever* want, thought I. Needless to
5086 say, somebody hit the limit. I have increased the maximum line length to
5087 20 000 and the maximum data length of concatenated lines to 100 000. I have
5088 also fixed two bugs, because there was no checking on these buffers. Tsk,
5089 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5090 buffer is too small.
5092 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5093 lsearch lookups. Now it does.
5095 46. When parsing a route_list item in a manualroute router, a fixed-length
5096 buffer was used for the list of hosts. I made this 1024 bytes long,
5097 thinking that nobody would ever have a list of hosts that long. Wrong.
5098 Somebody had a whole pile of complicated expansion conditions, and the
5099 string was silently truncated, leading to an expansion error. It turns out
5100 that it is easier to change to an unlimited length (owing to other changes
5101 that have happened since this code was originally written) than to build
5102 structure for giving a limitation error. The length of the item that
5103 expands into the list of hosts is now unlimited.
5105 47. The lsearch lookup could not handle data where the length of text line was
5106 more than 4095 characters. Such lines were truncated, leading to shortened
5107 data being returned. It should now handle lines of any length.
5109 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5110 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5113 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5114 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5115 no longer happens in comment lines. A list of replacements is now placed
5116 at the head of all of the source files, except those whose only change is
5117 to replace PERL_COMMAND in the very first #! line.
5119 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5120 messages on the queue, with a bottom-up merge sort, using code contributed
5121 by Michael Haardt. This should make operations like -bp somewhat faster on
5122 large queues. It won't affect queue runners, except when queue_run_in_order
5125 51. Installed eximstats 1.31 in the distribution.
5127 52. Added support for SRV lookups to the dnslookup router.
5129 53. If an ACL referred to $message_body or $message_body_end, the value was not
5130 reset for any messages that followed in the same SMTP session.
5132 54. The store-handling optimization for building very long strings was not
5133 differentiating between the different store pools. I don't think this
5134 actually made any difference in practice, but I've tidied it.
5136 55. While running the routers to verify a sender address, $sender_address
5137 was still set to the sender address. This is wrong, because when routing to
5138 send a bounce to the sender, it would be empty. Therefore, I have changed
5139 it so that, while verifying a sender address, $sender_address is set to <>.
5140 (There is no change to what happens when verifying a recipient address.)
5142 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5143 target A or AAAA records (if not already returned) without resetting the
5144 qualify_single or search_parents options of the DNS resolver. These are
5145 inappropriate in this case because the targets of MX and SRV records must
5146 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5147 target that, when qualified, matched something in the local domain. These
5148 two options are now turned off when doing these lookups.
5150 57. It seems that at least some releases of Reiserfs (which does not have the
5151 concept of a fixed number of inodes) returns zero and not -1 for the
5152 number of available inodes. This interacted badly with check_spool_inodes,
5153 which assumed that -1 was the "no such thing" setting. What I have done is
5154 to check that the total number of inodes is greater than zero before doing
5155 the test of how many are available.
5157 58. When a "warn" ACL statement has a log_message modifier, the message is
5158 remembered, and not repeated. This is to avoid a lot of repetition when a
5159 message has many recipients that cause the same warning to be written.
5160 Howewer, Exim was preserving the list of already written lines for an
5161 entire SMTP session, which doesn't seem right. The memory is now reset if a
5162 new message is started.
5164 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5165 debugging output unless log_rewrite was also set.
5167 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5168 of (int)(handle) when we know that handle contains (void *)(-1).
5170 61. The Exim daemon panic-logs an error return when it closes the incoming
5171 connection. However "connection reset by peer" seems to be common, and
5172 isn't really an error worthy of noting specially, so that particular error
5175 62. When Exim is trying to find all the local interfaces, it used to panic and
5176 die if the ioctl to get the interface flags failed. However, it seems that
5177 on at least one OS (Solaris 9) it is possible to have an interface that is
5178 included in the list of interfaces, but for which you get a failure error
5179 for this call. This happens when the interface is not "plumbed" into a
5180 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5181 failure of the "get flags" call assumes that the interface is down.
5183 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5184 makes life easier for people who are doing arithmetic on fields extracted
5185 from dates, where you often get leading zeros that should not be
5186 interpreted as octal.
5188 64. Added qualify_domain to the redirect router, to override the global
5191 65. If a pathologically long header line contained very many addresses (the
5192 report of this problem mentioned 10 000) and each of them was rewritten,
5193 Exim could use up a very large amount of memory. (It kept on making new
5194 copies of the header line as it rewrote, and never released the old ones.)
5195 At the expense of a bit more processing, the header rewriting function has
5196 been changed so that it no longer eats memory in this way.
5198 66. The generation of the Received: header has been moved from the time that a
5199 message starts to be received, to the time that it finishes. The timestamp
5200 in the Received: header should now be very close to that of the <= log
5201 line. There are two side-effects of this change:
5203 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5204 logged header lines no longer include the local Received: line, because
5205 it has not yet been created. The same applies to a copy of the message
5206 that is returned to a non-SMTP sender when a message is rejected.
5208 (b) When a filter file is tested using -bf, no additional Received: header
5209 is added to the test message. After some thought, I decided that this
5212 This change does not affect the value of $received_for. It is still set
5213 after address rewriting, but before local_scan() is called.
5215 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5217 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5218 gave an unhelpful panic error message, and a defer error. I have managed to
5219 change this behaviour so that it now rejects any supplied certificate,
5220 which seems right, as the list of acceptable certificates is empty.
5222 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5223 gave an unhelpful defer error. I have not managed to make this reject any
5224 supplied certificates, but the error message it gives is "no certificate
5225 supplied", which is not helpful.
5227 70. exigrep's output now also includes lines that are not associated with any
5228 message, but which match the given pattern. Implemented by a patch from
5229 Martin Sluka, which also tidied up the Perl a bit.
5231 71. Recipient callout verification, like sender verification, was using <> in
5232 the MAIL FROM command. This isn't really the right thing, since the actual
5233 sender may affect whether the remote host accepts the recipient or not. I
5234 have changed it to use the actual sender in the callout; this means that
5235 the cache record is now keyed on a recipient/sender pair, not just the
5236 recipient address. There doesn't seem to be a real danger of callout loops,
5237 since a callout by the remote host to check the sender would use <>.
5238 [SEE ABOVE: changed after hitting problems.]
5240 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5241 temporary errors. However, in the case of such a code being given after
5242 the end of a data transmission (i.e. after ".") Exim was failing to write
5243 a retry record for the message. (Yes, there was some broken host that was
5244 actually sending 8xx at this point.)
5246 73. An unknown lookup type in a host list could cause Exim to panic-die when
5247 the list was checked. (An example that provoked this was putting <; in the
5248 middle of a list instead of at the start.) If this happened during a DATA
5249 ACL check, a -D file could be left lying around. This kind of configuration
5250 error no longer causes Exim to die; instead it causes a defer errror. The
5251 incident is still logged to the main and panic logs.
5253 74. Buglet left over from Exim 3 conversion. The message "too many messages
5254 in one connection" was written to the rejectlog but not the mainlog, except
5255 when address rewriting (yes!) was being logged.
5257 75. Added write_rejectlog option.
5259 76. When a system filter was run not as root (that is, when system_filter_user
5260 was set), the values of the $n variables were not being returned to the
5261 main process; thus, they were not subsequently available in the $sn
5264 77. Added +return_path_on_delivery log selector.
5266 78. A connection timeout was being treated differently from recipients deferred
5267 when testing hosts_max_try with a message that was older than the host's
5268 retry timeout. (The host should not be counted, thus allowing all hosts to
5269 be tried at least once before bouncing.) This may have been the cause of an
5270 occasionally reported bug whereby a message would remain on the queue
5271 longer than the retry timeout, but would be bounced if a delivery was
5272 forced. I say "may" because I never totally pinned down the problem;
5273 setting up timeout/retry tests is difficult. See also the next item.
5275 79. The ultimate address timeout was not being applied to errors that involved
5276 a combination of host plus message (for example, a timeout on a MAIL
5277 command). When an address resolved to a number of possible hosts, and they
5278 were not all tried for each delivery (e.g. because of hosts_max_try), a
5279 message could remain on the queue longer than the retry timeout.
5281 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5284 81. Fixed an obscure SMTP outgoing bug which required at least the following
5285 conditions: (a) there was another message waiting for the same server;
5286 (b) the server returned 5xx to all RCPT commands in the first message so
5287 that the message was not completed; (c) the server dropped the connection
5288 or gave a negative response to the RSET that Exim sends to abort the
5289 transaction. The observed case was a dropped connection after DATA that had
5290 been sent in pipelining mode. That is, the server had advertised PIPELINING
5291 but was not implementing it correctly. The effect of the bug was incorrect
5292 behaviour, such as trying another host, and this could lead to a crash.
5298 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5299 and daemon.c were passed as pointers to ints; they should have been
5300 pointers to socklen_t variables (which are typically unsigned ints).
5302 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5305 3. Fixed a really odd bug that affected only the testing scheme; patching a
5306 certain fixed string in the binary changed the value of another string that
5307 happened to be identical to the end of the original first string.
5309 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5310 name", it returns that address as the IP address. On some operating
5311 systems (e.g. Solaris), it also passes back the IP address string as the
5312 "host name". However, on others (e.g. Linux), it passes back an empty
5313 string. Exim wasn't checking for this, and was changing the host name to an
5314 empty string, assuming it had been canonicized.
5316 5. Although rare, it is permitted to have more than one PTR record for a given
5317 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5318 all the names associated with an address, because they do in Solaris.
5319 However, it seems that they do not in Linux for data that comes from the
5320 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5321 I found this out when I moved to a new Linux workstation and tried to run
5322 the Exim test suite.
5324 To get round this problem I have changed the code so that it now does its
5325 own call to the DNS to look up PTR records when searching for a host name.
5326 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5327 addresses that are only in /etc/hosts are still found.
5329 This behaviour is, however, controlled by an option called host_lookup_
5330 order, which defaults to "bydns:byaddr". If people want to use the other
5331 order, or indeed, just use one or the other means of lookup, they can
5332 specify it in this variable.
5334 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5335 some operating systems, this comes back from gethostbyaddr() as an empty
5336 string, and this is what Exim used to test for. However, it seems that in
5337 other systems, "." is yielded. Exim now tests for this case too.
5339 7. The values of check_spool_space and check_log_space are now held internally
5340 as a number of kilobytes instead of an absolute number of bytes. If a
5341 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5342 kilobyte. This means that much larger values can be stored.
5344 8. Exim monitor: an attempt to get the action menu when not actually pointing
5345 at a message produces an empty menu entitled "No message selected". This
5346 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5347 no entries in it ("Shell widget menu has zero width and/or height"). So I
5348 have added a single, blank menu entry in this case.
5350 9. Added ${quote_local_part.
5352 10. MIME decoding is now applied to the contents of Subject: header lines when
5355 11. Now that a reference to $sender_host_address automatically causes a reverse
5356 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5357 host lookup before query-style lookups in lists that might use this
5358 variable. This has therefore been abolished, and the "net-" prefix is no
5359 longer necessary for query-style lookups.
5361 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5362 have been a typo for LFLAGS, so it has been changed.
5364 13. The install script calls Exim with "-C /dev/null" in order to find the
5365 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5366 to be output. Howeve, since Exim outputs its version number before the
5367 error, it didn't break the script. It just looked ugly. I fixed this by
5368 always allowing "-C /dev/null" if the caller is root.
5370 14. Ignore overlarge ACL variable number when reading spool file - insurance
5371 against a later release with more variables having written the file.
5373 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5374 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5375 this, as well as the form without the "IPv6" on the front.
5377 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5378 OS/Makefile-Darwin file.
5380 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5381 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5383 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5384 variable. However, some OS use ints for inodes. I've added cast to long int
5385 to get rid of the compiler warning.
5387 19. I had forgotten to lock out "/../" in configuration file names when
5388 ALT_CONFIG_PREFIX was set.
5390 20. Routers used for verification do not need to specify transports. However,
5391 if such a router generated a host list, and callout was configured, Exim
5392 crashed, because it could not find a port number from the (non-existent)
5393 transport. It now assumes port 25 in this circumstance.
5395 21. Added the -t option to exigrep.
5397 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5398 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5399 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5400 them is set, LOOKUP_LSEARCH is forced.
5402 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5404 24. Added sender and host information to the "rejected by local_scan()" log
5405 line; previously there was no indication of these.
5407 25. Added .include_if_exists.
5409 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5410 Linux. It turns out that not all file systems support this. Apparently some
5411 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5412 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5413 not supported on the file descriptor, is now ignored when Exim is trying to
5414 sync a directory. This applies only to Linux.
5416 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5418 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5419 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5422 29. Added support for maildirsize files from supplied patch (modified a bit).
5424 30. The use of :fail: followed by an empty string could lead Exim to respond to
5425 sender verification failures with (e.g.):
5427 550 Verification failed for <xxx>
5428 550 Sender verify failed
5430 where the first response line was missing the '-' that indicates it is not
5431 the final line of the response.
5433 31. The loop for finding the name of the user that called Exim had a hardwired
5434 limit of 10; it now uses the value of finduser_retries, which is used for
5435 all other user lookups.
5437 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5440 33. Exim was neglecting to zero errno before one call of strtol() when
5441 expanding a string and expecting an integer value. On some systems this
5442 resulted in spurious "integer overflow" errors. Also, it was casting the
5443 result into an int without checking.
5445 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5446 did not work. The code looks as if it has *never* worked, though it appears
5447 to have been documented since at least releast 1.62. I have made it work.
5449 35. The "timeout_DNS" error in retry rules, also documented since at least
5450 1.62, also never worked. As it isn't clear exactly what this means, and
5451 clearly it isn't a major issue, I have abolished the feature by treating it
5452 as "timeout", and writing a warning to the main and panic logs.
5454 36. The display of retry rules for -brt wasn't always showing the error code
5457 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5458 timeout_connect_A, timeout_connect_MX.
5460 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5461 to the empty sender.
5463 39. The daemon was not analysing the content of -oX till after it had closed
5464 stderr and disconnected from the controlling terminal. This meant that any
5465 syntax errors were only noted on the panic log, and the return code from
5466 the command was 0. By re-arranging the code a little, I've made the
5467 decoding happen first, so such errors now appear on stderr, and the return
5468 code is 1. However, the actual setting up of the sockets still happens in
5469 the disconnected process, so errors there are still only recorded on the
5472 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5473 connections (as happens on some IP stacks) was logged at start up time as
5474 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5475 it from "IPv6 and IPv4", which means that two separate sockets are being
5478 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5479 says whether AF_INET or AF_INET6 was passed as an argument.
5481 42. Exiwhat output was messed up when time zones were included in log
5484 43. Exiwhat now gives more information about the daemon's listening ports,
5485 and whether -tls-on-connect was used.
5487 44. The "port" option of the smtp transport is now expanded.
5489 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5490 silently ignored. Now an error message is written to the main and panic
5493 46. There's a new ACL modifier called "logwrite" which writes to a log file
5494 as soon as it is encountered.
5496 47. Added $local_user_uid and $local_user_gid at routing time.
5498 48. Exim crashed when trying to verify a sender address that was being
5501 49. Exim was recognizing only a space character after ".include". It now also
5502 recognizes a tab character.
5504 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5505 extracting the relevant information from the specification. The man page no
5506 longer contains scrambled data for the -d option, and I've added a section
5507 at the front about calling Exim under different names.
5509 51. Added "extra_headers" argument to the "mail" command in filter files.
5511 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5514 53. Installed eximstats 1.29.
5516 54. Added transport_filter_timeout as a generic transport option.
5518 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5519 Cc: header lines. This was required by RFC 822, but it not required by RFC
5522 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5523 incoming messages that did not have them. Now it does so only if the
5524 message originates locally, that is, if there is no associated remote host
5525 address. When Resent- header lines are present, this applies to the Resent-
5526 lines rather than the non-Resent- lines.
5528 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5529 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5531 58. Messages for configuration errors now include the name of the main
5532 configuration file - useful now that there may be more than one file in a
5533 list (.included file names were always shown).
5535 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5536 for those rare installations that do not start the daemon as root or run it
5537 setuid root. I've cut out the call to initgroups() if the daemon is not
5540 60. The Exim user and group can now be bound into the binary as text strings
5541 that are looked up at the start of Exim's processing.
5543 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5545 62. Added $mailstore_basename variable.
5547 63. Installed patch to sieve.c from Michael Haardt.
5549 64. When Exim failed to open the panic log after failing to open the main log,
5550 the original message it was trying to log was written to stderr and debug
5551 output, but if they were not available (the usual case in production), it
5552 was lost. Now it is written to syslog before the two lines that record the
5553 failures to open the logs.
5555 65. Users' Exim filters run in subprocesses under the user's uid. It is
5556 possible for a "deliver" command or an alias in a "personal" command to
5557 provoke an address rewrite. If logging of address rewriting is configured,
5558 this fails because the process is not running as root or exim. There may be
5559 a better way of dealing with this, but for the moment (because 4.30 needs
5560 to be released), I have disabled address rewrite logging when running a
5561 filter in a non-root, non-exim process.
5567 1. The buildconfig auxiliary program wasn't quoting the value set for
5568 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5569 not defined. This bug was masked in 4.22 by the effect that was fixed in
5572 2. Some messages that were rejected after a message id was allocated were
5573 shown as "incomplete" by exigrep. It no longer does this for messages that
5574 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5576 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5577 have allow_domain_literals set, the ID did not get logged in the <= line.
5578 Domain literals are now always recognized in Message-ID: header lines.
5580 4. The first argument for a ${extract expansion item is the key name or field
5581 number. Leading and trailing spaces in this item were not being ignored,
5582 causing some misleading effects.
5584 5. When deliver_drop_privilege was set, single queue runner processes started
5585 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5586 same command in the process it spins off) were not dropping privilege.
5588 6. When the daemon running as "exim" started a queue runner, it always
5589 re-executed Exim in the spun-off process. This is a waste of effort when
5590 deliver_drop_privilege is set. The new process now just calls the
5591 queue-runner function directly.
5597 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5600 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5601 ignored. Though the use of -oP was forcing the writing of a pid file, it
5602 was always written to the default place.
5604 3. If the message "no IP address found for host xxxx" is generated during
5605 incoming verification, it is now followed by identification of the incoming
5606 connection (so you can more easily find what provoked it).
5608 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5610 5. Added some features to "harden" Exim a bit more against certain attacks:
5612 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5613 be put in Local/Makefile. This is like the never_users runtime option,
5614 but it cannot be overridden. The default setting is "root".
5616 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5617 prefix string with which any file named in a -C command line option
5620 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5621 is retained for -C and -D only if the caller of Exim is root. Without
5622 it, the exim user may also use -C and -D and retain privilege.
5624 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5625 command line option is disabled.
5627 6. Macro names set by the -D option must start with an upper case letter, just
5628 like macro names defined in the configuration file.
5630 7. Added "dereference=" facility to LDAP.
5632 8. Two instances of the typo "uknown" in the source files are fixed.
5634 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5635 the Configure-Makefile script screwed up while processing it.
5637 10. Incorporated PCRE 4.4.
5639 11. The SMTP synchronization check was not operating right at the start of an
5640 SMTP session. For example, it could not catch a HELO sent before the client
5641 waited for the greeting. There is now a check for outstanding input at the
5642 point when the greeting is written. Because of the duplex, asynchronous
5643 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5644 way, but not yet received, when the check is performed.
5646 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5647 on TCP/IP sockets, because this apparently causes some broken clients to
5650 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5651 unchanged) from the Cygwin maintainer.
5653 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5654 was supported instead of testing for mbx. Effectively a typo.
5656 15. The spa authenticator server code was not checking that the input it
5657 received was valid base64.
5659 16. The debug output line for the "set" modifier in ACLs was not showing the
5660 name of the variable that was being set.
5662 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5664 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5665 lookup on its own. Something else was needed to trigger the lookup. For
5666 example, a match in host_lookup or the need for a host name in a host list.
5667 Now, if $sender_host_name is referenced and the host name has not yet been
5668 looked up, a lookup is performed. If the lookup fails, the variable remains
5669 empty, and $host_lookup_failed is set to "1".
5671 19. Added "eqi" as a case-independent comparison operator.
5673 20. The saslauthd authentication condition could segfault if neither service
5674 nor realm was specified.
5676 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5677 error message that was logged was misleading, and incoming SMTP
5678 connections were dropped. The message is now more accurate, and temporary
5679 errors are given to SMTP connections.
5681 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5682 (except RSET) with a 421 error, until QUIT is received. However, it was
5683 failing to send a response to QUIT.
5685 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5686 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5687 ACL is now run after the helo_try_verify_hosts code.
5689 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5690 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5691 case-independent, so other case variants are also recognized). Apparently
5692 some systems use these upper case variants.
5694 25. If more than two messages were waiting for the same host, and a transport
5695 filter was specified for the transport, Exim sent two messages over the
5696 same TCP/IP connection, and then failed with "socket operation on non-
5697 socket" when it tried to send the third.
5699 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5701 27. The extern definition of crypt16() in expand.c was not being excluded when
5702 the OS had its own crypt16() function.
5704 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5705 as a preferred synonym for return_size_limit, both as an option and as an
5708 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5710 30. Changed the default configuration ACL to relax the local part checking rule
5711 for addresses that are not in any local domains. For these addresses,
5712 slashes and pipe symbols are allowed within local parts, but the sequence
5713 /../ is explicitly forbidden.
5715 31. SPA server authentication was not clearing the challenge buffer before
5718 32. log_message in a "warn" ACL statement was writing to the reject log as
5719 well as to the main log, which contradicts the documentation and doesn't
5720 seem right (because no rejection is happening). So I have stopped it.
5722 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5723 However, I am unable to do any testing of this.
5725 34. Fixed an infelicity in the appendfile transport. When checking directories
5726 for a mailbox, to see if any needed to be created, it was accidentally
5727 using path names with one or more superfluous leading slashes; tracing
5728 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5730 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5731 contents are added to the log line that is written for every discarded
5732 recipient. (Previously a log_message setting was ignored.)
5734 36. The ${quote: operator now quotes the string if it is empty.
5736 37. The install script runs exim in order to find its version number. If for
5737 some reason other than non-existence or emptiness, which it checks, it
5738 could not run './exim', it was installing it with an empty version number,
5739 i.e. as "exim-". This error state is now caught, and the installation is
5742 38. An argument was missing from the function that creates an error message
5743 when Exim fails to connect to the socket for saslauthd authentication.
5744 This could cause Exim to crash, or give a corrupted message.
5746 39. Added isip, isip4, and isip6 to ${if conditions.
5748 40. The ACL variables $acl_xx are now saved with the message, and can be
5749 accessed later in routers, transports, and filters.
5751 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5752 strings in the file are not string-expanded.
5754 42. If a MAIL command specified a SIZE value that was too large to fit into an
5755 int variable, the check against message_size_limit failed. Such values are
5756 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5757 day this will have to be increased, but I don't think I want to be around
5758 when emails are that large.
5765 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5766 iconv() is not standard in FreeBSD.
5768 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5769 IPv6 enabled. The observed symptom was a segmentation fault on return from
5770 the function os_common_find_running_interfaces() in src/os.c.
5772 3. In the check_special_case() function in daemon.c I had used "errno" as an
5773 argument name, which causes warnings on some systems. This was basically a
5774 typo, since it was named "eno" in the comments!
5776 4. The code that waits for the clock to tick (at a resolution of some fraction
5777 of a second) so as to ensure message-id uniqueness was always waiting for
5778 at least one whole tick, when it could have waited for less. [This is
5779 almost certainly not relevant at current processor speeds, where it is
5780 unlikely to ever wait at all. But we try to future-proof.]
5782 5. The function that sleeps for a time interval that includes fractions of a
5783 second contained a race. It did not block SIGALRM between setting the
5784 timer, and suspending (a couple of lines later). If the interval was short
5785 and the sigsuspend() was delayed until after it had expired, the suspension
5786 never ended. On busy systems this could lead to processes getting stuck for
5789 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5790 process, before it forks any delivery processes. The open lookup caching
5791 mechanism meant that the open file or database connection was passed into
5792 the delivery process. The problem was that delivery processes always tidy
5793 up cached lookup data. This could cause a problem for the next delivery
5794 process started by the queue runner, because the external queue runner
5795 process does not know about the closure. So the next delivery process
5796 still has data in the lookup cache. In the case of a file lookup, there was
5797 no problem because closing a file descriptor in a subprocess doesn't affect
5798 the parent. However, if the lookup was caching a connection to a database,
5799 the connection was closed, and the second delivery process was likely to
5800 see errors such as "PGSQL: query failed: server closed the connection
5801 unexpectedly". The problem has been fixed by closing all cached lookups
5802 in a queue runner before running a delivery process.
5804 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5805 seem to have the "const" qualifier which it has on other OS. I've
5808 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5809 *of the same type* (client or server) with the same public name that an
5810 error should be diagnosed.
5812 9. When Exim looked up a host name for an IP address, but failed to find the
5813 original IP address when looking up the host name (a safety check), it
5814 output the message "<ip address> does not match any IP for NULL", which was
5815 confusing, to say the least. The bug was that the host name should have
5816 appeared instead of "NULL".
5818 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5819 user that is built into the binary, and the -C or -D options is used, root
5820 privilege is dropped before the configuration file is read. In addition,
5821 logging is switched to stderr instead of the normal log files. If the
5822 configuration then re-defines the Exim user, the unprivileged environment
5823 is probably not what is expected, so Exim logs a panic warning message (but
5826 However, if deliver_drop_privilege is set, the unprivileged state may well
5827 be exactly what is intended, so the warning has been cut out in that case,
5828 and Exim is allowed to try to write to its normal log files.
5834 1. smtp_return_error_details was not giving details for temporary sender
5835 or receiver verification errors.
5837 2. Diagnose a configuration error if two authenticators have the same public
5840 3. Exim used not to create the message log file for a message until the first
5841 delivery attempt. This could be confusing when incoming messages were held
5842 for policy or load reasons. The message log file is now created at the time
5843 the message is received, and an initial "Received" line is written to it.
5845 4. The automatically generated man page for command line options had a minor
5846 bug that caused no ill effects; however, a more serious problem was that
5847 the procedure for building the man page automatically didn't always
5848 operate. Consequently, release 4.20 contains an out-of-date version. This
5849 shouldn't happen again.
5851 5. When building Exim with embedded Perl support, the script that builds the
5852 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5853 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5855 6. The freeze_tell option was not being used for messages that were frozen on
5856 arrival, either by an ACL or by local_scan().
5858 7. Added the smtp_incomplete_transaction log selector.
5860 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5861 was accepting AUTH without a new EHLO.
5863 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5864 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5867 10. Logging of TCP/IP connections (when configured) now happens in the main
5868 daemon process instead of the child process, so that the TCP/IP connection
5869 count is more accurate (but it can never be perfect).
5871 11. The use of "drop" in a nested ACL was not being handled correctly in the
5872 outer ACL. Now, if condition failure induced by the nested "drop" causes
5873 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5874 or "require"), the connection is dropped.
5876 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5877 that yield "discard" can now be used with an "accept" or a "discard" verb,
5878 but an error is generated for any others (because I can't see a useful way
5879 to define what should happen).
5881 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5882 are now processed in the same way as lines in the Exim configuration file.
5883 In particular, continuation lines are supported.
5885 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5887 15. Added -ti meaning -t -i.
5889 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5890 domains, and warn by logging if others are found.
5892 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5893 returned from GIFCONF when Exim is trying to find the list of interfaces on
5894 a host. The code in os.c has been modified to copy each ifreq to an aligned
5895 structure in all cases.
5897 Also, in some cases, the returned ifreq's were being copied to a 'struct
5898 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5899 means the last couple of bytes of an IPv6 address could be chopped if the
5900 ifreq contained only a normal sockaddr (14 bytes storage).
5902 18. Named domain lists were not supported in the hosts_treat_as_local option.
5903 An entry such as +xxxx was not recognized, and was treated as a literal
5906 19. Ensure that header lines added by a DATA ACL are included in the reject log
5907 if the ACL subsequently rejects the message.
5909 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5910 MD5 (which is deprecated).
5912 21. When testing a filter file using -bf, Exim was writing a message when it
5913 took the sender from a "From " line in the message, but it was not doing so
5914 when it took $return_path from a Return-Path: header line. It now does.
5916 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5917 with a valid header line field name (a series of printing characters
5918 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5920 23. Changed "disc" in the source to "disk" to conform to the documentation and
5921 the book and for uniformity.
5923 24. Ignore Sendmail's -Ooption=value command line item.
5925 25. When execve() failed while trying to run a command in a pipe transport,
5926 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5927 could be confused with a return value of 69 from the command itself. This
5928 has been changed to 127, the value the shell returns if it is asked to run
5929 a non-existent command. The wording for the related log line suggests a
5930 non-existent command as the problem.
5932 26. If received_header_text expands to an empty string, do not add a Received:
5933 header line to the message. (Well, it adds a token one on the spool, but
5934 marks it "old" so that it doesn't get used or transmitted.)
5936 27. Installed eximstats 1.28 (addition of -nt option).
5938 28. There was no check for failure on the call to getsockname() in the daemon
5939 code. This can fail if there is a shortage of resources on the system, with
5940 ENOMEM, for example. A temporary error is now given on failure.
5942 29. Contrary to the C standard, it seems that in some environments, the
5943 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5944 program. Exim now does this explicitly; it affects the formatting of
5945 timestamps using strftime().
5947 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5948 complaints. I've now initialized all the variables, to avoid this.
5950 32. Header lines added by a system filter were not being "seen" during
5951 transport-time rewrites.
5953 33. The info_callback() function passed to OpenSSL is set up with type void
5954 (*)(SSL *, int, int), as described somewhere. However, when calling the
5955 function (actually a macro) that sets it up, the type void(*)() is
5956 expected. I've put in a cast to prevent warnings from picky compilers.
5958 34. If a DNS black list lookup found a CNAME record, but there were no A
5959 records associated with the domain it pointed at, Exim crashed.
5961 35. If a DNS black list lookup returned more than one A record, Exim ignored
5962 all but the first. It now scans all returned addresses if a particular IP
5963 value is being sought. In this situation, the contents of the
5964 $dnslist_value variable are a list of all the addresses, separated by a
5967 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5968 to accept a host name and all its aliases if the forward lookup for any of
5969 them yielded the IP address of the incoming connection. Now it accepts only
5970 those names whose forward lookup yields the correct IP address. Any other
5971 names are discarded. This closes a loophole whereby a rogue DNS
5972 administrator could create reverse DNS records to break through a
5973 wildcarded host restriction in an ACL.
5975 37. If a user filter or a system filter that ran in a subprocess used any of
5976 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5977 the wrong values were passed to the pipe command ($thisaddress had the
5978 value of $0, $0 had the value of $1, etc). This bug was introduced by
5979 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5981 38. Improved the line breaking for long SMTP error messages from ACLs.
5982 Previously, if there was no break point between 40 and 75 characters, Exim
5983 left the rest of the message alone. Two changes have been made: (a) I've
5984 reduced the minimum length to 35 characters; (b) if it can't find a break
5985 point between 35 and 75 characters, it looks ahead and uses the first one
5986 that it finds. This may give the occasional overlong line, but at least the
5987 remaining text gets split now.
5989 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5990 file descriptors might be low, and that setting 1000 would always raise it.
5991 It turns out that in some environments, the limit is already over 1000 and
5992 that lowering it causes trouble. So now Exim takes care not to decrease it.
5994 40. When delivering a message, the value of $return_path is set to $sender_
5995 address at the start of routing (routers may change the value). By an
5996 oversight, this default was not being set up when an address was tested by
5997 -bt or -bv, which affected the outcome if any router or filter referred to
6000 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6001 log only when "message" or "log_message" are set. However, if one of the
6002 conditions was an address verification, or a call to a nested ACL, the
6003 messages generated by the underlying test were being passed through. This
6004 no longer happens. The underlying message is available in $acl_verify_
6005 message for both "message" and "log_message" expansions, so it can be
6006 passed through if needed.
6008 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6009 new expansion $bh_ to give the encoded byte string without charset
6010 translation. Translation happens only if iconv() is available; HAVE_ICONV
6011 indicates this at build time. HEADERS_CHARSET gives the charset to
6012 translate to; headers_charset can change it in the configuration, and
6013 "headers charset" can change it in an individual filter file.
6015 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6016 that creates RFC 2047 encoded "words" labels them as that charset instead
6017 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6018 expansion operator; (ii) when Exim creates a From: line for a local
6019 message; (iii) when a header line is rewritten to include a "phrase" part.
6021 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6022 buggy, causing it to skip the first lines of messages whose message ID
6023 ended in 'D'. This would not have bitten before Exim release 4.14, because
6024 message IDs were unlikely to end in 'D' before then. The effect was to have
6025 incorrect size information for certain domains.
6027 45. #include "config.h" was missing at the start of the crypt16.c module. This
6028 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6031 46. If there was a timeout during a "random" callout check, Exim treated it as
6032 a failure of the random address, and carried on sending RSET and the real
6033 address. If the delay was just some slowness somewhere, the response to the
6034 original RCPT would be taken as a response to RSET and so on, causing
6035 mayhem of various kinds.
6037 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6038 when I implemented it. It didn't allow for the fact that some option values
6039 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6040 the right test for positive values.
6042 48. Domain names in DNS records are case-independent. Exim always looks them up
6043 in lower case. Some resolvers return domain names in exactly the case they
6044 appear in the zone file, that is, they may contain uppercase letters. Not
6045 all resolvers do this - some return always lower case. Exim was treating a
6046 change of case by a resolver as a change of domain, similar to a widening
6047 of a domain abbreviation. This triggered its re-routing code and so it was
6048 trying to route what was effectively the same domain again. This normally
6049 caused routing to fail (because the router wouldn't handle the domain
6050 twice). Now Exim checks for this case specially, and just changes the
6051 casing of the domain that it ultimately uses when it transmits the message
6054 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6057 50. If a filter generated a file delivery with a non-absolute name (possible if
6058 no home directory exists for the router), the forbid_file option was not
6061 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6062 the existing equality matching.
6064 52. Exim was using ints instead of ino_t variables in some places where it was
6065 dealing with inode numbers.
6067 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6068 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6069 called TMPDIR, and if it finds it is different, it changes its value.
6071 54. The smtp_printf() function is now made available to local_scan() so
6072 additional output lines can be written before returning. There is also an
6073 smtp_fflush() function to enable the detection of a dropped connection.
6074 The variables smtp_input and smtp_batched_input are exported to
6077 55. Changed the default runtime configuration: the message "Unknown user"
6078 has been removed from the ACL, and instead placed on the localuser router,
6079 using the cannot_route_message feature. This means that any verification
6080 failures that generate their own messages won't get overridden. Similarly,
6081 the "Unrouteable address" message that was in the ACL for unverifiable
6082 relay addresses has also been removed.
6084 56. Added hosts_avoid_esmtp to the smtp transport.
6086 57. The exicyclog script was not checking for the esoteric option
6087 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6088 will work only if exicyclog is run under the appropriate euid.
6090 58. Following a discussion on the list, the rules by which Exim recognises line
6091 endings on incoming messages have been changed. The -dropcr and drop_cr
6092 options are now no-ops, retained only for backwards compatibility. The
6093 following line terminators are recognized: LF CRLF CR. However, special
6094 processing applies to CR:
6096 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6097 nor a local message in the state where . is a terminator.
6099 (ii) If a bare CR is encountered in a header line, an extra space is added
6100 after the line terminator so as not to end the header. The reasoning
6101 behind this is that bare CRs in header lines are most likely either
6102 to be mistakes, or people trying to play silly games.
6104 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6105 was being incorrectly given as 18 bytes larger than it should have been.
6106 This is a VOB (very old bug).
6108 60. This may never have affected anything current, but just in case it has:
6109 When the local host is found other than at the start of a list of hosts,
6110 the local host, those with the same MX, and any that follow, are discarded.
6111 When the list in question was part of a longer list of hosts, the following
6112 hosts (not currently being processed) were also being discarded. This no
6113 longer happens. I'm not sure if this situation could ever has previously
6116 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6119 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6120 header. This is something that is recommended in a new Internet Draft, and
6121 is something that is documented as being done by Sendmail. There are two
6122 possible values. For messages generated by the autoreply transport, Exim
6125 Auto-Submitted: auto-replied
6127 whereas for all other generated messages (e.g. bounces) it adds
6129 Auto-Submitted: auto-generated
6131 63. The "personal" condition in filters now includes a test for the
6132 Auto-Submitted: header. If it contains the string "auto-" the message it
6133 not considered personal.
6135 64. Added rcpt_include_affixes as a generic transport option.
6137 65. Added queue_only_override (default true).
6139 66. Added the syslog_duplication option.
6141 67. If what should have been the first header line of a message consisted of
6142 a space followed by a colon, Exim was mis-interpreting it as a header line.
6143 It isn't of course - it is syntactically invalid and should therefore be
6144 treated as the start of the message body. The misbehaviour could have
6145 caused a number of strange effects, including loss of data in subsequent
6146 header lines, and spool format errors.
6148 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6149 client host had authenticated. This control can now be exercised by an ACL
6150 for more flexibility.
6152 69. By default, callouts do not happen when testing with -bh. There is now a
6153 variant, -bhc, which does actually run the callout code, including
6154 consulting and updating the callout cache.
6156 70. Added support for saslauthd authentication, courtesy of Alexander
6159 71. If statvfs() failed on the spool or log directories while checking their
6160 size for availability, Exim confusingly gave the error "space shortage".
6161 Furthermore, in debugging mode it crashed with a floating point exception.
6162 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6163 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6164 serious problem, Exim now writes to the main and panic logs when this
6165 happens, with details of the failure. It then refuses to accept the
6166 incoming message, giving the message "spool directory problem" or "log
6167 directory problem" with a 421 code for SMTP messages.
6169 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6170 0, 1, and 2 exist, because some OS complain for execs without them (see
6171 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6172 descriptors. However, the code omitted to check that the open succeeded,
6173 causing mysterious errors if for some reason the permissions on /dev/null
6174 got screwed. Now Exim writes a message to the main and panic logs, and
6175 bombs out if it can't open /dev/null.
6177 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6178 interact so that it is all more flexible. It is supposed to remain
6179 backwards compatible. Also added extra_local_interfaces.
6181 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6182 to bomb out with an assertion failure - to the client this appears as a
6183 connection drop. This problem occurs in the part of the code that was taken
6184 from the Samba project. Fortunately, the assertion is in a very simple
6185 function, so I have fixed this by reproducing the function inline in the
6186 one place where it is called, and arranging for authentication to fail
6187 instead of killing the process with assert().
6189 75. The SPA client code was not working when the server requested OEM rather
6190 than Unicode encoding.
6192 76. Added code to make require_files with a specific uid setting more usable in
6193 the case where statting the file as root fails - usually a non-root-mounted
6194 NFS file system. When this happens and the failure is EACCES, Exim now
6195 forks a subprocess and does the per-uid checking as the relevant uid.
6197 77. Added process_log_path.
6199 78. If log_file_path was not explicitly set, a setting of check_log_space or
6200 check_log_inodes was ignored.
6202 79. If a space check for the spool or log partitions fails, the incident is now
6203 logged. Of course, in the latter case the data may get lost...
6205 80. Added the %p formatting code to string_format() so that it can be used to
6206 print addresses in debug_print(). Adjusted all the address printing in the
6207 debugging in store.c to use %p rather than %d.
6209 81. There was a concern that a line of code in smtp_in.c could overflow a
6210 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6211 initially expressed, the concern was not well-founded, because trailing
6212 spaces are removed early. However, if the trailing spaces were followed by
6213 a NULL, they did not get removed, so the overflow was possible. Two fixes
6216 (a) I re-wrote the offending code in a cleaner fashion.
6217 (b) If an incoming SMTP command contains a NULL character, it is rejected
6220 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6221 runs initgroups(), so that if the Exim user is in any additional groups,
6222 they will be used during message reception.
6228 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob