1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
54 TL/01 When running the test suite, the README says that variables such as
55 no_msglog_check are global and can be placed anywhere in a specific
56 test's script, however it was observed that placement needed to be near
57 the beginning for it to behave that way. Changed the runtest perl
58 script to read through the entire script once to detect and set these
59 variables, reset to the beginning of the script, and then run through
60 the script parsing/test process like normal.
62 TL/02 The BSD's have an arc4random API. One of the functions to induce
63 adding randomness was arc4random_stir(), but it has been removed in
64 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
65 function when detected.
67 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
68 cause callback expansion.
70 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
71 syntax errors in an expansion can be treated as a string instead of
72 logging or causing an error, due to the internal use of bool_lax
73 instead of bool when processing it.
75 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
76 server certificates when making smtp deliveries.
78 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
80 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
82 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
84 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
85 Merged patch from Sebastian Wiedenroth.
87 JH/05 Fix results-pipe from transport process. Several recipients, combined
88 with certificate use, exposed issues where response data items split
89 over buffer boundaries were not parsed properly. This eventually
90 resulted in duplicates being sent. This issue only became common enough
91 to notice due to the introduction of conection certificate information,
92 the item size being so much larger. Found and fixed by Wolfgang Breyha.
94 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
95 size buffer was used, resulting in syntax errors when an expansion
98 JH/07 Add support for directories of certificates when compiled with a GnuTLS
99 version 3.3.6 or later.
101 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
102 is EXPERIMENTAL_EVENT, the main-configuration and transport options
103 both become "event_action", the variables become $event_name, $event_data
104 and $event_defer_errno. There is a new variable $verify_mode, usable in
105 routers, transports and related events. The tls:cert event is now also
106 raised for inbound connections, if the main configuration event_action
109 TL/06 In test suite, disable OCSP for old versions of openssl which contained
110 early OCSP support, but no stapling (appears to be less than 1.0.0).
112 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
113 server certificate names available under the smtp transport option
114 "tls_verify_cert_hostname" now do not permit multi-component wildcard
117 JH/10 Time-related extraction expansions from certificates now use the main
118 option "timezone" setting for output formatting, and are consistent
119 between OpenSSL and GnuTLS compilations. Bug 1541.
121 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
122 encoded parameter in the incoming message. Bug 1558.
124 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
125 include certificate info, eximon was claiming there were spoolfile
128 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
130 JH/14 Log delivery-related information more consistently, using the sequence
131 "H=<name> [<ip>]" wherever possible.
133 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
134 are problematic for Debian distribution, omit them from the release
137 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
139 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
142 JH/17 Fix a null-indirection in certextract expansions when a nondefault
143 output list separator was used.
148 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
149 checkers that were complaining about end of non-void function with no
152 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
153 This was a regression intruduced in 4.83 by another bugfix.
155 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
157 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
158 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
164 TF/01 Correctly close the server side of TLS when forking for delivery.
166 When a message was received over SMTP with TLS, Exim failed to clear up
167 the incoming connection properly after forking off the child process to
168 deliver the message. In some situations the subsequent outgoing
169 delivery connection happened to have the same fd number as the incoming
170 connection previously had. Exim would try to use TLS and fail, logging
171 a "Bad file descriptor" error.
173 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
174 utilities have not been installed.
176 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
177 temporary space as the ACL may create new global variables.
179 TL/01 LDAP support uses per connection or global context settings, depending
180 upon the detected version of the libraries at build time.
182 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
183 to extract and use the src ip:port in logging and expansions as if it
184 were a direct connection from the outside internet. PPv2 support was
185 updated based on HAProxy spec change in May 2014.
187 JH/02 Add ${listextract {number}{list}{success}{fail}}.
189 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
190 Properly escape header and check for NULL return.
192 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
195 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
197 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
198 characters in header names, implemented as a verify condition.
199 Contributed by Michael Fischer v. Mollard.
201 TL/05 Rename SPF condition results err_perm and err_temp to standardized
202 results permerror and temperror. Previous values are deprecated but
203 still accepted. In a future release, err_perm and err_temp will be
204 completely removed, which will be a backward incompatibility if the
205 ACL tests for either of these two old results. Patch contributed by
206 user bes-internal on the mailing list.
208 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
210 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
211 selectors, in both main and reject logs.
213 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
216 JH/07 Add malware type "sock" for talking to simple daemon.
218 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
220 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
221 routers/transports under cutthrough routing.
223 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
224 numbers. Touch up "bool" conditional to keep the same definition.
226 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
228 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
230 JH/12 Expand items in router/transport headers_add or headers_remove lists
231 individually rather than the list as a whole. Bug 1452.
233 Required for reasonable handling of multiple headers_ options when
234 they may be empty; requires that headers_remove items with embedded
235 colons must have them doubled (or the list-separator changed).
237 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
238 view the policy declared in the DMARC record. Currently, $dmarc_status
239 is a combined value of both the record presence and the result of the
242 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
244 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
245 dnslookup router and the smtp transport (applying to the forward
248 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
249 of ldap servers used for a specific lookup. Patch provided by Heiko
252 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
253 New variable $lookup_dnssec_authenticated for observability.
255 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
256 Patch submitted by Lars Timman.
258 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
260 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
261 Requires trusted mode and valid format message id, aborts otherwise.
262 Patch contributed by Heiko Schlichting.
264 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
265 certextract with support for various fields. Bug 1358.
267 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
268 is requested by default, modifiable by smtp transport option
271 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
272 operate on certificate variables to give certificate fingerprints
273 Also new ${sha256:cert_variable}.
275 JH/23 The PRDR feature is moved from being Experimental into the mainline.
277 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
278 Christian Aistleitner.
280 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
282 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
283 file. Patch from Wolfgang Breyha.
285 JH/25 Expand the coverage of the delivery $host and $host_address to
286 client authenticators run in verify callout. Bug 1476.
288 JH/26 Port service names are now accepted for tls_on_connect_ports, to
289 align with daemon_smtp_ports. Bug 72.
291 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
292 support and error reporting did not work properly.
294 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
295 and is readable. Patch from Andrew Colin Kissa.
297 TL/14 Enhance documentation of ${run expansion and how it parses the
298 commandline after expansion, particularly in the case when an
299 unquoted variable expansion results in an empty value.
301 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
303 PP/02 Fix internal collision of T_APL on systems which support RFC3123
304 by renaming away from it. Addresses GH issue 15, reported by
307 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
309 TL/15 SECURITY: prevent double expansion in math comparison functions
310 (can expand unsanitized data). Not remotely exploitable.
317 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
319 PP/02 Make -n do something, by making it not do something.
320 When combined with -bP, the name of an option is not output.
322 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
325 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
326 $sender_host_name and config options to manage this, and basic check
329 PP/05 DSCP support for outbound connections and control modifier for inbound.
331 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
332 (Only plugin which currently uses this is kerberos4, which nobody should
333 be using, but we should make it available and other future plugins might
334 conceivably use it, even though it would break NAT; stuff *should* be
335 using channel bindings instead).
337 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
338 name; added for Sendmail compatibility; requires admin caller.
339 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
340 just ignore it); requires trusted caller.
341 Also parse but ignore: -Ac -Am -X<logfile>
344 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
346 TL/02 Add +smtp_confirmation as a default logging option.
348 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
349 Patch by Magnus Holmgren from 2007-02-20.
351 TL/04 Bugzilla 1281 - Spec typo.
352 Bugzilla 1283 - Spec typo.
353 Bugzilla 1290 - Spec grammar fixes.
355 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
357 TL/06 Add Experimental DMARC support using libopendmarc libraries.
359 TL/07 Fix an out of order global option causing a segfault. Reported to dev
360 mailing list by by Dmitry Isaikin.
362 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
364 JH/02 Support "G" suffix to numbers in ${if comparisons.
366 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
368 NM/01 Bugzilla 1197 - Spec typo
369 Bugzilla 1196 - Spec examples corrections
371 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
373 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
374 gnutls_enable_pkcs11, but renamed to more accurately indicate its
377 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
378 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
380 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
381 "acl {{name}{arg}...}", and optional args on acl condition
384 JH/05 Permit multiple router/transport headers_add/remove lines.
386 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
388 JH/07 Avoid using a waiting database for a single-message-only transport.
389 Performance patch from Paul Fisher. Bugzilla 1262.
391 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
394 JH/09 Add $headers_added variable, with content from use of ACL modifier
395 add_header (but not yet added to the message). Bugzilla 199.
397 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
398 Pulled from Bugzilla 817 by Wolfgang Breyha.
400 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
402 (nb: this is the same fix as in Exim 4.80.1)
404 JH/11 Add A= logging on delivery lines, and a client_set_id option on
407 JH/12 Add optional authenticated_sender logging to A= and a log_selector
410 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
412 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
413 advertise SMTP AUTH mechanism to us, instead of a generic
414 protocol violation error. Also, make Exim more robust to bad
415 data from the Dovecot auth socket.
417 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
419 When a queue runner is handling a message, Exim first routes the
420 recipient addresses, during which it prunes them based on the retry
421 hints database. After that it attempts to deliver the message to
422 any remaining recipients. It then updates the hints database using
425 So if a recipient address works intermittently, it can get repeatedly
426 deferred at routing time. The retry hints record remains fresh so the
427 address never reaches the final cutoff time.
429 This is a fairly common occurrence when a user is bumping up against
430 their storage quota. Exim had some logic in its local delivery code
431 to deal with this. However it did not apply to per-recipient defers
432 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
434 This change adds a proper retry rule check during routing so that the
435 final cutoff time is checked against the message's age. We only do
436 this check if there is an address retry record and there is not a
437 domain retry record; this implies that previous attempts to handle
438 the address had the retry_use_local_parts option turned on. We use
439 this as an approximation for the destination being like a local
440 delivery, as in LMTP.
442 I suspect this new check makes the old local delivery cutoff check
443 redundant, but I have not verified this so I left the code in place.
445 TF/02 Correct gecos expansion when From: is a prefix of the username.
447 Test 0254 submits a message to Exim with the header
451 When I ran the test suite under the user fanf2, Exim expanded
452 the header to contain my full name, whereas it should have added
453 a Resent-Sender: header. It erroneously treats any prefix of the
454 username as equal to the username.
456 This change corrects that bug.
458 GF/01 DCC debug and logging tidyup
459 Error conditions log to paniclog rather than rejectlog.
460 Debug lines prefixed by "DCC: " to remove any ambiguity.
462 TF/03 Avoid unnecessary rebuilds of lookup-related code.
464 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
465 Bug spotted by Jeremy Harris; was flawed since initial commit.
466 Would have resulted in OCSP responses post-SNI triggering an Exim
467 NULL dereference and crash.
469 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
471 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
472 Bug detection, analysis and fix by Samuel Thibault.
473 Bugzilla 1331, Debian bug #698092.
475 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
477 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
478 Server implementation by Todd Lyons, client by JH.
479 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
480 config variable "prdr_enable" controls whether the server
481 advertises the facility. If the client requests PRDR a new
482 acl_data_smtp_prdr ACL is called once for each recipient, after
483 the body content is received and before the acl_smtp_data ACL.
484 The client is controlled by bolth of: a hosts_try_prdr option
485 on the smtp transport, and the server advertisement.
486 Default client logging of deliveries and rejections involving
487 PRDR are flagged with the string "PRDR".
489 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
490 fclose(). Diagnosis by Todd Lyons.
492 PP/17 Update configure.default to handle IPv6 localhost better.
493 Patch by Alain Williams (plus minor tweaks).
496 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
497 This is now consistent with GnuTLS, and is now documented: the
498 previous undocumented portable approach to treating the option as
499 unset was to force an expansion failure. That still works, and
500 an empty string is now equivalent.
502 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
503 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
504 not performing validation itself.
506 PP/20 Added force_command boolean option to pipe transport.
507 Patch from Nick Koston, of cPanel Inc.
509 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
512 TF/04 Added udpsend ACL modifer and hexquote expansion operator
514 PP/21 Fix eximon continuous updating with timestamped log-files.
515 Broken in a format-string cleanup in 4.80, missed when I repaired the
516 other false fix of the same issue.
517 Report and fix from Heiko Schlichting.
520 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
521 Report from Prashanth Katuri.
523 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
524 It's SecureTransport, so affects any MacOS clients which use the
525 system-integrated TLS libraries, including email clients.
527 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
528 using a MIME ACL for non-SMTP local injection.
529 Report and assistance in diagnosis by Warren Baker.
531 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
533 JH/16 Fix comparisons for 64b. Bugzilla 1385.
535 TL/09 Add expansion variable $authenticated_fail_id to keep track of
536 last id that failed so it may be referenced in subsequent ACL's.
538 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
541 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
542 ldap library initialization, allowing self-signed CA's to be
543 used. Also properly sets require_cert option later in code by
544 using NULL (global ldap config) instead of ldap handle (per
545 session). Bug diagnosis and testing by alxgomz.
547 TL/12 Enhanced documentation in the ratelimit.pl script provided in
548 the src/util/ subdirectory.
550 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
551 renamed to Transport Post Delivery Action by Jeremy Harris, as
554 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
555 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
556 redis_servers = needs to be configured which will be used by the redis
557 lookup. Patch from Warren Baker, of The Packet Hub.
559 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
561 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
562 hostname or reverse DNS when processing a host list. Used suggestions
563 from multiple comments on this bug.
565 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
567 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
568 Missed a few lines, added it to make the runtest require no keyboard
571 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
572 contains upper case chars. Make router use caseful_local_part.
574 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
575 support when GnuTLS has been built with p11-kit.
581 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
583 This, or similar/improved, will also be change PP/11 of 4.82.
589 PP/01 Handle short writes when writing local log-files.
590 In practice, only affects FreeBSD (8 onwards).
591 Bugzilla 1053, with thanks to Dmitry Isaikin.
593 NM/01 Bugzilla 949 - Documentation tweak
595 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
598 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
600 PP/02 Implemented gsasl authenticator.
602 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
604 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
605 `pkg-config foo` for cflags/libs.
607 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
608 with rest of GSASL and with heimdal_gssapi.
610 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
611 `pkg-config foo` for cflags/libs for the TLS implementation.
613 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
614 properties get this fed in as external SSF. A number of robustness
615 and debugging improvements to the cyrus_sasl authenticator.
617 PP/08 cyrus_sasl server now expands the server_realm option.
619 PP/09 Bugzilla 1214 - Log authentication information in reject log.
620 Patch by Jeremy Harris.
622 PP/10 Added dbmjz lookup type.
624 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
626 PP/12 MAIL args handles TAB as well as SP, for better interop with
627 non-compliant senders.
628 Analysis and variant patch by Todd Lyons.
630 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
631 Bug report from Lars Müller <lars@samba.org> (via SUSE),
632 Patch from Dirk Mueller <dmueller@suse.com>
634 PP/13 tls_peerdn now print-escaped for spool files.
635 Observed some $tls_peerdn in wild which contained \n, which resulted
636 in spool file corruption.
638 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
639 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
640 or write after TLS renegotiation, which otherwise led to messages
643 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
644 as a tracking header (ie: a signed header comes before the signature).
645 Patch from Wolfgang Breyha.
647 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
648 comma-sep list; embedded commas doubled.
650 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
652 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
654 Report and patch from Dmitry Banschikov.
656 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
657 Removed SSL_clear() after SSL_new() which led to protocol negotiation
658 failures. We appear to now support TLS1.1+ with Exim.
660 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
661 lets Exim select keys and certificates based upon TLS SNI from client.
662 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
663 before an outbound SMTP session. New log_selector, +tls_sni.
665 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
666 NULL dereference. Report and patch from Alun Jones.
668 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
669 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
670 Not seeing resolver debug output on NetBSD, but suspect this is a
671 resolver implementation change.
673 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
674 Left warnings. Added "eximon gdb" invocation mode.
676 PP/21 Defaulting "accept_8bitmime" to true, not false.
678 PP/22 Added -bw for inetd wait mode support.
680 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
681 locate the relevant includes and libraries. Made this the default.
683 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
684 Bugzilla 1246, report and most of solution from Tomasz Kusy.
686 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
687 This may cause build issues on older platforms.
689 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
690 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
691 gnutls_require_mac & gnutls_require_protocols (no longer supported).
692 Added SNI support via GnuTLS too.
693 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
695 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
697 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
698 Applied second patch from Janne, implementing suggestion to default
699 multiple-strings-in-record handling to match SPF spec.
701 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
703 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
704 read-only, out of scope).
705 Patch from Wolfgang Breyha, report from Stuart Northfield.
707 PP/29 Fix three issues highlighted by clang analyser static analysis.
708 Only crash-plausible issue would require the Cambridge-specific
709 iplookup router and a misconfiguration.
710 Report from Marcin Mirosław.
712 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
714 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
715 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
716 As part of this, removing so much warning spew let me fix some minor
717 real issues in debug logging.
719 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
720 assignment on my part. Fixed.
722 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
723 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
724 Janne Snabb (who went above and beyond: thank you).
726 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
727 string otherwise requires a connection and a bunch more work and it's
728 relatively easy to get wrong. Should also expose TLS library linkage
731 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
732 64-bit ${eval} (JH/03).
734 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
735 GNU libc to support some of the 64-bit stuff, should not lead to
736 conflicts. Defined before os.h is pulled in, so if a given platform
737 needs to override this, it can.
739 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
740 protection layer was required, which is not implemented.
741 Bugzilla 1254, patch from Wolfgang Breyha.
743 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
744 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
745 tls_dhparam take prime identifiers. Also unbreak combination of
746 OpenSSL+DH_params+TLSSNI.
748 PP/39 Disable SSLv2 by default in OpenSSL support.
754 PP/01 Solaris build fix for Oracle's LDAP libraries.
755 Bugzilla 1109, patch from Stephen Usher.
757 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
759 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
762 TF/02 Fix a couple more cases where we did not log the error message
763 when unlink() failed. See also change 4.74-TF/03.
765 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
766 lock up or crash if it happened to be inside a call to libc when it
767 got a SIGUSR1 from exiwhat.
769 The SIGUSR1 handler appends the current process status to the process
770 log which is later printed by exiwhat. It used to use the general
771 purpose logging code to do this, but several functions it calls are
772 not safe for signals.
774 The new output code in the SIGUSR1 handler is specific to the process
775 log, and simple enough that it's easy to inspect for signal safety.
776 Removing some special cases also simplifies the general logging code.
777 Removing the spurious timestamps from the process log simplifies
780 TF/04 Improved ratelimit ACL condition.
782 The /noupdate option has been deprecated in favour of /readonly which
783 has clearer semantics. The /leaky, /strict, and /readonly update modes
784 are mutually exclusive. The update mode is no longer included in the
785 database key; it just determines when the database is updated. (This
786 means that when you upgrde Exim will forget old rate measurements.)
788 Exim now checks that the per_* options are used with an update mode that
789 makes sense for the current ACL. For example, when Exim is processing a
790 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
791 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
792 must specify per_mail/readonly. If you omit the update mode it defaults to
793 /leaky where that makes sense (as before) or /readonly where required.
795 The /noupdate option is now undocumented but still supported for
796 backwards compatibility. It is equivalent to /readonly except that in
797 ACLs where /readonly is required you may specify /leaky/noupdate or
798 /strict/noupdate which are treated the same as /readonly.
800 A useful new feature is the /count= option. This is a generalization
801 of the per_byte option, so that you can measure the throughput of other
802 aggregate values. For example, the per_byte option is now equivalent
803 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
805 The per_rcpt option has been generalized using the /count= mechanism
806 (though it's more complicated than the per_byte equivalence). When it is
807 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
808 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
809 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
810 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
811 Note that using per_rcpt with a non-readonly update mode in more than
812 one ACL will cause the recipients to be double-counted. (The per_mail
813 and per_byte options don't have this problem.)
815 The handling of very low rates has changed slightly. If the computed rate
816 is less than the event's count (usually one) then this event is the first
817 after a long gap. In this case the rate is set to the same as this event's
818 count, so that the first message of a spam run is counted properly.
820 The major new feature is a mechanism for counting the rate of unique
821 events. The new per_addr option counts the number of different
822 recipients that someone has sent messages to in the last time period. It
823 behaves like per_rcpt if all the recipient addresses are different, but
824 duplicate recipient addresses do not increase the measured rate. Like
825 the /count= option this is a general mechanism, so the per_addr option
826 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
827 example, measure the rate that a client uses different sender addresses
828 with the options per_mail/unique=$sender_address. There are further
829 details in the main documentation.
831 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
833 TF/06 Removed a few PCRE remnants.
835 TF/07 Automatically extract Exim's version number from tags in the git
836 repository when doing development or release builds.
838 PP/02 Raise smtp_cmd_buffer_size to 16kB.
839 Bugzilla 879. Patch from Paul Fisher.
841 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
842 Heavily based on revision 40f9a89a from Simon Arlott's tree.
845 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
847 PP/05 Variable $av_failed, true if the AV scanner deferred.
848 Bugzilla 1078. Patch from John Horne.
850 PP/06 Stop make process more reliably on build failure.
851 Bugzilla 1087. Patch from Heiko Schlittermann.
853 PP/07 Make maildir_use_size_file an _expandable_ boolean.
854 Bugzilla 1089. Patch from Heiko Schlittermann.
856 PP/08 Handle ${run} returning more data than OS pipe buffer size.
857 Bugzilla 1131. Patch from Holger Weiß.
859 PP/09 Handle IPv6 addresses with SPF.
860 Bugzilla 860. Patch from Wolfgang Breyha.
862 PP/10 GnuTLS: support TLS 1.2 & 1.1.
864 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
867 PP/11 match_* no longer expand right-hand-side by default.
868 New compile-time build option, EXPAND_LISTMATCH_RHS.
869 New expansion conditions, "inlist", "inlisti".
871 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
873 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
875 PP/14 fix log_write() format string regression from TF/03.
876 Bugzilla 1152. Patch from Dmitry Isaikin.
882 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
884 PP/02 Harmonised TLS library version reporting; only show if debugging.
885 Layout now matches that introduced for other libraries in 4.74 PP/03.
887 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
889 PP/04 New "dns_use_edns0" global option.
891 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
894 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
895 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
897 TK/01 Updated PolarSSL code to 0.14.2.
898 Bugzilla 1097. Patch from Andreas Metzler.
900 PP/07 Catch divide-by-zero in ${eval:...}.
903 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
906 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
907 format-string attack -- SECURITY: remote arbitrary code execution.
909 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
910 time unintentionally subject to list matching rules, letting the header
911 cause arbitrary Exim lookups (of items which can occur in lists, *not*
912 arbitrary string expansion). This allowed for information disclosure.
914 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
915 INT_MIN/-1 -- value coerced to INT_MAX.
921 NM/01 Workround for PCRE version dependency in version reporting
924 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
925 This fixes portability to compilers other than gcc, notably
926 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
928 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
929 makefiles for portability to HP-UX and POSIX correctness.
931 PP/01 Permit LOOKUP_foo enabling on the make command-line.
932 Also via indented variable definition in the Makefile.
933 (Debugging by Oliver Heesakkers).
935 PP/02 Restore caching of spamd results with expanded spamd_address.
936 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
938 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
939 Improves build reliability. Fix from: Frank Elsner
941 NM/02 Fix wide character breakage in the rfc2047 coding
942 Fixes bug 1064. Patch from Andrey N. Oktyabrski
944 NM/03 Allow underscore in dnslist lookups
945 Fixes bug 1026. Patch from Graeme Fowler
947 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
948 Code patches from Adam Ciarcinski of NetBSD.
950 NM/04 Fixed exiqgrep to cope with mailq missing size issue
953 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
954 is logged, to avoid truncation. Patch from John Horne.
956 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
957 Patch from Jakob Hirsch.
959 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
960 SQL string expansion failure details.
961 Patch from Andrey Oktyabrski.
963 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
964 Patch from Simon Arlott.
966 PP/09 New lookups functionality failed to compile on old gcc which rejects
967 extern declarations in function scope.
968 Patch from Oliver Fleischmann
970 PP/10 Use sig_atomic_t for flags set from signal handlers.
971 Check getgroups() return and improve debugging.
972 Fixed developed for diagnosis in bug 927 (which turned out to be
975 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
976 Patch from Mark Zealey.
978 PP/12 Bugzilla 1056: Improved spamd server selection.
979 Patch from Mark Zealey.
981 PP/13 Bugzilla 1086: Deal with maildir quota file races.
982 Based on patch from Heiko Schlittermann.
984 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
985 Patch from Uwe Doering, sign-off by Michael Haardt.
987 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
988 variable declaration deep within a block. Bug and patch from
991 PP/15 lookups-Makefile IRIX compatibilty coercion.
993 PP/16 Make DISABLE_DKIM build knob functional.
995 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
996 Patch by Simon Arlott
998 TF/03 Fix valgrind.h portability to C89 compilers that do not support
999 variable argument macros. Our copy now differs from upstream.
1005 TF/01 Failure to get a lock on a hints database can have serious
1006 consequences so log it to the panic log.
1008 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1009 controlled using the smtp_confirmation log selector.
1011 TF/03 Include the error message when we fail to unlink a spool file.
1013 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1014 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1015 for maintaining out-of-tree patches for some time.
1017 PP/01 Bugzilla 139: Documentation and portability issues.
1018 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1019 Handle per-OS dynamic-module compilation flags.
1021 PP/02 Let /dev/null have normal permissions.
1022 The 4.73 fixes were a little too stringent and complained about the
1023 permissions on /dev/null. Exempt it from some checks.
1024 Reported by Andreas M. Kirchwitz.
1026 PP/03 Report version information for many libraries, including
1027 Exim version information for dynamically loaded libraries. Created
1028 version.h, now support a version extension string for distributors
1029 who patch heavily. Dynamic module ABI change.
1031 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1032 privilege escalation vulnerability whereby the Exim run-time user
1033 can cause root to append content of the attacker's choosing to
1036 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1039 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1040 If dropping privileges for untrusted macros, we disabled normal logging
1041 on the basis that it would fail; for the Exim run-time user, this is not
1042 the case, and it resulted in successful deliveries going unlogged.
1043 Fixed. Reported by Andreas Metzler.
1049 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1050 only prepend for the Resent-* case. Fixes regression introduced in
1051 Exim 4.70 by NM/22 for Bugzilla 607.
1053 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1054 increasing numbers of administrators be bitten by this.
1056 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1058 PP/03 Bugzilla 994: added openssl_options main configuration option.
1060 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1062 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1064 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1066 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1067 without a peer certificate, leading to a segfault because of an
1068 assumption that peers always have certificates. Be a little more
1069 paranoid. Problem reported by Martin Tscholak.
1071 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1072 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1073 NB: ClamAV planning to remove STREAM in "middle of 2010".
1074 CL also introduces -bmalware, various -d+acl logging additions and
1075 more caution in buffer sizes.
1077 PP/09 Implemented reverse_ip expansion operator.
1079 PP/10 Bugzilla 937: provide a "debug" ACL control.
1081 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1083 PP/12 Bugzilla 973: Implement --version.
1085 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1087 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1089 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1091 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1092 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1093 ignore trailing whitespace.
1095 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1097 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1100 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1101 Notification from Dr Andrew Aitchison.
1103 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1104 ExtendedDetectionInfo response format.
1105 Notification from John Horne.
1107 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1110 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1111 XSL and documented dependency on system catalogs, with examples of how
1114 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1117 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1118 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1119 configuration file which is writeable by the Exim user or group.
1121 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1122 of configuration files to cover files specified with the -C option if
1123 they are going to be used with root privileges, not just the default
1126 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1127 option (effectively making it always true).
1129 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1130 files to be used while preserving root privileges.
1132 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1133 that rogue child processes cannot use them.
1135 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1136 run-time user, instead of root.
1138 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1139 Exim run-time user without dropping privileges.
1141 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1142 result string, instead of calling string_vformat() twice with the same
1145 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1146 for other users. Others should always drop root privileges if they use
1147 -C on the command line, even for a whitelisted configure file.
1149 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1151 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1157 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1158 $data_path, and $header_path variables; fixed documentation bugs and
1161 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1162 exipick to access non-standard spools, including the "frozen" queue
1165 NM/01 Bugzilla 965: Support mysql stored procedures.
1166 Patch from Alain Williams
1168 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1170 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1171 Patch from Andreas Metzler
1173 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1174 Patch from Kirill Miazine
1176 NM/05 Bugzilla 671: Added umask to procmail example.
1178 JJ/03 installed exipick 20100323.0, fixing doc bug
1180 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1181 directory. Notification and patch from Dan Rosenberg.
1183 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1185 TK/02 Improve log output when DKIM signing operation fails.
1187 MH/01 Treat the transport option dkim_domain as a colon separated
1188 list, not as a single string, and sign the message with each element,
1189 omitting multiple occurences of the same signer.
1191 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1192 Bugzilla 985, 986. Patch by Simon Arlott
1194 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1195 Patch by Simon Arlott
1197 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1198 MBX locking. Notification from Dan Rosenberg.
1204 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1206 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1208 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1210 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1212 NM/04 Bugzilla 915: Flag broken perl installation during build.
1218 TK/01 Added patch by Johannes Berg that expands the main option
1219 "spamd_address" if it starts with a dollar sign.
1221 TK/02 Write list of recipients to X-Envelope-Sender header when building
1222 the mbox-format spool file for content scanning (suggested by Jakob
1225 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1226 (http://www.dcc-servers.net/) support via dccifd. Activated by
1227 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1229 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1230 by Mark Daniel Reidel <mr@df.eu>.
1232 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1233 When building exim an external PCRE library is now needed -
1234 PCRE is a system library on the majority of modern systems.
1235 See entry on PCRE_LIBS in EDITME file.
1237 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1238 conversation. Added nologin parameter to request.
1239 Patch contributed by Kirill Miazine.
1241 TF/01 Do not log submission mode rewrites if they do not change the address.
1243 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1245 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1246 log files in place. Contributed by Roberto Lima.
1248 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1250 TF/03 Bugzilla 615: When checking the local_parts router precondition
1251 after a local_part_suffix or local_part_prefix option, Exim now
1252 does not use the address's named list lookup cache, since this
1253 contains cached lookups for the whole local part.
1255 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1256 Robert Millan. Documentation is in experimental-spec.txt.
1258 TF/04 Bugzilla 668: Fix parallel build (make -j).
1260 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1262 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1263 Patch provided by Jan Srzednicki.
1265 TF/05 Leading white space used to be stripped from $spam_report which
1266 wrecked the formatting. Now it is preserved.
1268 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1269 that they are available at delivery time.
1271 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1273 TF/08 TLS error reporting now respects the incoming_interface and
1274 incoming_port log selectors.
1276 TF/09 Produce a more useful error message if an SMTP transport's hosts
1277 setting expands to an empty string.
1279 NM/06 Bugzilla 744: EXPN did not work under TLS.
1280 Patch provided by Phil Pennock.
1282 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1283 Patch provided by Richard Godbee.
1285 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1286 acl_smtp_notquit, added index entry.
1288 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1289 Patch provided by Eugene Bujak.
1291 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1292 accept(). Patch provided by Maxim Dounin.
1294 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1295 Patch provided by Phil Pennock.
1297 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1299 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1300 Patch provided by Brad "anomie" Jorsch.
1302 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1303 Patch provided by Dean Brooks.
1305 TK/05 Add native DKIM support (does not depend on external libraries).
1307 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1308 Patch provided by Graeme Fowler.
1310 NM/16 Bugzilla 851: Documentation example syntax fix.
1312 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1314 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1317 NM/19 Bugzilla 745: TLS version reporting.
1318 Patch provided by Phil Pennock.
1320 NM/20 Bugzilla 167: bool: condition support.
1321 Patch provided by Phil Pennock.
1323 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1324 clients. Patch provided by Phil Pennock.
1326 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1327 Patch provided by Brad "anomie" Jorsch.
1329 NM/23 Bugzilla 687: Fix misparses in eximstats.
1330 Patch provided by Heiko Schlittermann.
1332 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1333 Patch provided by Heiko Schlittermann.
1335 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1336 plus update to original patch.
1338 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1340 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1341 Patch provided by David Brownlee.
1343 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1345 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1347 NM/30 Bugzilla 888: TLS documentation bugfixes.
1349 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1351 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1352 Unlike the original bugzilla I have changed all shell scripts in src tree.
1354 NM/33 Bugzilla 898: Transport filter timeout fix.
1355 Patch by Todd Rinaldo.
1357 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1358 Patch by Serge Demonchaux.
1360 NM/35 Bugzilla 39: Base64 decode bug fixes.
1361 Patch by Jakob Hirsch.
1363 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1365 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1367 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1369 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1375 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1376 ALT-N's libdkim that I have put here:
1377 http://duncanthrax.net/exim-experimental/
1379 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1380 were called 'true' and it seems that C99 defines that as a reserved
1381 keyword to be used with 'bool' variable types. That means you could
1382 not include C99-style headers which use bools without triggering
1383 build errors in sieve.c.
1385 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1386 as mailq or other aliases. Changed the --help handling significantly
1387 to do whats expected. exim_usage() emits usage/help information.
1389 SC/01 Added the -bylocaldomain option to eximstats.
1391 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1393 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1395 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1401 PH/01 Another patch from the Sieve maintainer.
1403 PH/02 When an IPv6 address is converted to a string for single-key lookup
1404 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1405 dots are used instead of colons so that keys in lsearch files need not
1406 contain colons. This was done some time before quoting was made available
1407 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1408 (notated using the quote facility) so as to distinguish them from IPv4
1409 keys. This meant that lookups for IP addresses in host lists did not work
1410 for iplsearch lookups.
1412 This has been fixed by arranging for IPv6 addresses to be expressed with
1413 colons if the lookup type is iplsearch. This is not incompatible, because
1414 previously such lookups could never work.
1416 The situation is now rather anomolous, since one *can* have colons in
1417 ordinary lsearch keys. However, making the change in all cases is
1418 incompatible and would probably break a number of configurations.
1420 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1423 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1424 conversion specification without a maximum field width, thereby enabling
1425 a rogue spamd server to cause a buffer overflow. While nobody in their
1426 right mind would setup Exim to query an untrusted spamd server, an
1427 attacker that gains access to a server running spamd could potentially
1428 exploit this vulnerability to run arbitrary code as the Exim user.
1430 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1431 $primary_hostname instead of what libspf2 thinks the hosts name is.
1433 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1434 a directory entry by the name of the lookup key. Previously, if a
1435 symlink pointed to a non-existing file or a file in a directory that
1436 Exim lacked permissions to read, a lookup for a key matching that
1437 symlink would fail. Now it is enough that a matching directory entry
1438 exists, symlink or not. (Bugzilla 503.)
1440 PH/03 The body_linecount and body_zerocount variables are now exported in the
1443 PH/04 Added the $dnslist_matched variable.
1445 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1446 This means they are set thereafter only if the connection becomes
1449 PH/06 Added the client_condition to authenticators so that some can be skipped
1450 by clients under certain conditions.
1452 PH/07 The error message for a badly-placed control=no_multiline_responses left
1453 "_responses" off the end of the name.
1455 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1457 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1458 (without spaces) instead of just copying the configuration text.
1460 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1462 PH/11 Added $max_received_linelength.
1464 PH/12 Added +ignore_defer and +include_defer to host lists.
1466 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1467 way in which PCRE > 7.0 is built.
1469 PH/14 Implemented queue_only_load_latch.
1471 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1472 MAIL command. The effect was to mangle the value on 64-bit systems.
1474 PH/16 Another patch from the Sieve maintainer.
1476 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1478 PH/18 If a system quota error occurred while trying to create the file for
1479 a maildir delivery, the message "Mailbox is full" was not appended to the
1480 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1481 only to a quota excession during the actual writing of the file.
1483 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1484 characters?) which causes problems in log lines. The DN values are now
1485 passed through string_printing() before being added to log lines.
1487 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1488 and InterBase are left for another time.)
1490 PH/21 Added message_body_newlines option.
1492 PH/22 Guard against possible overflow in moan_check_errorcopy().
1494 PH/23 POSIX allows open() to be a macro; guard against that.
1496 PH/24 If the recipient of an error message contained an @ in the local part
1497 (suitably quoted, of course), incorrect values were put in $domain and
1498 $local_part during the evaluation of errors_copy.
1504 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1505 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1508 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1509 issue a MAIL command.
1511 PH/02 In an ACL statement such as
1513 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1515 if a client was not listed at all, or was listed with a value other than
1516 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1517 the condition was not true (as it should be), so access was not denied.
1518 The bug was that the ! inversion was incorrectly passed on to the second
1519 item. This has been fixed.
1521 PH/03 Added additional dnslists conditions == and =& which are different from
1522 = and & when the dns lookup returns more than one IP address.
1524 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1525 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1527 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1528 FSYNC, which compiles an option called disable_fsync that allows for
1529 bypassing fsync(). The documentation is heavily laced with warnings.
1531 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1533 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1534 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1535 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1536 including adding "make clean"; (3) Added -fPIC when compiling the test
1537 dynamically loaded module, to get rid of a warning.
1539 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1540 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1541 = 0s. The bug is otherwise harmless.
1543 PH/07 There was a bug in the dovecot authenticator such that the value of
1544 $auth1 could be overwritten, and so not correctly preserved, after a
1545 successful authentication. This usually meant that the value preserved by
1546 the server_setid option was incorrect.
1548 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1550 PH/09 Installed PCRE release 7.0.
1552 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1553 run for batched SMTP input. It is now run at the start of every message
1554 in the batch. While fixing this I discovered that the process information
1555 (output by running exiwhat) was not always getting set for -bs and -bS
1556 input. This is fixed, and it now also says "batched" for BSMTP.
1558 PH/11 Added control=no_pipelining.
1560 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1561 patch, slightly modified), and move the expansion of helo_data till after
1562 the connection is made in the smtp transport (so it can use these
1565 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1567 PH/14 Added log_selector = +pid.
1569 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1571 PH/16 Add ${if forany and ${if forall.
1573 PH/17 Added dsn_from option to vary the From: line in DSNs.
1575 PH/18 Flush SMTP output before performing a callout, unless control =
1576 no_callout_flush is set.
1578 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1579 was true (the default) a successful delivery failed to delete the retry
1580 item, thus causing premature timeout of the address. The bug is now
1583 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1585 PH/21 Long custom messages for fakedefer and fakereject are now split up
1586 into multiline reponses in the same way that messages for "deny" and
1587 other ACL rejections are.
1589 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1590 with slight modification.
1592 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1593 draft, changing the syntax and factoring some duplicate code".
1595 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1596 for deliveries of the second and subsequent messages over the same SMTP
1599 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1600 ${reduce, with only minor "tidies".
1602 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1604 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1605 expansion side effects.
1607 PH/27 When a message times out after an over-quota error from an Exim-imposed
1608 quota, the bounce message says "mailbox is full". This message was not
1609 being given when it was a system quota that was exceeded. It now should
1612 MH/03 Made $recipients available in local_scan(). local_scan() already has
1613 better access to the recipient list through recipients_list[], but
1614 $recipients can be useful in postmaster-provided expansion strings.
1616 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1617 in the case of a MAIL command with additional options following the
1618 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1619 were accidentally chopped off.
1621 PH/29 SMTP synchronization checks are implemented when a command is read -
1622 there is a check that no more input is waiting when there shouldn't be
1623 any. However, for some commands, a delay in an ACL can mean that it is
1624 some time before the response is written. In this time, more input might
1625 arrive, invalidly. So now there are extra checks after an ACL has run for
1626 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1627 pipelining has not been advertised.
1629 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1631 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1632 returned in the SMTP response when smtp_return_error_details was false.
1633 This has been fixed.
1635 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1636 instead of the C I/O that was originally supplied, because problems were
1637 reported on Solaris.
1639 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1640 Exim which did not show up earlier: it was assuming that a call to
1641 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1642 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1643 was a macro that became an assignment, so it seemed to work. This has
1644 changed to a proper function call with a void return, hence the compile
1645 error. Exim's code has been fixed.
1647 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1650 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1652 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1654 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1655 because it turns out to be too big or there is a timeout) there is no
1656 "Completed" line in the log. When some messages of this type were
1657 selected by exigrep, they were listed as "not completed". Others were
1658 picked up by some special patterns. I have improved the selection
1659 criteria to be more general.
1661 PH/38 The host_find_failed option in the manualroute router can now be set
1662 to "ignore", to completely ignore a host whose IP address cannot be
1663 found. If all hosts are ignored, the behaviour is controlled by the new
1664 host_all_ignored option.
1666 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1667 homing or because of multiple MX records with /mx) generated more than
1668 one IP address, and the following item turned out to be the local host,
1669 all the secondary addresses of the first item were incorrectly removed
1670 from the list, along with the local host and any following hosts (which
1671 is what is supposed to happen).
1673 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1674 whoever called Exim into the -H file. In the case of the daemon it was
1675 behaving confusingly. When first started, it used values for whoever
1676 started the daemon, but after a SIGHUP it used the Exim user (because it
1677 calls itself on a restart). I have changed the code so that it now always
1680 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1681 message are rejected with the same error (e.g. no authentication or bad
1682 sender address), and a DATA command is nevertheless sent (as can happen
1683 with PIPELINING or a stupid MUA), the error message that was given to the
1684 RCPT commands is included in the rejection of the DATA command. This is
1685 intended to be helpful for MUAs that show only the final error to their
1688 PH/42 Another patch from the Sieve maintainer.
1690 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1691 Eximstats - Fixed some broken HTML links and added missing column headers
1693 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1694 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1696 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1698 PH/43 Yet another patch from the Sieve maintainer.
1700 PH/44 I found a way to check for a TCP/IP connection going away before sending
1701 the response to the final '.' that terminates a message, but only in the
1702 case where the client has not sent further data following the '.'
1703 (unfortunately, this is allowed). However, in many cases there won't be
1704 any further data because there won't be any more messages to send. A call
1705 to select() can be used: if it shows that the input is "ready", there is
1706 either input waiting, or the socket has been closed. An attempt to read
1707 the next input character can distinguish the two cases. Previously, Exim
1708 would have sent an OK response which the client would never have see.
1709 This could lead to message repetition. This fix should cure that, at
1710 least in a lot of common cases.
1712 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1713 advertised in response to EHLO.
1719 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1720 fixed by 4.65/MH/01 (is this a record?) are fixed:
1722 (i) An empty string was always treated as zero by the numeric comparison
1723 operators. This behaviour has been restored.
1725 (ii) It is documented that the numeric comparison operators always treat
1726 their arguments as decimal numbers. This was broken in that numbers
1727 starting with 0 were being interpreted as octal.
1729 While fixing these problems I realized that there was another issue that
1730 hadn't been noticed. Values of message_size_limit (both the global option
1731 and the transport option) were treated as octal if they started with 0.
1732 The documentation was vague. These values are now always treated as
1733 decimal, and I will make that clear in the documentation.
1739 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1740 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1743 MH/01 Don't check that the operands of numeric comparison operators are
1744 integers when their expansion is in "skipping" mode (fixes bug
1745 introduced by 4.64-PH/07).
1747 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1748 child addresses, Exim now panics and dies. Previously, because the count
1749 is held in a short int, deliveries were likely to be lost. As such a
1750 large number of recipients for a single message is ridiculous
1751 (performance will be very, very poor), I have chosen to impose a limit
1752 rather than extend the field.
1758 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1759 leftover -K file (the existence of which was triggered by #402).
1760 While we were at it, introduced process PID as part of the -K
1761 filename. This should rule out race conditions when creating
1764 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1765 processing considerably. Previous code took too long for large mails,
1766 triggering a timeout which in turn triggers #401.
1768 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1769 in the DK code in transports.c. sendfile() is not really portable,
1770 hence the _LINUX specificness.
1772 TF/01 In the add_headers option to the mail command in an Exim filter,
1773 there was a bug that Exim would claim a syntax error in any
1774 header after the first one which had an odd number of characters
1777 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1778 callout verification, Exim cached a "reject" for the entire domain. This
1779 is correct for most verifications, but it is not correct for a recipient
1780 verification with use_sender or use_postmaster set, because in that case
1781 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1782 case of MAIL FROM:<> rejection from other early rejections (e.g.
1783 rejection of HELO). When verifying a recipient using a non-null MAIL
1784 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1785 Whatever the result of the callout, the value of the domain cache is
1786 left unchanged (for any other kind of callout, getting as far as trying
1787 RCPT means that the domain itself is ok).
1789 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1792 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1793 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1794 session, the daemon ignores ECONNECT errors and logs others; it now
1795 ignores EPIPE as well.
1797 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1798 (quoted-printable decoding).
1800 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1801 later the small subsequent patch to fix an introduced bug.
1803 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1805 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1807 PH/08 An error is now given if message_size_limit is specified negative.
1809 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1810 to be given (somewhat) arbitrary names.
1812 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1815 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1816 miscellaneous code fixes
1818 PH/10 Added the log_reject_target ACL modifier to specify where to log
1821 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1822 hostname. This is wrong, because it relates to the incoming message (and
1823 probably the interface on which it is arriving) and not to the outgoing
1824 callout (which could be using a different interface). This has been
1825 changed to use the value of the helo_data option from the smtp transport
1826 instead - this is what is used when a message is actually being sent. If
1827 there is no remote transport (possible with a router that sets up host
1828 addresses), $smtp_active_hostname is used.
1830 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1831 tweaks were necessary in order to get it to work (see also 21 below):
1832 (a) The code assumed that strncpy() returns a negative number on buffer
1833 overflow, which isn't the case. Replaced with Exim's string_format()
1835 (b) There were several signed/unsigned issues. I just did the minimum
1836 hacking in of casts. There is scope for a larger refactoring.
1837 (c) The code used strcasecmp() which is not a standard C function.
1838 Replaced with Exim's strcmpic() function.
1839 (d) The code set only $1; it now sets $auth1 as well.
1840 (e) A simple test gave the error "authentication client didn't specify
1841 service in request". It would seem that Dovecot has changed its
1842 interface. Fortunately there's a specification; I followed it and
1843 changed what the client sends and it appears to be working now.
1845 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1848 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1849 address A is aliased to B and C, where B exists and C does not. Without
1850 -v the output is "A verified" because verification stops after a
1851 successful redirection if more than one address is generated. However,
1852 with -v the child addresses are also verified. Exim was outputting "A
1853 failed to verify" and then showing the successful verification for C,
1854 with its parentage. It now outputs "B failed to verify", showing B's
1855 parentage before showing the successful verification of C.
1857 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1858 look up a TXT record in a specific list after matching in a combined
1861 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1862 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1863 they consult the DNS. I had assumed they would set it the way they
1864 wanted; and indeed my experiments on Linux seem to show that in some
1865 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1866 To be on the safe side, however, I have now made the interface to
1867 host_find_byname() similar to host_find_bydns(), with an argument
1868 containing the DNS resolver options. The host_find_byname() function now
1869 sets these options at its start, just as host_find_bydns() does. The smtp
1870 transport options dns_qualify_single and dns_search_parents are passed to
1871 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1872 of host_find_byname() use the default settings of RES_DEFNAMES
1873 (qualify_single) but not RES_DNSRCH (search_parents).
1875 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1876 spool_read_header() do less string testing, by means of a preliminary
1877 switch on the second character of optional "-foo" lines. (This is
1878 overdue, caused by the large number of possibilities that now exist.
1879 Originally there were few.) While I was there, I also converted the
1880 str(n)cmp tests so they don't re-test the leading "-" and the first
1881 character, in the hope this might squeeze out yet more improvement.
1883 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1884 flag allowing group syntax was set by the header_syntax check but not
1885 turned off, possible causing trouble later; (2) The flag was not being
1886 set at all for the header_verify test, causing "group"-style headers to
1887 be rejected. I have now set it in this case, and also caused header_
1888 verify to ignore an empty address taken from a group. While doing this, I
1889 came across some other cases where the code for allowing group syntax
1890 while scanning a header line wasn't quite right (mostly, not resetting
1891 the flag correctly in the right place). These bugs could have caused
1892 trouble for malformed header lines. I hope it is now all correct.
1894 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1895 with the "reply" argument non-NULL. The code, however (which originally
1896 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1897 but it didn't always do it. This confused somebody who was copying the
1898 code for some other use. I have removed all the tests.
1900 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1901 feature that was used to support insecure browsers during the U.S. crypto
1902 embargo. It requires special client support, and Exim is probably the
1903 only MTA that supported it -- and would never use it because real RSA is
1904 always available. This code has been removed, because it had the bad
1905 effect of slowing Exim down by computing (never used) parameters for the
1906 RSA_EXPORT functionality.
1908 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1909 authenticator to fail if there's a tab character in the incoming data
1910 (there should never be unless someone is messing about, as it's supposed
1911 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1912 if the connection is using TLS or if the remote IP is the same as the
1913 local IP, and the "valid-client-cert option" if a client certificate has
1916 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1917 authenticators. This can be used for authorization after authentication
1918 succeeds. (In the case of plaintext, it servers for both authentication
1921 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1922 if any retry times were supplied.
1924 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1925 connection, where there is no client IP address to check. In this
1926 situation, the verify now always succeeds.
1928 PH/25 Applied John Jetmore's -Mset patch.
1930 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1932 PH/27 In a string expansion for a processed (not raw) header when multiple
1933 headers of the same name were present, leading whitespace was being
1934 removed from all of them, but trailing whitespace was being removed only
1935 from the last one. Now trailing whitespace is removed from each header
1936 before concatenation. Completely empty headers in a concatenation (as
1937 before) are ignored.
1939 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1940 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1942 PH/29 [Removed. This was a change that I later backed out, and forgot to
1943 correct the ChangeLog entry (that I had efficiently created) before
1944 committing the later change.]
1946 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1947 address errors (4xx response to RCPT) over the same connection as other
1948 messages routed to the same hosts. Such deliveries are always "forced",
1949 so retry times are not inspected. This resulted in far too many retries
1950 for the affected addresses. The effect occurred only when there were more
1951 hosts than the hosts_max_try setting in the smtp transport when it had
1952 the 4xx errors. Those hosts that it had tried were not added to the list
1953 of hosts for which the message was waiting, so if all were tried, there
1954 was no problem. Two fixes have been applied:
1956 (i) If there are any address or message errors in an SMTP delivery, none
1957 of the hosts (tried or untried) are now added to the list of hosts
1958 for which the message is waiting, so the message should not be a
1959 candidate for sending over the same connection that was used for a
1960 successful delivery of some other message. This seems entirely
1961 reasonable: after all the message is NOT "waiting for some host".
1962 This is so "obvious" that I'm not sure why it wasn't done
1963 previously. Hope I haven't missed anything, but it can't do any
1964 harm, as the worst effect is to miss an optimization.
1966 (ii) If, despite (i), such a delivery is accidentally attempted, the
1967 routing retry time is respected, so at least it doesn't keep
1968 hammering the server.
1970 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1971 in ${readsocket because some servers need this prod.
1973 PH/32 Added some extra debug output when updating a wait-xxx database.
1975 PH/33 The hint "could be header name not terminated by colon", which has been
1976 given for certain expansion errors for a long time, was not being given
1977 for the ${if def:h_colon_omitted{... case.
1979 PH/34 The spec says: "With one important exception, whenever a domain list is
1980 being scanned, $domain contains the subject domain." There was at least
1981 one case where this was not true.
1983 PH/35 The error "getsockname() failed: connection reset by peer" was being
1984 written to the panic log as well as the main log, but it isn't really
1985 panic-worthy as it just means the connection died rather early on. I have
1986 removed the panic log writing for the ECONNRESET error when getsockname()
1989 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1990 runs only) independently of the message's sender address. This meant
1991 that, if the 4xx error was in fact related to the sender, a different
1992 message to the same recipient with a different sender could confuse
1993 things. In particualar, this can happen when sending to a greylisting
1994 server, but other circumstances could also provoke similar problems.
1995 I have changed the default so that the retry time for these errors is now
1996 based a combination of the sender and recipient addresses. This change
1997 can be overridden by setting address_retry_include_sender=false in the
2000 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2001 remote server are returned as part of bounce messages. This was not
2002 happening for LMTP over a pipe (the lmtp transport), but now it is the
2003 same for both kinds of LMTP.
2005 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2006 in header lines that were in fact CNAMEs. This is no longer the case.
2008 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2009 and queue runs started by the daemon processed all messages. This has
2010 been fixed so that -R and -S can now usefully be given with -q<time>.
2012 PH/40 Import PCRE release 6.7 (fixes some bugs).
2014 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2016 PH/42 Give an error if -q is specified more than once.
2018 PH/43 Renamed the variables $interface_address and $interface_port as
2019 $received_ip_address and $received_port, to make it clear that these
2020 values apply to message reception, and not to the outgoing interface when
2021 a message is delivered. (The old names remain recognized, of course.)
2023 PH/44 There was no timeout on the connect() call when using a Unix domain
2024 socket in the ${readsocket expansion. There now is.
2026 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2027 be meaningful with "accept".
2029 SC/01 Eximstats V1.43
2030 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2032 SC/02 Eximstats V1.44
2033 Use a glob alias rather than an array ref in the generated
2034 parser. This improves both readability and performance.
2036 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2037 Collect SpamAssassin and rejection statistics.
2038 Don't display local sender or destination tables unless
2039 there is data to show.
2040 Added average volumes into the top table text output.
2042 SC/04 Eximstats V1.46
2043 Collect data on the number of addresses (recipients)
2044 as well as the number of messages.
2046 SC/05 Eximstats V1.47
2047 Added 'Message too big' to the list of mail rejection
2048 reasons (thanks to Marco Gaiarin).
2050 SC/06 Eximstats V1.48
2051 Mainlog lines which have GMT offsets and are too short to
2052 have a flag are now skipped.
2054 SC/07 Eximstats V1.49 (Alain Williams)
2055 Added the -emptyok flag.
2057 SC/08 Eximstats V1.50
2058 Fixes for obtaining the IP address from reject messages.
2060 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2061 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2062 whitesspace changes from 4.64-PH/27
2064 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2067 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2068 are found, allow negative numbers in numeric criteria)
2070 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2072 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2073 to match changes made in 4.64-PH/43
2075 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2077 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2078 "domain" is actually an IP address, and force a failure. This locks out
2079 those revolvers/nameservers that support "A-for-A" lookups, in
2080 contravention of the specifications.
2082 PH/48 When a host name was looked up from an IP address, and the subsequent
2083 forward lookup of the name timed out, the host name was left in
2084 $sender_host_name, contrary to the specification.
2086 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2087 restricted to single-key lookups, Exim was not diagnosing an error if
2088 * or *@ was used with a query-style lookup.
2090 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2092 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2093 long ago, but noone interested enough thought of it. Let's just say that
2094 the "1.1" means that there are some new functions that weren't there at
2095 some point in the past.
2097 PH/51 Error processing for expansion failure of helo_data from an smtp
2098 transport during callout processing was broken.
2100 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2101 tested/used via the -bh/-bhc/-bs options.
2103 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2104 bug, fixed in subsequent PCRE releases).
2106 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2107 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2109 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2115 SC/01 Use a glob alias rather than an array ref in eximstats generated
2116 parser. This improves both readability and performance.
2118 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2119 Don't display local sender or destination tables in eximstats unless
2120 there is data to show.
2121 Added average volumes into the eximstats top table text output.
2123 SC/03 Collect data on the number of addresses (recipients) as well
2124 as the number of messages in eximstats.
2126 TF/01 Correct an error in the documentation for the redirect router. Exim
2127 does (usually) call initgroups() when daemonizing.
2129 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2130 with consistent privilege compared to when running as a daemon.
2132 TF/03 Note in the spec that $authenticated_id is not set for local
2133 submissions from trusted users.
2135 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2136 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2138 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2139 by adding some example configuration directives to the default
2140 configuration file. A little bit of work is required to uncomment the
2141 directives and define how usernames and passwords are checked, but
2142 there is now a framework to start from.
2144 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2145 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2146 without this. I don't know how relevant this is to other LDAP libraries.
2148 PH/02 Add the verb name to the "unknown ACL verb" error.
2150 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2152 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2154 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2155 directory not expanded when it should be if an expanded home directory
2156 was set for the address (which is overridden by the transport).
2158 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2161 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2162 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2163 because it is too late at that time, and has no effect.
2165 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2166 security issue with \' (bugzilla #107). I could not use the
2167 PQescapeStringConn() function, because it needs a PGconn value as one of
2170 PH/08 When testing addresses using -bt, indicate those final addresses that
2171 are duplicates that would not cause an additional delivery. At least one
2172 person was confused, thinking that -bt output corresponded to deliveries.
2173 (Suppressing duplicates isn't a good idea as you lose the information
2174 about possibly different redirections that led to the duplicates.)
2176 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2177 systems where poll() doesn't work, in particular OS X.
2179 PH/10 Added more information to debugging output for retry time not reached.
2181 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2182 operations in malware.c.
2184 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2187 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2188 syslog_duplication set false, log lines that would normally be written
2189 both the the main log and to the reject log were not written to syslog at
2192 PH/14 In the default configuration, change the use of "message" in ACL warn
2193 statements to "add_header".
2195 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2196 not followed by a command (e.g. "seen endif").
2198 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2199 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2202 PH/17 Added extra conditions to the default value of delay_warning_condition
2206 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2207 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2208 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2211 The Auto-Submitted: and various List- headers are standardised, whereas I
2212 don't think Precedence: ever was.
2214 PH/18 Refactored debugging code in route_finduser() to show more information,
2215 in particular, the error code if getpwnam() issues one.
2217 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2218 This is apparently needed in addition to the PH/07 change above to avoid
2219 any possible encoding problems.
2221 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2222 but not after initializing Perl.
2224 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2225 output them only if debugging. By default they are written stderr,
2226 apparently, which is not desirable.
2228 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2231 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2234 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2236 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2237 authenticated or an ident call has been made. Suppress the default
2238 values for $authenticated_id and $authenticated_sender (but permit -oMai
2239 and -oMas) when testing with -bh.
2241 PH/24 Re-jigged the order of the tests in the default configuration so that the
2242 tests for valid domains and recipients precede the DNS black list and CSA
2243 tests, on the grounds that those ones are more expensive.
2245 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2246 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2247 This bug exists in every version of Exim that I still have, right back to
2250 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2251 However, an attempt to turn on case-sensitivity in a regex key by
2252 including (?-i) didn't work because the subject string was already
2253 lowercased, and the effects were non-intuitive. It turns out that a
2254 one-line patch can be used to allow (?-i) to work as expected.
2260 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2261 other effects) broke the use of negated acl sub-conditions.
2263 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2266 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2267 "Deny" causes Exim to reject the incoming connection with a 554 error.
2268 Unfortunately, if there is a major crisis, such as a disk failure,
2269 tcp-wrappers gives "deny", whereas what one would like would be some
2270 kind of temporary error. A kludge has been added to help with this.
2271 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2272 554 error is used if errno is still zero or contains ENOENT (which occurs
2273 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2276 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2278 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2279 errors. Otherwise a message that provokes a temporary error (when other
2280 messages do not) can cause a whole host to time out.
2282 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2283 addresses were routed directly to files or pipes from a redirect router.
2284 File deliveries just didn't batch; pipe deliveries might have suffered
2287 PH/06 A failure to get a lock for a hints database would erroneously always say
2288 "Failed to get write lock", even when it was really a read lock.
2290 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2291 of 0600. This has been changed to use the value of the lockfile_mode
2292 option (which defaults to 0600).
2294 PH/08 Applied small patch from the Sieve maintainer.
2296 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2297 folder from quota calculations, a direct delivery into this folder messed
2298 up the contents of the maildirsize file. This was because the regex was
2299 used only to exclude .Trash (or whatever) when the size of the mailbox
2300 was calculated. There was no check that a delivery was happening into an
2301 excluded directory. This bug has been fixed by ignoring all quota
2302 processing for deliveries into excluded directories.
2304 PH/10 Added the maildirfolder_create_regex option to appendfile.
2310 PH/01 The code for finding all the local interface addresses on a FreeBSD
2311 system running IPv6 was broken. This may well have applied to all BSD
2312 systems, as well as to others that have similar system calls. The broken
2313 code found IPv4 interfaces correctly, but gave incorrect values for the
2314 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2315 that it would not match correctly against @[] and not recognize the IPv6
2318 PH/02 The ipliteral router was not recognizing addresses of the form user@
2319 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2321 PH/03 Added disable_ipv6.
2323 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2324 decoded form, because it is most often used to construct To: headers
2325 lines in autoreplies, and the decoded form may well be syntactically
2326 invalid. However, $reply_address has leading white space removed, and all
2327 newlines turned into spaces so that the autoreply transport does not
2330 PH/05 If group was specified without a user on a router, and no group or user
2331 was specified on a transport, the group from the router was ignored.
2333 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2334 for visible compile-time settings that can be used to change these
2335 numbers, for those that want even more. Backwards compatibility with old
2336 spool files has been maintained. However, going back to a previous Exim
2337 release will lost any variables that are in spool files.
2339 PH/07 Two small changes when running in the test harness: increase delay when
2340 passing a TCP/IP connection to a new process, in case the original
2341 process has to generate a bounce, and remove special handling of
2342 127.0.0.2 (sic), which is no longer necessary.
2344 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2345 be the same on different OS.
2347 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2350 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2351 whether --show-vars was specified or not
2353 JJ/02 exipick: Added support for new ACL variable spool format introduced
2356 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2357 syntactically invalid From: or Reply-to: line, and a filter used this to
2358 generate an autoreply, and therefore failed to obtain an address for the
2359 autoreply, Exim could try to deliver to a non-existent relative file
2360 name, causing unrelated and misleading errors. What now happens is that
2361 it logs this as a hard delivery error, but does not attempt to create a
2364 PH/11 The exinext utility has a -C option for testing purposes, but although
2365 the given file was scanned by exinext itself; it wasn't being passed on
2366 when Exim was called.
2368 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2369 an end-of-file indication when reading a command response.
2371 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2372 compiled. In many other places in Exim, IPv6 addresses are always
2373 recognized, so I have changed this. It also means that IPv4 domain
2374 literals of the form [IPV4:n.n.n.n] are now always recognized.
2376 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2377 used if the router is not running as root, for example, when verifying at
2378 ACL time, or when using -bh. The debugging output from this situation was
2379 non-existent - all you got was a failure to exec. I have made two
2382 (a) Failures to set uid/gid, the current directory, or a process leader
2383 in a subprocess such as that created by queryprogram now generate
2384 suitable debugging ouput when -d is set.
2386 (b) The queryprogram router detects when it is not running as root,
2387 outputs suitable debugging information if -d is set, and then runs
2388 the subprocess without attempting to change uid/gid.
2390 PH/15 Minor change to Makefile for building test_host (undocumented testing
2393 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2394 additional section of a DNS packet that returns MX or SRV records.
2395 Instead, it always explicitly searches for A/AAAA records. This avoids
2396 major problems that occur when a DNS server includes only records of one
2397 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2398 fixed another bug: if SRV records were looked up and the corresponding
2399 address records were *not* found in the additional section, the port
2400 values from the SRV records were lost.
2402 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2403 using the correct key (the original address) when searching the retry
2404 rules in order to find which one to use for generating the retry hint.
2406 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2407 adding the default one. Similarly, if it contains a Reply-To: header, the
2408 errors_reply_to option, if set, is not used.
2410 PH/19 When calculating a retry time, Exim used to measure the "time since
2411 failure" by looking at the "first failed" field in the retry record. Now
2412 it does not use this if it is later than than the arrival time of the
2413 message. Instead it uses the arrival time. This makes for better
2414 behaviour in cases where some deliveries succeed, thus re-setting the
2415 "first failed" field. An example is a quota failure for a huge message
2416 when small messages continue to be delivered. Without this change, the
2417 "time since failure" will always be short, possible causing more frequent
2418 delivery attempts for the huge message than are intended.
2419 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2421 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2422 $1, $2, $3) because the numerical variables can be reset during some
2423 expansion items (e.g. "match"), thereby losing the authentication data.
2425 PH/21 Make -bV show the size of off_t variables so that the test suite can
2426 decide whether to run tests for quotas > 2G.
2428 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2429 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2430 filecount value is greater than 2G or if a quota value is greater than 2G
2431 on a system where the size of off_t is not greater than 4, a panic error
2434 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2435 never match. The debug and -bh output now contains an explicit error
2436 message indicating a malformed IPv4 address or mask.
2438 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2439 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2440 PH/23 above applies.
2442 PH/25 Do not write to syslog when running in the test harness. The only
2443 occasion when this arises is a failure to open the main or panic logs
2444 (for which there is an explicit test).
2446 PH/26 Added the /no_tell option to "control=freeze".
2448 PH/27 If a host name lookup failed very early in a connection, for example, if
2449 the IP address matched host_lookup and the reverse lookup yielded a name
2450 that did not have a forward lookup, an error message of the form "no IP
2451 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2452 could be logged. Now it outputs the IP address instead of "NULL".
2454 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2455 allows the sender and the authenticated sender to be set when
2456 submitting a message from within Exim. Since child_open_exim() is
2457 documented for local_scan(), the new function should be too.
2459 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2460 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2461 results in an empty string is now treated as unset.
2463 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2465 PH/31 Added sender_verify_fail logging option.
2467 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2468 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2469 tidied the source and removed it altogether.
2471 PH/33 When a queue run was abandoned because the load average was too high, a
2472 log line was always written; now it is written only if the queue_run log
2473 selector is set. In addition, the log line for abandonment now contains
2474 information about the queue run such as the pid. This is always present
2475 in "start" and "stop" lines but was omitted from the "abandon" line.
2477 PH/34 Omit spaces between a header name and the colon in the error message that
2478 is given when verify = headers_syntax fails (if there are lots of them,
2479 the message gets confusing).
2481 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2482 names, as there are now some PTR records that contain slashes. This check
2483 is only to protect against broken name servers that fall over on strange
2484 characters, so the fact that it applies to all lookups doesn't matter.
2486 PH/36 Now that the new test suite is complete, we can remove some of the
2487 special code in Exim that was needed for the old test suite. For example,
2488 sorting DNS records because real resolvers return them in an arbitrary
2489 order. The new test suite's fake resolver always returns records in the
2492 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2493 bounces) except when queue_only is set, to avoid logging races between
2494 the different processes.
2496 PH/38 Panic-die if .include specifies a non-absolute path.
2498 PH/39 A tweak to the "H" retry rule from its user.
2500 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2501 a label. They prevented compilation on older perls.
2503 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2504 a warning to be raised on newish perls.
2506 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2507 on queue. Changes to match documented behaviour of showing count of
2508 messages matching specified criteria.
2510 PH/40 Changed the default ident timeout from 30s to 5s.
2512 PH/41 Added support for the use of login_cap features, on those BSD systems
2513 that have them, for controlling the resources used by pipe deliveries.
2515 PH/42 The content-scanning code uses fopen() to create files in which to put
2516 message data. Previously it was not paying any attention to the mode of
2517 the files. Exim runs with umask(0) because the rest of the code creates
2518 files with open(), and sets the required mode explicitly. Thus, these
2519 files were ending up world-writeable. This was not a big issue, because,
2520 being within the spool directory, they were not world-accessible. I have
2521 created a function called modefopen, which takes an additional mode
2522 argument. It sets umask(777), creates the file, chmods it to the required
2523 mode, then resets the umask. All the relevant calls to fopen() in the
2524 content scanning code have been changed to use this function.
2526 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2527 to 24 hours. This avoids potential overflow problems when processing G
2528 and H retry rules. I suspect nobody ever tinkers with this value.
2530 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2532 PH/45 When the plaintext authenticator is running as a client, the server's
2533 challenges are checked to ensure they are valid base64 strings. By
2534 default, the authentication attempt is cancelled if an invalid string is
2535 received. Setting client_ignore_invalid_base64 true ignores these errors.
2536 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2537 they are received. Thus, the responses can be made to depend on the
2538 challenges. If an invalid string is ignored, an empty string is placed in
2541 PH/46 Messages that are created by the autoreply transport now contains a
2542 References: header, in accordance with RFCs 2822 and 3834.
2544 PH/47 Added authenticated_sender_force to the smtp transport.
2546 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2548 PH/49 Installed latest patch from the Sieve maintainer.
2550 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2551 was also set, the appendfile transport was unnecessarily scanning a
2552 directory of message files (e.g. for maildir delivery) to find the count
2553 of files (along with the size), even though it did not need this
2554 information. It now does the scan only if it needs to find either the
2555 size of the count of files.
2557 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2559 PH/52 Two bugs concerned with error handling when the smtp transport is
2562 (i) Exim was not creating retry information for temporary errors given
2563 for individual recipients after the DATA command when the smtp transport
2564 was used in LMTP mode. This meant that they could be retried too
2565 frequently, and not timed out correctly.
2567 (ii) Exim was setting the flag that allows error details to be returned
2568 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2569 recipients that were returned after the DATA command.
2571 PH/53 This is related to PH/52, but is more general: for any failing address,
2572 when detailed error information was permitted to be returned to the
2573 sender, but the error was temporary, then after the final timeout, only
2574 "retry timeout exceeded" was returned. Now it returns the full error as
2575 well as "retry timeout exceeded".
2577 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2578 do this, and (what is worse) MTAs that accept it.
2580 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2581 will now be deprecated.
2583 PH/56 New os.c-cygwin from the Cygwin maintainer.
2585 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2586 formats (previously only available in exim formats via -bpr, -bpru,
2587 and -bpra. Now also available in native and exiqgrep formats)
2589 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2590 with very large, slow to parse queues
2592 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2594 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2596 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2597 responses to authentication challenges, though it was showing the
2598 challenges; (ii) I've removed the CR characters from the debug output for
2601 PH/58 Allow for the insertion of a newline as well as a space when a string
2602 is turned into more than one encoded-word during RFC 2047 encoding. The
2603 Sieve code now uses this.
2605 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2606 data_4xx, lost_connection, tls_required.
2608 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2609 message was being sent as an SMTP response.
2611 PH/61 Add -l and -k options to exicyclog.
2613 PH/62 When verifying, if an address was redirected to one new address, so that
2614 verification continued, and the new address failed or deferred after
2615 having set something in $address_data, the value of $address_data was not
2616 passed back to the ACL. This was different to the case when no
2617 redirection occurred. The value is now passed back in both cases.
2619 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2620 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2621 use login_cap.h, so on its own it isn't the distinguishing feature. The
2622 new name refers directly to the setclassresources() function.
2624 PH/65 Added configuration files for NetBSD3.
2626 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2628 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2629 is preferred over IPv4.
2631 PH/68 The bounce_return_message and bounce_return_body options were not being
2632 honoured for bounces generated during the reception of non-SMTP messages.
2633 In particular, this applied to messages rejected by the ACL. This bug has
2634 been fixed. However, if bounce_return_message is true and bounce_return_
2635 body is false, the headers that are returned for a non-SMTP message
2636 include only those that have been read before the error was detected.
2637 (In the case of an ACL rejection, they have all been read.)
2639 PH/69 The HTML version of the specification is now built in a directory called
2640 spec_html instead of spec.html, because the latter looks like a path with
2641 a MIME-type, and this confuses some software.
2643 PH/70 Catch two compiler warnings in sieve.c.
2645 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2646 function verify_get_ident() calls ip_connect() to connect a socket, but
2647 if the "connect()" function timed out, ip_connect() used to close the
2648 socket. However, verify_get_ident() also closes the socket later, and in
2649 between Exim writes to the log, which may get opened at this point. When
2650 the socket was closed in ip_connect(), the log could get the same file
2651 descriptor number as the socket. This naturally causes chaos. The fix is
2652 not to close the socket in ip_connect(); the socket should be closed by
2653 the function that creates it. There was only one place in the code where
2654 this was missing, in the iplookup router, which I don't think anybody now
2655 uses, but I've fixed it anyway.
2657 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2658 well as to direct DNS lookups. Otherwise the handling of names in host
2659 lists is inconsistent and therefore confusing.
2665 PH/01 Two changes to the default runtime configuration:
2667 (1) Move the checks for relay_from_hosts and authenticated clients from
2668 after to before the (commented out) DNS black list checks.
2670 (2) Add control=submission to the relay_from_hosts and authenticated
2671 clients checks, on the grounds that messages accepted by these
2672 statements are most likely to be submissions.
2674 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2676 (1) Generate an error if the third argument for the ${prvs expansion is
2679 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2682 (3) Reset the variables that are obtained from the first argument of
2683 ${prvscheck and used in the second argument before leaving the code,
2684 because their memory is reclaimed, so using them afterwards may do
2687 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2688 one (it's much easier than Tom thought :-).
2690 (5) Because of (4), we can now allow for the use of $prvscheck_result
2691 inside the third argument.
2693 PH/03 For some reason, the default setting of PATH when running a command from
2694 a pipe transport was just "/usr/bin". I have changed it to
2697 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2698 anything to be listed in the output from -bV.
2700 PH/05 When a filter generated an autoreply, the entire To: header line was
2701 quoted in the delivery log line, like this:
2703 => >A.N.Other <ano@some.domain> <original@ddress> ...
2705 This has been changed so that it extracts the operative address. There
2706 may be more than one such address. If so, they are comma-separated, like
2709 => >ano@some.domain,ona@other.domain <original@ddress> ...
2711 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2712 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2713 not being looked up in the rDNS to get a host name, Exim was showing the
2714 IP address twice in Received: lines, even though the IP addresses were
2715 identical. For example:
2717 Received: from [1.2.3.4] (helo=[1.2.3.4])
2719 However, if the real host name was known, it was omitting the HELO data
2720 if it matched the actual IP address. This has been tidied up so that it
2721 doesn't show the same IP address twice.
2723 PH/07 When both +timestamp and +memory debugging was on, the value given by
2724 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2725 called by the debug printing, thereby overwriting the timestamp buffer.
2726 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2728 PH/08 When the original message was included in an autoreply transport, it
2729 always said "this is a copy of the message, including all the headers",
2730 even if body_only or headers_only was set. It now gives an appropriate
2733 PH/09 Applied a patch from the Sieve maintainer which:
2735 o fixes some comments
2736 o adds the (disabled) notify extension core
2737 o adds some debug output for the result of if/elsif tests
2738 o points to the current vacation draft in the documentation
2739 and documents the missing references header update
2743 o fixes a bug in processing the envelope test (when testing
2744 multiple envelope elements, the last element determinted the
2747 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2748 Electronic Mail") by including:
2750 Auto-submitted: auto-generated
2752 in the messages that it generates (bounce messages and others, such as
2753 warnings). In the case of bounce messages for non-SMTP mesages, there was
2754 also a typo: it was using "Auto_submitted" (underscore instead of
2755 hyphen). Since every message generated by Exim is necessarily in response
2756 to another message, thes have all been changed to:
2758 Auto-Submitted: auto-replied
2760 in accordance with these statements in the RFC:
2762 The auto-replied keyword:
2764 - SHOULD be used on messages sent in direct response to another
2765 message by an automatic process,
2767 - MUST NOT be used on manually-generated messages,
2769 - MAY be used on Delivery Status Notifications (DSNs) and Message
2770 Disposition Notifications (MDNs),
2772 - MUST NOT be used on messages generated by automatic or periodic
2773 processes, except for messages which are automatic responses to
2776 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2777 to the default Received: header definition.
2779 PH/12 Added log selector acl_warn_skipped (default on).
2781 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2782 variables because (a) they are in the wrong storage pool and (b) even if
2783 they were copied, it wouldn't work properly because of the caching.
2785 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2786 checking when decoding. Apparently there are clients that generate
2787 overlong encoded strings. Why am I not surprised?
2789 PH/15 If the first argument of "${if match_address" was not empty, but did not
2790 contain an "@" character, Exim crashed. Now it writes a panic log message
2791 and treats the condition as false.
2793 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2795 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2796 extension "envelope-auth". The code is finished and in agreement with
2797 other implementations, but there is no documentation so far and in fact,
2798 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2799 not changing the active code.
2801 Print executed "if" and "elsif" statements when debugging is used. This
2802 helps a great deal to understand what a filter does.
2804 Document more things not specified clearly in RFC3028. I had all this
2805 sorted out, when out of a sudden new issues came to my mind. Oops."
2807 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2810 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2811 canonical form (as documented). However, after a host name lookup from
2812 the IP address, check_host() was doing a simple string comparison with
2813 addresses acquired from the DNS when checking that the found name did
2814 have the original IP as one of its addresses. Since any found IPv6
2815 addresses are likely to be in abbreviated form, the comparison could
2816 fail. Luckily, there already exists a function for doing the comparison
2817 by converting both addresses to binary, so now that is used instead of
2818 the text comparison.
2820 PH/20 There was another similar case to PH/19, when a complete host name was
2821 given in a host list; looking up its IP address could give an abbreviated
2822 form, whereas the current host's name might or might not be abbreviated.
2823 The same fix has been applied.
2829 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2830 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2833 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2834 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2836 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2838 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2839 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2840 Also, for "Auto-Submitted", treat anything other than "no" as
2841 non-personal, in accordance with RFC 3834. (Previously it treated
2842 anything starting "auto-" as non-personal.)
2844 TF/01 The control=submission/name=... option had a problem with syntax
2845 errors if the name included a slash character. The /name= option
2846 now slurps the rest of the string, so it can include any characters
2847 but it must come last in the list of options (after /sender_retain
2850 PH/05 Some modifications to the interface to the fake nameserver for the new
2858 TK/01 Added the "success_on_redirect" address verification option. See
2859 NewStuff for rationale and an example.
2861 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2863 PH/02 Patch to exigrep to allow it to work on syslog lines.
2865 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2866 fread() to skip over the body file's header line, because in Cygwin the
2867 header line is locked and is inaccessible.
2869 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2870 co-exist for some time) to make it clear that it is the Exim ID that is
2871 referenced, not the Message-ID: header line.
2873 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2874 string_format() function, because snprintf() does not exist on all
2877 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2878 ${stat: expansion item.
2880 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2881 protocol synchronization error", to keep the pedants happy.
2883 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2884 well as for IRIX systems, when gcc is being used. See the host.c source
2887 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2889 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2892 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2893 transport and to the smtp transport in LMTP mode.
2895 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2897 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2898 run a filter in a subprocess. This could lead to confusion in subsequent
2899 lookups in the parent process. There should also be a search_tidyup() at
2900 the end of the subprocess.
2902 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2903 only if the host matched helo_try_verify_hosts, which caused the
2904 verification to occur when the EHLO/HELO command was issued. The ACL just
2905 tested the remembered result. Now, if a previous verification attempt has
2906 not happened, "verify = helo" does it there and then.
2908 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2910 TK/03 Fix log output including CR from clamd.
2912 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2913 exist provoked a memory error which could cause a segfault.
2915 PH/15 Installed PCRE 6.2
2917 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2919 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2920 of the problem. Specifically, suggested +O2 rather than +O1 for the
2923 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2925 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2926 record, the port was not correctly logged when the outgoing_port log
2927 selector was set (it logged the transort's default port).
2929 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2930 fallback_hosts, and "hosts" in the smtp transport.
2932 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2933 host errors such as "Connection refused".
2935 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2936 authentication with radiusclient 0.4.9:
2938 - Error returned from rc_read_config was caught wrongly
2939 - Username/password not passed on to radius server due to wrong length.
2941 The presumption is that some radiusclient API changes for 4.51/PH/17
2942 were not taken care of correctly. The code is still untested by me (my
2943 Linux distribution still has 0.3.2 of radiusclient), but it was
2944 contributed by a Radius user.
2946 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2947 expanding the "port" option of the smtp transport.
2949 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2950 while reading a MIME header. Thanks to Tom Hughes for a patch.
2952 PH/24 Include config.h inside local_scan.h so that configuration settings are
2955 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2956 that in an ACL for RCPT (for example), you can examine exactly what was
2959 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2960 commands, but it was not correctly comparing the address with the actual
2961 client host address. Thus, it would show the EHLO address in Received:
2962 header lines when this was not necessary.
2964 PH/27 Added the % operator to ${eval:}.
2966 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2967 it should be ignoring failures (because with -C, for example, it has lost
2968 privilege). It wasn't ignoring creation failures other than "already
2971 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2974 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2975 because an input line was too long, either on its own, or by virtue of
2976 too many continuations, the temporary file was not being removed, and the
2977 return code was incorrect.
2979 PH/31 Missing "BOOL" in function definition in filtertest.c.
2981 PH/32 Applied Sieve patches from the maintainer.
2983 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2985 PH/33 Added "verify = not_blind".
2987 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2988 Local/Makefile (with some defaults set). These are used in built scripts
2989 such as exicyclog, but they have never been used in the exim_install
2990 script (though there are many overriding facilities there). I have
2991 arranged that the exim_install script now takes note of these two
2994 PH/35 Installed configuration files for Dragonfly.
2996 PH/36 When a locally submitted message by a trusted user did not contain a
2997 From: header, and the sender address was obtained from -f or from an SMTP
2998 MAIL command, and the trusted user did not use -F to supply a sender
2999 name, $originator_name was incorrectly used when constructing a From:
3000 header. Furthermore, $originator_name was used for submission mode
3001 messages from external hosts without From: headers in a similar way,
3002 which is clearly wrong.
3004 PH/37 Added control=suppress_local_fixups.
3006 PH/38 When log_selector = +received_sender was set, and the addition of the
3007 sender made the log line's construction buffer exactly full, or one byte
3008 less than full, an overflow happened when the terminating "\n" was
3011 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3012 when the result of a list match is failure because a DNS lookup failed.
3014 PH/40 RM_COMMAND is now used in the building process.
3016 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3017 the "build-* directories that it finds.
3019 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3020 address in a domain literal was a prefix of an interface address.
3022 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3023 when verifying a sender address, unless rewrite_headers is false.
3025 PH/44 Wrote a long comment about why errors_to addresses are verified as
3026 recipients, not senders.
3028 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3029 the ratelimit ACL was added.
3031 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3033 PH/46 Added extra information about PostgreSQL errors to the error string.
3035 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3036 avoiding the need to install special zones in a real server. This is
3037 backwards compatible; if it can't find the fake resolver, it drops back.
3038 Thus, both old and new test suites can be run.
3040 TF/02 Added util/ratelimit.pl
3042 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3043 clock is set back in time.
3045 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3046 Candler <B.Candler@pobox.com>.
3048 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3049 OK for addresses that are the result of redirections.
3051 PH/48 A number of further additions for the benefit of the new test suite,
3052 including a fake gethostbyname() that interfaces to the fake DNS resolver
3055 TF/06 The fix for widen_domains has also been applied to qualify_single and
3056 search_parents which are the other dnslookup options that can cause
3059 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3062 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3064 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3065 service" if it encountered an MX record with an empty target hostname.
3066 The message is now "an MX or SRV record indicated no SMTP service".
3068 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3069 if the DNS of the sending site is misconfigured. This is quite a
3070 common situation. This change restores the behaviour of treating a
3071 helo verification defer as a failure.
3073 PH/51 If self=fail was set on a router, the bounce message did not include the
3074 actual error message.
3080 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3082 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3083 command itself ended in error, the underlying message about the transport
3084 filter timeout was being overwritten with the pipe command error. Now the
3085 underlying error message should be appended to the second error message.
3087 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3089 PH/02 Reduce the amount of output that "make" produces by default. Full output
3090 can still be requested.
3092 PH/03 The warning log line about a condition test deferring for a "warn" verb
3093 was being output only once per connection, rather than after each
3094 occurrence (because it was using the same function as for successful
3095 "warn" verbs). This seems wrong, so I have changed it.
3097 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3098 it should not have, which might have caused a crash in the right
3099 circumstances, but probably never did.
3101 PH/04 Installed a modified version of Tony Finch's patch to make submission
3102 mode fix the return path as well as the Sender: header line, and to
3103 add a /name= option so that you can make the user's friendly name appear
3106 TF/03 Added the control = fakedefer ACL modifier.
3108 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3109 Mark Lowes for thorough testing.
3111 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3113 TK/03 Merged latest SRS patch from Miles Wilton.
3115 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3116 with the definition in sysexits.h (which is #included earlier).
3117 Fortunately, Exim does not actually use EX_OK. The code used to try to
3118 preserve the sysexits.h value, by assumimg that macro definitions were
3119 scanned for macro replacements. I have been disabused of this notion,
3120 so now the code just undefines EX_OK before #including unistd.h.
3122 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3123 in the smtp transport. When a block could not be written in a single
3124 write() function, the timeout was being re-applied to each part-write.
3125 This seems wrong - if the receiver was accepting one byte at a time it
3126 would take for ever. The timeout is now adjusted when this happens. It
3127 doesn't have to be particularly precise.
3129 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3130 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3132 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3133 without a domain if the check to <postmaster@domain> fails.
3135 SC/01 Eximstats: added -xls and the ability to specify output files
3136 (patch written by Frank Heydlauf).
3138 SC/02 Eximstats: use FileHandles for outputing results.
3140 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3142 SC/04 Eximstats: fixed display of large numbers with -nvr option
3144 SC/05 Eximstats: fixed merging of reports with empty tables.
3146 SC/06 Eximstats: added the -include_original_destination flag
3148 SC/07 Eximstats: removed tabs and trailing whitespace.
3150 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3152 TK/06 MBOX spool code: Add real "From " MBOX separator line
3153 so the .eml file is really in mbox format (even though
3154 most programs do not really care). Patch from Alex Miller.
3156 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3157 The latter is generated from $received_to and is only set if the
3158 message has one envelope recipient. SA can use these headers,
3159 obviously out-of-the-box. Patch from Alex Miller.
3161 PH/08 The ${def test on a variable was returning false if the variable's
3162 value was "0", contrary to what the specification has always said!
3163 The result should be true unless the variable is empty.
3165 PH/09 The syntax error of a character other than { following "${if
3166 def:variable_name" (after optional whitespace) was not being diagnosed.
3167 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3168 accidental colon was present, for example, could give incorrect results.
3170 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3171 result is used (not including appendfile, where other changes are about
3174 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3175 This involved changing a lot of size variables from int to off_t. It
3176 should work with maildirs and everything.
3178 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3179 spamd dying while we are connected to it.
3181 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3184 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3185 function for BDB 4.3.
3187 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3189 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3190 the sources for PCRE have been re-arranged and more files are now
3193 PH/15 The code I had for printing potentially long long variables in PH/11
3194 above was not the best (it lost precision). The length of off_t variables
3195 is now inspected at build time, and an appropriate printing format (%ld
3196 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3197 to be "long long int" or "long int". This is needed for the internal
3198 formatting function string_vformat().
3200 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3201 the configuration file to be ":syslog", then the script "guesses" where
3202 the logs files are, rather than using the compiled in default. In our
3203 case the guess is not the same as the compiled default, so the script
3204 suddenly stopped working when I started to use syslog. The patch checks
3205 to see if log_file_path is "". If so, it attempts to read it from exim
3206 with no configuration file to get the compiled in version, before it
3207 falls back to the previous guessing code."
3209 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3210 implementing BATV in an Exim configuration. See NewStuff for the gory
3213 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3214 Makefile that are specific to HP-UX.
3216 PH/18 If the "use_postmaster" option was set for a recipient callout together
3217 with the "random" option, the postmaster address was used as the MAIL
3218 FROM address for the random test, but not for the subsequent recipient
3219 test. It is now used for both.
3221 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3222 patch removes a few documentation additions to RFC 3028, because the
3223 latest draft now contains them. It adds the new en;ascii-case comparator
3224 and a new error check for 8bit text in MIME parts. Comparator and
3225 require names are now matched exactly. I enabled the subaddress
3226 extension, but it is not well tested yet (read: it works for me)."
3228 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3229 rework some of the code of TK/09 above to avoid the hardwired use of
3230 "%lld" and "long long". Replaced the call to snprintf() with a call to
3233 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3234 records point to non-existent hosts", "retry timeout exceeded", and
3235 "retry time not reached for any host after a long failure period".
3237 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3238 experimental DomainKeys support:
3240 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3241 (2) On an error such as an illegally used "control", the wrong name for
3242 the control was given.
3244 These problems did NOT occur unless DomainKeys support was compiled.
3246 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3248 PH/24 Added ${if match_ip condition.
3250 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3251 databases so that it will be absolutely obvious if a crash occurs in the
3252 DB library. This is a regular occurrence (often caused by mis-matched
3255 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3256 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3257 picked up on a user's system that detects such things. There doesn't seem
3258 to be a gcc warning option for this - only an attribute that has to be
3259 put on the function's prototype. It seems that in Fedora Core 4 they have
3260 set this on a number of new functions. No doubt there will be more in due
3263 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3264 specify a transport. However, if an address that was verified by such a
3265 router was the subject of a callout, Exim crashed because it tried to
3266 read the rcpt_include_affixes from the non-existent transport. Now it
3267 just assumes that the setting of that option is false. This bug was
3268 introduced by 4.51/PH/31.
3270 PH/28 Changed -d+all to exclude +memory, because that information is very
3271 rarely of interest, but it makes the output a lot bigger. People tend to
3272 do -d+all out of habit.
3274 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3275 code in os-type was giving problems when libc.so lives in lib64, like on
3278 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3279 aren't the modern standard, and it seems that some systems' include files
3280 don't always have them. Exim was already checking for some of the newer
3281 ones like T_AAAA, and defining it itself. I've added checks for all the
3282 record types that Exim uses.
3284 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3285 not automatically generating a new one, as it is supposed to. This
3286 prevented TLS from working. If the file did exist, but contained invalid
3287 data, a new version was generated, as expected. It was only the case of a
3288 non-existent file that was broken.
3290 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3291 with a change in libdomainkeys > 0.64.
3293 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3294 from DNS. If the selector record carries the flag, it now has
3295 precedence over the domain-wide flag.
3297 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3299 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3300 the use of a port name that isn't defined in /etc/services) occurred, the
3301 message was deferred as in a normal delivery, and thus remained on the
3302 spool, instead of being failed because of the mua_wrapper setting. This
3303 is now fixed, and I tidied up some of the mua_wrapper messages at the
3306 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3307 the messages in a hash of arrays rather than using individual hashes.
3308 This is a bit cleaner and results in dramatic memory savings, albeit
3309 at a slight CPU cost.
3311 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3312 as requested by Marc Sherman.
3314 SC/10 Eximstats: added histograms for user specified patterns as requested
3317 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3319 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3320 fopen() in the content-scanning modules that did not already have it.
3326 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3327 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3329 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3331 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3333 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3334 to test Sieve filters that use "vacation".
3336 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3337 that changes the way the GnuTLS parameters are stored in the cache file.
3338 The new format can be generated externally. For backward compatibility,
3339 if the data in the cache doesn't make sense, Exim assumes it has read an
3340 old-format file, and it generates new data and writes a new file. This
3341 means that you can't go back to an older release without removing the
3344 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3345 work if there are any delivery delays because "one_time" forces the
3346 parent to be marked "delivered", so its unseen clone is never tried
3347 again. For this reason, Exim now forbids the simultaneous setting of
3350 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3351 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3352 entry if you want to know the details. The fix, however, neglected to
3353 consider the case when local delivery batching is involved. The test for
3354 "previously delivered" was not happening when checking to see if an
3355 address could be batched with a previous (undelivered) one; under
3356 certain circumstances this could lead to multiple deliveries to the same
3359 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3360 in its include files, and this causes problems building Exim.
3362 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3363 header_syntax) but Exim was just ignoring anything given after a slash.
3364 In particular, this caused confusion with an attempt to use "verify =
3365 reverse_host_lookup/defer_ok". An error is now given when options are
3366 supplied for verify items that do not have them. (Maybe reverse_host_
3367 lookup should have a defer_ok option, but that's a different point.)
3369 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3370 defined by RFC 821) to 2048, because there were problems with some AUTH
3371 commands, and RFC 1869 says the size should be increased for extended
3372 SMTP commands that take arguments.
3374 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3377 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3378 "unknown" error; now it says that the functionality isn't in the binary.
3380 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3381 an address' error message when a string expansion fails (syntax or
3382 whatever). Otherwise the password may appear in the log. Following change
3383 PH/42 below, there is no longer a chance of it appearing in a bounce
3386 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3388 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3389 of its addresses was ever tried. (Bugzilla bug #2.)
3391 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3392 the result incorrectly in the debug output. (It correctly added a newline
3393 to what was transported.)
3395 TF/01 Added $received_time.
3397 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3398 commented out examples of how to interface to a virus scanner and to
3399 SpamAssassin. Also added commented examples of av_scanner and
3400 spamd_address settings.
3402 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3403 and controls are allowed in which ACLs. There were a couple of minor
3404 errors. Some of the entries in the conditions table (which is a table of
3405 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3406 negation of where the condition IS allowed.
3408 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3410 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3411 header file does not have a version number, so I've had to invent a new
3412 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3413 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3414 radiusclient), but it was contributed by a Radius user.
3416 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3417 files or directories, for OpenSSL.
3419 PH/19 When an Exim process that is running as root has to create an Exim log
3420 file, it does so in a subprocess that runs as exim:exim so as to get the
3421 ownership right at creation (otherwise, other Exim processes might see
3422 the file with the wrong ownership). There was no test for failure of this
3423 fork() call, which would lead to the process getting stuck as it waited
3424 for a non-existent subprocess. Forks do occasionally fail when resources
3425 run out. I reviewed all the other calls to fork(); they all seem to check
3428 PH/20 When checking for unexpected SMTP input at connect time (before writing
3429 the banner), Exim was not dealing correctly with a non-positive return
3430 from the read() function. If the client had disconnected by this time,
3431 the result was a log entry for a synchronization error with an empty
3432 string after "input=" when read() returned zero. If read() returned -1
3433 (an event I could not check), uninitialized data bytes were printed.
3434 There were reports of junk text (parts of files, etc) appearing after
3437 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3439 PH/22 Added support for macro redefinition, and (re)definition in between
3440 driver and ACL definitions.
3442 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3443 forgetting to use the resulting value; it was using the unexpanded value.
3445 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3446 hadn't been configured. The fix is from Juergen Kreileder, who
3447 understands it better than I do:
3449 "Here's what I see happening with three configured cyrus_sasl
3450 authenticators configured (plain, login, cram-md5):
3452 On startup auth_cyrus_sasl_init() gets called for each of these.
3453 This means three calls to sasl_listmech() without a specified mech_list.
3454 => SASL tests which mechs of all available mechs actually work
3455 => three warnings about OTP not working
3456 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3458 With the patch, sasl_listmech() also gets called three times. But now
3459 SASL's mech_list option is set to the server_mech specified in the the
3460 authenticator. Or in other words, the answer from sasl_listmech()
3461 gets limited to just the mech you're testing for (which is different
3463 => the return list contains just 'plain' or 'login', 'cram-md5' or
3464 nothing depending on the value of ob->server_mech.
3466 I've just tested the patch: Authentication still works fine,
3467 unavailable mechs specified in the exim configuration are still
3468 caught, and the auth.log warnings about OTP are gone."
3470 PH/25 When debugging is enabled, the contents of the command line are added
3471 to the debugging output, even when log_selector=+arguments is not
3474 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3475 answer is "GNU", and only if the return is "GNU/something" is the answer
3478 PH/27 $acl_verify_message is now set immediately after the failure of a
3479 verification in an ACL, and so is available in subsequent modifiers. In
3480 particular, the message can be preserved by coding like this:
3482 warn !verify = sender
3483 set acl_m0 = $acl_verify_message
3485 Previously, $acl_verify_message was set only while expanding "message"
3486 and "log_message" when a very denied access.
3488 PH/28 Modified OS/os.c-Linux with
3490 -#ifndef OS_LOAD_AVERAGE
3491 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3493 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3494 nomenclature these days.)
3496 PH/29 Installed patch from the Sieve maintainer that adds the options
3497 sieve_useraddress and sieve_subaddress to the redirect router.
3499 PH/30 In these circumstances:
3500 . Two addresses routed to the same list of hosts;
3501 . First host does not offer TLS;
3502 . First host accepts first address;
3503 . First host gives temporary error to second address;
3504 . Second host offers TLS and a TLS session is established;
3505 . Second host accepts second address.
3506 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3507 and peerdn, if requested) that were in fact used only for the second
3510 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3511 attention to any local part prefix or suffix that was matched by the
3512 router that accepted the address. It now behaves in the same way as it
3513 does for delivery: the affixes are removed from the local part unless
3514 rcpt_include_affixes is set on the transport.
3516 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3517 timeout during the DATA phase of an incoming message.
3519 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3520 applied a patch sanctioned by the Sieve maintainer.
3522 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3523 the uid or gid is negative. A case of a negative gid caused this to be
3524 noticed. The fix allows for either to be negative.
3526 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3527 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3530 PH/36 The change PH/12 above was broken. Fixed it.
3532 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3533 the grounds that routing the same address twice would always produce the
3534 same answer. This might have been true once, but it is certainly no
3535 longer true now. Routing a child address may depend on the previous
3536 routing that produced that child. Some complicated redirection strategies
3537 went wrong when messages had multiple recipients, and made Exim's
3538 behaviour dependent on the order in which the addresses were given.
3540 I have moved the duplicate checking until after the routing is complete.
3541 Exim scans the addresses that are assigned to local and remote
3542 transports, and removes any duplicates. This means that more work will be
3543 done, as duplicates will always all be routed, but duplicates are
3544 presumably rare, so I don't expect this is of any significance.
3546 For deliveries to pipes, files, and autoreplies, the duplicate checking
3547 still happens during the routing process, since they are not going to be
3550 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3551 It corrects a timeout issue with spamd. This is Ian's comment: "The
3552 background is that sometimes spamd either never reads data from a
3553 connection it has accepted, or it never writes response data. The exiscan
3554 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3555 blindly assumes that writes won't block so it may never time out."
3557 PH/39 Allow G after quota size as well as K and M.
3559 PH/40 The value set for $authenticated_id in an authenticator may not contain
3560 binary zeroes or newlines because the value is written to log lines and
3561 to spool files. There was no check on this. Now the value is run through
3562 the string_printing() function so that such characters are converted to
3563 printable escape sequences.
3565 PH/41 $message_linecount is a new variable that contains the total number of
3566 lines in the message. Compare $body_linecount, which is the count for the
3569 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3570 bounce and delay warning messages, except in certain special cases, which
3573 (a) An SMTP error message from a remote host;
3574 (b) A message specified in a :fail: redirection;
3575 (c) A message specified in a "fail" command in a system filter;
3576 (d) A message specified in a FAIL return from the queryprogram router;
3577 (e) A message specified by the cannot_route_message router option.
3579 In these cases only, Exim does include the error details in bounce and
3580 warning messages. There are also a few cases where bland messages such
3581 as "unrouteable address" or "local delivery error" are given.
3583 PH/43 $value is now also set for the "else" part of a ${run expansion.
3585 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3586 being worked on, but at least Exim now implements the latest version to
3589 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3590 process to complete was treated as a delivery failure, a timeout while
3591 writing the message to the pipe was logged, but erroneously treated as a
3592 successful delivery. Such timeouts include transport filter timeouts. For
3593 consistency with the overall process timeout, these timeouts are now
3594 treated as errors, giving rise to delivery failures by default. However,
3595 there is now a new Boolean option for the pipe transport called
3596 timeout_defer, which, if set TRUE, converts the failures into defers for
3597 both kinds of timeout. A transport filter timeout is now identified in
3600 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3601 systems where "make" and "gmake" are different, calling "gmake" at top
3602 level broke things. I've arranged for the value of $(MAKE) to be passed
3603 from the Makefile to this script so that it can call the same version of
3607 A note about Exim versions 4.44 and 4.50
3608 ----------------------------------------
3610 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3611 changes of various kinds. As a consequence, a big documentation update was
3612 needed. This delayed the release for rather longer than seemed good, especially
3613 in the light of a couple of (minor) security issues. Therefore, the changes
3614 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3615 release. So 4.44 and 4.50 are in effect two different branches that both start
3618 I have left the 4.50 change log unchanged; it contains all the changes since
3619 4.43. The change log for 4.44 is below; many of its items are identical to
3620 those for 4.50. This seems to be the most sensible way to preserve the
3621 historical information.
3627 1. Minor wording change to the doc/README.SIEVE file.
3629 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3630 computation of the current number of files was incorrect.
3632 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3633 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3636 4. Give more explanation in the error message when the command for a transport
3637 filter fails to execute.
3639 5. There are several places where Exim runs a non-Exim command in a
3640 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3641 was being done only for the command run by the queryprogram router. It is
3642 now done for all such subprocesses. The other cases are: ${run, transport
3643 filters, and the commands run by the lmtp and pipe transports.
3645 6. Added CONFIGURE_GROUP build-time option.
3647 7. Some older OS have a limit of 256 on the maximum number of file
3648 descriptors. Exim was using setrlimit() to set 1000 as a large value
3649 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3650 systems. I've change it so that if it can't get 1000, it tries for 256.
3652 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3653 was an oversight, and furthermore, ever since the addition of extra
3654 controls (e.g. 4.43/32), the checks on when to allow different forms of
3655 "control" were broken. There should now be diagnostics for all cases when a
3656 control that does not make sense is encountered.
3658 9. Added the /retain_sender option to "control=submission".
3660 10. $recipients is now available in the predata ACL (oversight).
3662 11. Tidy the search cache before the fork to do a delivery from a message
3663 received from the command line. Otherwise the child will trigger a lookup
3664 failure and thereby defer the delivery if it tries to use (for example) a
3665 cached ldap connection that the parent has called unbind on.
3667 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3668 of $address_data from the recipient verification was clobbered by the
3669 sender verification.
3671 13. The value of address_data from a sender verification is now available in
3672 $sender_address_data in subsequent conditions in the ACL statement.
3674 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3676 15. Added a new option "connect=<time>" to callout options, to set a different
3679 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3680 was its contents. (It was OK if the option was not defined at all.)
3682 17. A "Completed" log line is now written for messages that are removed from
3683 the spool by the -Mrm option.
3685 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3686 information about exactly what failed.
3688 19. Added -dd to debug only the daemon process.
3690 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3691 handles timeouts, both on the server side and network timeouts. Renamed the
3692 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3694 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3695 It is now set to "smtps".
3697 22. $host_address is now set to the target address during the checking of
3698 ignore_target_hosts.
3700 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3701 being passed; this would have caused $sender_host_name to have been used if
3702 matching the list had actually called for a host name (not very likely,
3703 since this list is usually IP addresses). A host name is now passed as
3706 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3707 code that specifies a non-restarting handler (typically sigaction() in
3708 modern systems) in an attempt to fix a rare and obscure crash bug.
3710 25. Narrowed the window for a race in the daemon that could cause it to ignore
3711 SIGCHLD signals. This is not a major problem, because they are used only to
3712 wake it up if nothing else does.
3714 26. A malformed maildirsize file could cause Exim to calculate negative values
3715 for the mailbox size or file count. Odd effects could occur as a result.
3716 The maildirsize information is now recalculated if the size or filecount
3719 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3720 support for a long time. Removed HAVE_SYS_VFS_H.
3722 28. Installed the latest version of exipick from John Jetmore.
3724 29. In an address list, if the pattern was not a regular expression, an empty
3725 subject address (from a bounce message) matched only if the pattern was an
3726 empty string. Non-empty patterns were not even tested. This was the wrong
3727 because it is perfectly reasonable to use an empty address as part of a
3728 database query. An empty address is now tested by patterns that are
3729 lookups. However, all the other forms of pattern expect the subject to
3730 contain a local part and a domain, and therefore, for them, an empty
3731 address still always fails if the pattern is not itself empty.
3733 30. Exim went into a mad DNS loop when attempting to do a callout where the
3734 host was specified on an smtp transport, and looking it up yielded more
3735 than one IP address.
3737 31. Re-factored the code for checking spool and log partition space into a
3738 function that finds that data and another that does the check. The former
3739 is then used to implement four new variables: $spool_space, $log_space,
3740 $spool_inodes, and $log_inodes.
3742 32. The RFC2047 encoding function was originally intended for short strings
3743 such as real names; it was not keeping to the 75-character limit for
3744 encoded words that the RFC imposes. It now respects the limit, and
3745 generates multiple encoded words if necessary. To be on the safe side, I
3746 have increased the buffer size for the ${rfc2047: expansion operator from
3749 33. It is now permitted to omit both strings after an "if" condition; if the
3750 condition is true, the result is "true". As before, when the second string
3751 is omitted, a false condition yields an empty string. This makes it less
3752 cumbersome to write custom ACL and router conditions.
3754 34. Failure to deliver a bounce message always caused it to be frozen, even if
3755 there was an errors_to setting on the router. The errors_to setting is now
3758 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3759 canonical form (fully expanded) before being placed in
3760 $sender_host_address.
3762 36. The table in the code that translates DNS record types into text (T_A to
3763 "A" for instance) was missing entries for NS and CNAME. It is just possible
3764 that this could have caused confusion if both these types were looked up
3765 for the same domain, because the text type is used as part of Exim's
3766 per-process caching. But the chance of anyone hitting this buglet seems
3769 37. The dnsdb lookup has been extended in a number of ways.
3771 (1) There is a new type, "zns", which walks up the domain tree until it
3772 finds some nameserver records. It should be used with care.
3774 (2) There is a new type, "mxh", which is like "mx" except that it returns
3775 just the host names, not the priorities.
3777 (3) It is now possible to give a list of domains (or IP addresses) to be
3778 looked up. The behaviour when one of the lookups defers can be
3779 controlled by a keyword.
3781 (4) It is now possible to specify the separator character for use when
3782 multiple records are returned.
3784 38. The dnslists ACL condition has been extended: it is now possible to supply
3785 a list of IP addresses and/or domains to be looked up in a particular DNS
3788 39. Added log_selector=+queue_time_overall.
3790 40. When running the queue in the test harness, wait just a tad after forking a
3791 delivery process, to get repeatability of debugging output.
3793 41. Include certificate and key file names in error message when GnuTLS fails
3794 to set them up, because the GnuTLS error message doesn't include the name
3795 of the failing file when there is a problem reading it.
3797 42. Allow both -bf and -bF in the same test run.
3799 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3801 44. The "Exiscan patch" is now merged into the mainline Exim source.
3803 45. Sometimes the final signoff response after QUIT could fail to get
3804 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3805 < 0 before doing a fflush(). This bug looks as though it goes back to the
3806 introduction of TLS in release 3.20, but "sometimes" must have been rare
3807 because the tests only now provoked it.
3809 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3810 (this can affect the format of dates).
3812 47. exim_tidydb, when checking for the continued existence of a message for
3813 which it has found a message-specific retry record, was not finding
3814 messages that were in split spool directories. Consequently, it was
3815 deleting retry records that should have stayed in existence.
3817 48. Steve fixed some bugs in eximstats.
3819 49. The SPA authentication driver was not abandoning authentication and moving
3820 on to the next authenticator when an expansion was forced to fail,
3821 contradicting the general specification for all authenticators. Instead it
3822 was generating a temporary error. It now behaves as specified.
3824 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3825 (the order specifies the preference for clients). The order is now AES256,
3826 AES128, 3DES, ARCFOUR128.
3828 51. Small patch to Sieve code - explicitly set From: when generating an
3831 52. Exim crashed if a remote delivery caused a very long error message to be
3832 recorded - for instance if somebody sent an entire SpamAssassin report back
3833 as a large number of 550 error lines. This bug was coincidentally fixed by
3834 increasing the size of one of Exim's internal buffers (big_buffer) that
3835 happened as part of the Exiscan merge. However, to be on the safe side, I
3836 have made the code more robust (and fixed the comments that describe what
3839 53. Now that there can be additional text after "Completed" in log lines (if
3840 the queue_time_overall log selector is set), a one-byte patch to exigrep
3841 was needed to allow it to recognize "Completed" as not the last thing in
3844 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3845 patch that reportedly fixes this has been added. I am not expert enough to
3846 create a test for it. This is what the patch creator wrote:
3848 "I found a little strange behaviour of ldap code when working with
3849 Windows 2003 AD Domain, where users was placed in more than one
3850 Organization Units. When I tried to give exim partial DN, the exit code
3851 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3852 But simultaneously result of request was absolutely normal ldap result,
3853 so I produce this patch..."
3855 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3856 REFERENCE, so I have modified the code to exclude the patch when that macro
3859 55. Some experimental protocols are using DNS PTR records for new purposes. The
3860 keys for these records are domain names, not reversed IP addresses. The
3861 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3862 leaves it alone. Component reversal etc. now happens only for IP addresses.
3865 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3867 57. Double the size of the debug message buffer (to 2048) so that more of very
3868 long debug lines gets shown.
3870 58. The exicyclog utility now does better if the number of log files to keep
3871 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3873 59. Two changes related to the smtp_active_hostname option:
3875 (1) $smtp_active_hostname is now available as a variable.
3876 (2) The default for smtp_banner uses $smtp_active_hostname instead
3877 of $primary_hostname.
3879 60. The host_aton() function is supposed to be passed a string that is known
3880 to be a valid IP address. However, in the case of IPv6 addresses, it was
3881 not checking this. This is a hostage to fortune. Exim now panics and dies
3882 if the condition is not met. A case was found where this could be provoked
3883 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3884 components; fortuitously, this particular loophole had already been fixed
3885 by change 4.50/55 above.
3887 If there are any other similar loopholes, the new check in host_aton()
3888 itself should stop them being exploited. The report I received stated that
3889 data on the command line could provoke the exploit when Exim was running as
3890 exim, but did not say which command line option was involved. All I could
3891 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3892 running as the user.
3895 61. There was a buffer overflow vulnerability in the SPA authentication code
3896 (which came originally from the Samba project). I have added a test to the
3897 spa_base64_to_bits() function which I hope fixes it.
3900 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3901 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3903 63. The daemon start-up calls getloadavg() while still root for those OS that
3904 need the first call to be done as root, but it missed one case: when
3905 deliver_queue_load_max is set with deliver_drop_privilege. This is
3906 necessary for the benefit of the queue runner, because there is no re-exec
3907 when deliver_drop_privilege is set.
3909 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3910 This has been fixed.
3912 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3913 was in use, was not putting the data itself into the right store pool;
3914 consequently, it could be overwritten for a subsequent message in the same
3915 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3918 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3920 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3921 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3922 the function were treating the return as a boolean value, which happened to
3923 work because 0=false and not-0=true, but is not correct code.
3925 68. The host_aton() function was not handling scoped IPv6 addresses (those
3926 with, for example, "%eth0" on the end) correctly.
3928 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3929 negated items (that is, ~something) in unsigned ints. Some compilers
3930 apparently mutter when there is no cast.
3932 70. If an address verification called from an ACL failed, and did not produce a
3933 user-specific message (i.e. there was only a "system" message), nothing was
3934 put in $acl_verify_message. In this situation, it now puts the system
3937 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3938 session; change 4.31/43 added the unwanted input to the log line - except
3939 that it did not do this in the start of session case. It now does.
3941 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3942 This is wrong and can cause the other end to generate a synchronization
3943 error if it is another Exim or anything else that does the synchronization
3944 check. A QUIT command is no longer sent after a timeout.
3946 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3947 during host lookups.
3949 74. The defer_ok option of callout verification was not working if it was used
3950 when verifying addresses in header lines, that is, for this case:
3952 verify = header_sender/callout=defer_ok
3954 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3955 those file descriptors could be used for SMTP connections. If anything
3956 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3957 could be sent to the SMTP client, causing chaos. The daemon now opens
3958 stdin, stdout, and stderr to /dev/null when it puts itself into the
3961 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3962 log by default. The user can override this with suitable Perl magic.
3964 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3965 the log message when discard triggers, was not working for the DATA ACL or
3966 for the non-SMTP ACL.
3968 78. Error message wording change in sieve.c.
3970 79. If smtp_accept_max_per_host was set, the number of connections could be
3971 restricted to fewer than expected, because the daemon was trying to set up
3972 a new connection before checking whether the processes handling previous
3973 connections had finished. The check for completed processes is now done
3974 earlier. On busy systems, this bug wouldn't be noticed because something
3975 else would have woken the daemon, and it would have reaped the completed
3978 80. If a message was submitted locally by a user whose login name contained one
3979 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3980 It caused a spool format error. I have fixed the spool reading code. A
3981 related problem was that the "from" clause in the Received: line became
3982 illegal because of the space(s). It is now covered by ${quote_local_part.
3984 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3987 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3989 83. Patch from Sieve maintainer to fix unterminated string problem in
3990 "vacation" handling.
3992 84. Some minor changes to the Linux configuration files to help with other
3993 OS variants using glibc.
3995 85. One more patch for Sieve to update vacation handling to latest spec.
3998 ----------------------------------------------------
3999 See the note above about the 4.44 and 4.50 releases.
4000 ----------------------------------------------------
4006 1. Change 4.43/35 introduced a bug that caused file counts to be
4007 incorrectly computed when quota_filecount was set in an appendfile
4010 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4011 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4014 3. Give more explanation in the error message when the command for a transport
4015 filter fails to execute.
4017 4. There are several places where Exim runs a non-Exim command in a
4018 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4019 was being done only for the command run by the queryprogram router. It is
4020 now done for all such subprocesses. The other cases are: ${run, transport
4021 filters, and the commands run by the lmtp and pipe transports.
4023 5. Some older OS have a limit of 256 on the maximum number of file
4024 descriptors. Exim was using setrlimit() to set 1000 as a large value
4025 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4026 systems. I've change it so that if it can't get 1000, it tries for 256.
4028 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4029 was an oversight, and furthermore, ever since the addition of extra
4030 controls (e.g. 4.43/32), the checks on when to allow different forms of
4031 "control" were broken. There should now be diagnostics for all cases when a
4032 control that does not make sense is encountered.
4034 7. $recipients is now available in the predata ACL (oversight).
4036 8. Tidy the search cache before the fork to do a delivery from a message
4037 received from the command line. Otherwise the child will trigger a lookup
4038 failure and thereby defer the delivery if it tries to use (for example) a
4039 cached ldap connection that the parent has called unbind on.
4041 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4042 of $address_data from the recipient verification was clobbered by the
4043 sender verification.
4045 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4046 was its contents. (It was OK if the option was not defined at all.)
4048 11. A "Completed" log line is now written for messages that are removed from
4049 the spool by the -Mrm option.
4051 12. $host_address is now set to the target address during the checking of
4052 ignore_target_hosts.
4054 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4055 being passed; this would have caused $sender_host_name to have been used if
4056 matching the list had actually called for a host name (not very likely,
4057 since this list is usually IP addresses). A host name is now passed as
4060 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4061 code that specifies a non-restarting handler (typically sigaction() in
4062 modern systems) in an attempt to fix a rare and obscure crash bug.
4064 15. Narrowed the window for a race in the daemon that could cause it to ignore
4065 SIGCHLD signals. This is not a major problem, because they are used only to
4066 wake it up if nothing else does.
4068 16. A malformed maildirsize file could cause Exim to calculate negative values
4069 for the mailbox size or file count. Odd effects could occur as a result.
4070 The maildirsize information is now recalculated if the size or filecount
4073 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4074 support for a long time. Removed HAVE_SYS_VFS_H.
4076 18. Updated exipick to current release from John Jetmore.
4078 19. Allow an empty sender to be matched against a lookup in an address list.
4079 Previously the only cases considered were a regular expression, or an
4082 20. Exim went into a mad DNS lookup loop when doing a callout where the
4083 host was specified on the transport, if the DNS lookup yielded more than
4086 21. The RFC2047 encoding function was originally intended for short strings
4087 such as real names; it was not keeping to the 75-character limit for
4088 encoded words that the RFC imposes. It now respects the limit, and
4089 generates multiple encoded words if necessary. To be on the safe side, I
4090 have increased the buffer size for the ${rfc2047: expansion operator from
4093 22. Failure to deliver a bounce message always caused it to be frozen, even if
4094 there was an errors_to setting on the router. The errors_to setting is now
4097 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4098 canonical form (fully expanded) before being placed in
4099 $sender_host_address.
4101 24. Updated eximstats to version 1.33
4103 25. Include certificate and key file names in error message when GnuTLS fails
4104 to set them up, because the GnuTLS error message doesn't include the name
4105 of the failing file when there is a problem reading it.
4107 26. Expand error message when OpenSSL has problems setting up cert/key files.
4110 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4111 (this can affect the format of dates).
4113 28. exim_tidydb, when checking for the continued existence of a message for
4114 which it has found a message-specific retry record, was not finding
4115 messages that were in split spool directories. Consequently, it was
4116 deleting retry records that should have stayed in existence.
4118 29. eximstats updated to version 1.35
4119 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4120 1.35 - bugfix such that pie charts by volume are generated correctly
4122 30. The SPA authentication driver was not abandoning authentication and moving
4123 on to the next authenticator when an expansion was forced to fail,
4124 contradicting the general specification for all authenticators. Instead it
4125 was generating a temporary error. It now behaves as specified.
4127 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4128 (the order specifies the preference for clients). The order is now AES256,
4129 AES128, 3DES, ARCFOUR128.
4131 31. Small patch to Sieve code - explicitly set From: when generating an
4134 32. Exim crashed if a remote delivery caused a very long error message to be
4135 recorded - for instance if somebody sent an entire SpamAssassin report back
4136 as a large number of 550 error lines. This bug was coincidentally fixed by
4137 increasing the size of one of Exim's internal buffers (big_buffer) that
4138 happened as part of the Exiscan merge. However, to be on the safe side, I
4139 have made the code more robust (and fixed the comments that describe what
4142 33. Some experimental protocols are using DNS PTR records for new purposes. The
4143 keys for these records are domain names, not reversed IP addresses. The
4144 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4145 leaves it alone. Component reversal etc. now happens only for IP addresses.
4148 34. The host_aton() function is supposed to be passed a string that is known
4149 to be a valid IP address. However, in the case of IPv6 addresses, it was
4150 not checking this. This is a hostage to fortune. Exim now panics and dies
4151 if the condition is not met. A case was found where this could be provoked
4152 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4153 components; fortuitously, this particular loophole had already been fixed
4154 by change 4.50/55 or 4.44/33 above.
4156 If there are any other similar loopholes, the new check in host_aton()
4157 itself should stop them being exploited. The report I received stated that
4158 data on the command line could provoke the exploit when Exim was running as
4159 exim, but did not say which command line option was involved. All I could
4160 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4161 running as the user.
4164 35. There was a buffer overflow vulnerability in the SPA authentication code
4165 (which came originally from the Samba project). I have added a test to the
4166 spa_base64_to_bits() function which I hope fixes it.
4169 36. The daemon start-up calls getloadavg() while still root for those OS that
4170 need the first call to be done as root, but it missed one case: when
4171 deliver_queue_load_max is set with deliver_drop_privilege. This is
4172 necessary for the benefit of the queue runner, because there is no re-exec
4173 when deliver_drop_privilege is set.
4175 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4176 was in use, was not putting the data itself into the right store pool;
4177 consequently, it could be overwritten for a subsequent message in the same
4178 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4181 38. Sometimes the final signoff response after QUIT could fail to get
4182 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4183 < 0 before doing a fflush(). This bug looks as though it goes back to the
4184 introduction of TLS in release 3.20, but "sometimes" must have been rare
4185 because the tests only now provoked it.
4191 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4192 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4193 Now it returns an integer. A number of calls were still expecting a T/F
4194 return. Fortuitously, in all cases, the tests worked in OK situations,
4195 which is the norm. However, things would have gone wrong on any write
4196 failures on the smtp file descriptor. This function is used when sending
4197 messages over SMTP and also when doing verify callouts.
4199 2. When Exim is called to do synchronous delivery of a locally submitted
4200 message (the -odf or -odi options), it no longer closes stderr before doing
4203 3. Implemented the mua_wrapper option.
4205 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4207 5. Implemented the functions header_remove(), header_testname(),
4208 header_add_at_position(), and receive_remove_recipient(), and exported them
4211 6. If an ACL "warn" statement specified the addition of headers, Exim already
4212 inserted X-ACL-Warn: at the start if there was no header name. However, it
4213 was not making this test for the second and subsequent header lines if
4214 there were newlines in the string. This meant that an invalid header could
4215 be inserted if Exim was badly configured.
4217 7. Allow an ACL "warn" statement to add header lines at the start or after all
4218 the Received: headers, as well as at the end.
4220 8. Added the rcpt_4xx retry error code.
4222 9. Added postmaster_mailfrom=xxx to callout verification option.
4224 10. Added mailfrom=xxxx to the callout verification option, for verify=
4227 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4228 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4230 12. Inserted some casts to stop certain compilers warning when using pointer
4231 differences as field lengths or precisions in printf-type calls (mostly
4232 affecting debugging statements).
4234 13. Added optional readline() support for -be (dynamically loaded).
4236 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4237 same clock tick as a message's arrival, so that its received time was the
4238 same as the "first fail" time on the retry record, and that message
4239 remained on the queue past the ultimate address timeout, every queue runner
4240 would try a delivery (because it was past the ultimate address timeout) but
4241 after another failure, the ultimate address timeout, which should have then
4242 bounced the address, did not kick in. This was a "< instead of <=" error;
4243 in most cases the first failure would have been in the next clock tick
4244 after the received time, and all would be well.
4246 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4247 being recognized when the domain list was tested by the match_domain
4248 condition in an expansion string.
4250 16. Added the ${str2b64: operator.
4252 17. Exim was always calling setrlimit() to set a large limit for the number of
4253 processes, without checking whether the existing limit was already
4254 adequate. (It did check for the limit on file descriptors.) Furthermore,
4255 errors from getrlimit() and setrlimit() were being ignored. Now they are
4256 logged to the main and panic logs, but Exim does carry on, to try to do its
4257 job under whatever limits there are.
4259 18. Imported PCRE 5.0.
4261 19. Trivial typo in log message " temporarily refused connection" (the leading
4264 20. If the log selector return_path_on_delivery was set and an address was
4265 redirected to /dev/null, the delivery process crashed because it assumed
4266 that a return path would always be set for a "successful" delivery. In this
4267 case, the whole delivery is bypassed as an optimization, and therefore no
4270 21. Internal re-arrangement: the function for sending a challenge and reading
4271 a response while authentication was assuming a zero-terminated challenge
4272 string. It's now changed to take a pointer and a length, to allow for
4273 binary data in such strings.
4275 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4277 23. Exim was not respecting finduser_retries when seeking the login of the
4278 uid under which it was called; it was always trying 10 times. (The default
4279 setting of finduser_retries is zero.) Also, it was sleeping after the final
4280 failure, which is pointless.
4282 24. Implemented tls_on_connect_ports.
4284 25. Implemented acl_smtp_predata.
4286 26. If the domain in control=submission is set empty, Exim assumes that the
4287 authenticated id is a complete email address when it generates From: or
4288 Sender: header lines.
4290 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4291 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4292 chown and chgrp in /bin and hostname in /usr/bin.
4294 28. Exim was keeping the "process log" file open after each use, just as it
4295 does for the main log. This opens the possibility of it remaining open for
4296 long periods when the USR1 signal hits a daemon. Occasional processlog
4297 errors were reported, that could have been caused by this. Anyway, it seems
4298 much more sensible not to leave this file open at all, so that is what now
4301 29. The long-running daemon process does not normally write to the log once it
4302 has entered its main loop, and it closes the log before doing so. This is
4303 so that log files can straightforwardly be renamed and moved. However,
4304 there are a couple of unusual error situations where the daemon does write
4305 log entries, and I had neglected to close the log afterwards.
4307 30. The text of an SMTP error response that was received during a remote
4308 delivery was being truncated at 512 bytes. This is too short for some of
4309 the long messages that one sometimes sees. I've increased the limit to
4312 31. It is now possible to make retry rules that apply only when a message has a
4313 specific sender, in particular, an empty sender.
4315 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4316 it possible to be selective about when SMTP synchronization is enforced.
4318 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4320 32. Implemented hosts_connection_nolog.
4322 33. Added an ACL for QUIT.
4324 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4327 35. Added mailbox_size and mailbox_filecount to appendfile.
4329 36. Added control = no_multiline_responses to ACLs.
4331 37. There was a bug in the logic of the code that waits for the clock to tick
4332 in the case where the clock went backwards by a substantial amount such
4333 that the microsecond fraction of "now" was more than the microsecond
4334 fraction of "then" (but the whole seconds number was less).
4336 38. Added support for the libradius Radius client library this is found on
4337 FreeBSD (previously only the radiusclient library was supported).
4343 1. When certain lookups returned multiple values in the form name=value, the
4344 quoting of the values was not always being done properly. Specifically:
4345 (a) If the value started with a double quote, but contained no whitespace,
4347 (b) If the value contained whitespace other than a space character (i.e.
4348 tabs or newlines or carriage returns) it was not quoted.
4349 This fix has been applied to the mysql and pgsql lookups by writing a
4350 separate quoting function and calling it from the lookup code. The fix
4351 should probably also be applied to nisplus, ibase and oracle lookups, but
4352 since I cannot test any of those, I have not disturbed their existing code.
4354 2. A hit in the callout cache for a specific address caused a log line with no
4355 reason for rejecting RCPT. Now it says "Previous (cached) callout
4356 verification failure".
4358 3. There was an off-by-one bug in the queryprogram router. An over-long
4359 return line was truncated at 256 instead of 255 characters, thereby
4360 overflowing its buffer with the terminating zero. As well as fixing this, I
4361 have increased the buffer size to 1024 (and made a note to document this).
4363 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4364 when Exim is waiting for an SMTP response from a remote server, Exim
4365 restarts its select() call on the socket, thereby resetting its timeout.
4366 This is not a problem when such interrupts are rare. Somebody set up a cron
4367 job to run exiwhat every 2 minutes, which is less than the normal select()
4368 timeout (5 or 10 minutes). This meant that the select() timeout never
4369 kicked in because it was always reset. I have fixed this by comparing the
4370 time when an interrupt arrives with the time at the start of the first call
4371 to select(). If more time than the timeout has elapsed, the interrupt is
4372 treated as a timeout.
4374 5. Some internal re-factoring in preparation for the addition of Sieve
4375 extensions (by MH). In particular, the "personal" test is moved to a
4376 separate function, and given an option for scanning Cc: and Bcc: (which is
4377 not set for Exim filters).
4379 6. When Exim created an email address using the login of the caller as the
4380 local part (e.g. when creating a From: or Sender: header line), it was not
4381 quoting the local part when it contained special characters such as @.
4383 7. Installed new OpenBSD configuration files.
4385 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4386 try to make them clearer.
4388 9. Callout options, other than the timeout value, were being ignored when
4389 verifying sender addresses in header lines. For example, when using
4391 verify = header_sender/callout=no_cache
4393 the cache was (incorrectly) being used.
4395 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4396 only the Cygwin environment.
4398 11. When return_path_on_delivery was set as a log selector, if different remote
4399 addresses in the same message used different return paths and parallel
4400 remote delivery occurred, the wrong values would sometimes be logged.
4401 (Whenever a remote delivery process finished, the return path value from
4402 the most recently started remote delivery process was logged.)
4404 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4405 lines when AUTH and/or TLS are in use. This is the "received protocol"
4406 field. Exim used to use "asmtp" for authenticated SMTP, without any
4407 indication (in the protocol name) for TLS use. Now it follows the RFC and
4408 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4409 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4410 names appear in log lines as well as in Received: header lines.
4412 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4413 extensions, and comparison tests, and to fix some bugs.
4415 14. Changes to the "personal" filter test:
4417 (1) The test was buggy in that it was just doing the equivalent of
4418 "contains" tests on header lines. For example, if a user's address was
4419 anne@some.where, the "personal" test would incorrectly be true for
4421 To: susanne@some.where
4423 This test is now done by extracting each address from the header in turn,
4424 and checking the entire address. Other tests that are part of "personal"
4425 are now done using regular expressions (for example, to check local parts
4426 of addresses in From: header lines).
4428 (2) The list of non-personal local parts in From: addresses has been
4429 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4430 taken from the Sieve specification recommendations.
4432 (3) If the message contains any header line starting with "List-" it is
4433 treated as non-personal.
4435 (4) The test for "circular" in the Subject: header line has been removed
4436 because it now seems ill-conceived.
4438 15. Minor typos in src/EDITME comments corrected.
4440 16. Installed latest exipick from John Jetmore.
4442 17. If headers_add on a router specified a text string that was too long for
4443 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4444 of string_sprintf() is now avoided.
4446 18. $message_body_size was not set (it was always zero) when running the DATA
4447 ACL and the local_scan() function.
4449 19. For the "mail" command in an Exim filter, no default was being set for
4450 the once_repeat time, causing a random time value to be used if "once" was
4451 specified. (If the value happened to be <= 0, no repeat happened.) The
4452 default is now 0s, meaning "never repeat". The "vacation" command was OK
4453 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4454 (I found it when inspecting the code).
4456 20. There is now an overall timeout for performing a callout verification. It
4457 defaults to 4 times the callout timeout, which applies to individual SMTP
4458 commands during the callout. The overall timeout applies when there is more
4459 than one host that can be tried. The timeout is checked before trying the
4460 next host. This prevents very long delays if there are a large number of
4461 hosts and all are timing out (e.g. when the network connections are timing
4462 out). The value of the overall timeout can be changed by specifying an
4463 additional sub-option for "callout", called "maxwait". For example:
4465 verify = sender/callout=5s,maxwait=20s
4467 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4468 to the end before writing, but this should make it even safer).
4470 22. Exim was forgetting that it had advertised PIPELINING for the second and
4471 subsequent messages on an SMTP connection. It was also not resetting its
4472 memory on STARTTLS and an internal HELO.
4474 23. When Exim logs an SMTP synchronization error within a session, it now
4475 records whether PIPELINING has been advertised or not.
4477 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4478 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4479 rather than long int.
4481 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4483 26. Added the never_mail option to autoreply.
4489 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4490 crash if the getsockname() call failed; this can happen if a connection is
4491 closed very soon after it is established. The problem was simply in the
4492 order in which certain operations were done, causing Exim to try to write
4493 to the SMTP stream before it had set up the file descriptor. The bug has
4494 been fixed by making things happen in the correct order.
4500 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4501 before the connection was closed, thus losing the rejection response.
4503 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4504 some early Solaris releases, but causes trouble in current releases where
4505 socklen_t is defined.
4507 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4510 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4513 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4514 that releasing the top of it at the end releases what was used for sub-
4515 expansions (unless the block got too big). However, discard this block if
4516 the first thing is a variable or header, so that we can use its block when
4517 it is dynamic (useful for very large $message_headers, for example).
4519 6. Lookups now cache *every* query, not just the most recent. A new, separate
4520 store pool is used for this. It can be recovered when all lookup caches are
4521 flushed. Lookups now release memory at the end of their result strings.
4522 This has involved some general refactoring of the lookup sources.
4524 7. Some code has been added to the store_xxx() functions to reduce the amount
4525 of flapping under certain conditions.
4527 8. log_incoming_interface used to affect only the <= reception log lines. Now
4528 it causes the local interface and port to be added to several more SMTP log
4529 lines, for example "SMTP connection from", and rejection lines.
4531 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4533 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4535 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4536 could be overwritten at the end of the current message (or the start of a
4537 new message if it was set in a HELO ACL). The value is now preserved for
4538 the duration of the SMTP connection.
4540 12. If a transport had a headers_rewrite setting, and a matching header line
4541 contained an unqualified address, that address was qualified, even if it
4542 did not match any rewriting rules. The underlying bug was that the values
4543 of the flags that permit the existence of unqualified sender and recipient
4544 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4545 non-local messages, and by -bnq for local messages) were not being
4546 preserved with the message after it was received.
4548 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4549 "next input=" as part of the text comprising the host identity instead of
4550 the correct text. The code was using the same buffer for two different
4551 strings. However, depending on which order the printing function evaluated
4552 its arguments, the bug did not always show up. Under Linux, for example, my
4553 test suite worked just fine.
4555 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4556 doesn't work with some older versions of Perl. It has been changed to "my",
4557 which in any case is probably the better facility to use.
4559 15. A really picky compiler found some instances of statements for creating
4560 error messages that either had too many or two few arguments for the format
4563 16. The size of the buffer for calls to the DNS resolver has been increased
4564 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4565 for addresses that have a lot of PTR records. This alleviates a problem; it
4566 does not fully solve it.
4568 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4569 buffer now truncates the list and logs the incident, which is the same
4570 action as happens when Exim is looking up a host name and its aliases.
4571 Previously in this situation something unpredictable would happen;
4572 sometimes it was "internal error: store_reset failed".
4574 18. If a server dropped the connection unexpectedly when an Exim client was
4575 using GnuTLS and trying to read a response, the client delivery process
4576 crashed while trying to generate an error log message.
4578 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4579 string, for example:
4581 warn message = H1: something\nH2: something
4583 the text was added as a single header line from Exim's point of view
4584 though it ended up OK in the delivered message. However, searching for the
4585 second and subsequent header lines using $h_h2: did not work. This has been
4586 fixed. Similarly, if a system filter added multiple headers in this way,
4587 the routers could not see them.
4589 20. Expanded the error message when iplsearch is called with an invalid key to
4590 suggest using net-iplsearch in a host list.
4592 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4593 ACLs are no longer actually imposed (and a message to that effect is
4596 22. If a "gecos" field in a passwd entry contained escaped characters, in
4597 particular, if it contained a \" sequence, Exim got it wrong when building
4598 a From: or a Sender: header from that name. A second bug also caused
4599 incorrect handling when an unquoted " was present following a character
4600 that needed quoting.
4602 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4603 was not being matched caselessly.
4605 24. Arranged for all hyphens in the exim.8 source to be escaped with
4608 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4609 properly. Recipient callout cache records were still being keyed to include
4610 the sender, even when use_sender was set false. This led to far more
4611 callouts that were necessary. The sender is no longer included in the key
4612 when use_sender is false.
4614 26. Added "control = submission" modifier to ACLs.
4616 27. Added the ${base62d: operator to decode base 62 numbers.
4618 28. dnsdb lookups can now access SRV records.
4620 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4621 the configuration file.
4623 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4624 (-v) mode. This makes the output for a verbose queue run more intelligible.
4626 31. Added a use_postmaster feature to recipient callouts.
4628 32. Added the $body_zerocount variable, containing the number of binary zero
4629 bytes in the message body.
4631 33. The time of last modification of the "new" subdirectory is now used as the
4632 "mailbox time last read" when there is a quota error for a maildir
4635 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4637 35. Added +ignore_unknown as a special item in host lists.
4639 36. Code for decoding IPv6 addresses in host lists is now included, even if
4640 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4641 address was recognized as an IP address, but was then not correctly decoded
4642 into binary, causing unexpected and incorrect effects when compared with
4649 1. Very minor rewording of debugging text in manualroute to say "list of
4650 hosts" instead of "hostlist".
4652 2. If verify=header_syntax was set, and a header line with an unqualified
4653 address (no domain) and a large number of spaces between the end of the
4654 name and the colon was received, the reception process suffered a buffer
4655 overflow, and (when I tested it) crashed. This was caused by some obsolete
4656 code that should have been removed. The fix is to remove it!
4658 3. When running in the test harness, delay a bit after writing a bounce
4659 message to get a bit more predictability in the log output.
4661 4. Added a call to search_tidyup() just before forking a reception process. In
4662 theory, someone could use a lookup in the expansion of smtp_accept_max_
4663 per_host which, without the tidyup, could leave open a database connection.
4665 5. Added the variables $recipient_data and $sender_data which get set from a
4666 lookup success in an ACL "recipients" or "senders" condition, or a router
4667 "senders" option, similar to $domain_data and $local_part_data.
4669 6. Moved the writing of debug_print from before to after the "senders" test
4672 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4673 problems for message scanning, either using a data ACL, or using
4674 local_scan() because the Received: header was not generated till after they
4675 were called (in order to set the time as the time of reception completion).
4676 I have revised the way this works. The header is now generated after the
4677 body is received, but before the ACL or local_scan() are called. After they
4678 are run, the timestamp in the header is updated.
4684 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4685 before starting a queue runner without re-exec. This happened only when
4686 deliver_drop_privilege was set or when the Exim user was set to root. The
4687 effect of the bug was that timeouts during subsequent deliveries caused
4688 crashes instead of being properly handled. The handler is now left at its
4689 default (and expected) setting.
4691 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4692 message, again when deliver_drop_privilege is set or Exim is run as root.
4693 The bug described in (1) was not present in this case, but the tidying up
4694 of the other signals was missing. I have made the two cases consistent.
4696 3. The ignore_target_hosts setting on a manualroute router was being ignored
4697 for hosts that were looked up using the /MX notation.
4699 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4702 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4703 operated on the sender address. After changing the $sender_address to <>
4704 for the sender address verify, Exim was re-instated it as the original
4705 (before rewriting) address, but remembering that it had rewritten it, so it
4706 wasn't rewriting it again. This bug also had the effect of breaking the
4707 sender address verification caching when the sender address was rewritten.
4709 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4710 This has been changed so that if the ip literal address matches
4711 ignore_target_hosts, the router declines.
4713 7. Added expansion conditions match_domain, match_address, and match_local_
4714 part (NOT match_host).
4716 8. The placeholder for the Received: header didn't have a length field set.
4718 9. Added code to Exim itself and to exim_lock to test for a specific race
4719 condition that could lead to file corruption when using MBX delivery. The
4720 issue is with the lockfile that is created in /tmp. If this file is removed
4721 after a process has opened it but before that process has acquired a lock,
4722 there is the potential for a second process to recreate the file and also
4723 acquire a lock. This could lead to two Exim processes writing to the file
4724 at the same time. The added code performs the same test as UW imapd; it
4725 checks after acquiring the lock that its file descriptor still refers to
4726 the same named file.
4728 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4729 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4732 11. Added the smtp_active_hostname option. If used, this will typically be made
4733 to depend on the incoming interface address. Because $interface_address is
4734 not set up until the daemon has forked a reception process, error responses
4735 that can happen earlier (such as "too many connections") no longer contain
4738 12. If an expansion in a condition on a "warn" statement fails because a lookup
4739 defers, the "warn" statement is abandoned, and the next ACL statement is
4740 processed. Previously this caused the whole ACL to be aborted.
4742 13. Added the iplsearch lookup type.
4744 14. Added ident_timeout as a log selector.
4746 15. Added tls_certificate_verified as a log selector.
4748 16. Added a global option tls_require_ciphers (compare the smtp transport
4749 option of the same name). This controls incoming TLS connections.
4751 17. I finally figured out how to make tls_require_ciphers do a similar thing
4752 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4753 before starting the TLS session.
4755 18. Tabs are now shown as \t in -bP output.
4757 19. If the log selector return_path_on_delivery was set, Exim crashed when
4758 bouncing a message because it had too many Received: header lines.
4760 20. If two routers both had headers_remove settings, and the first one included
4761 a superfluous trailing colon, the final name in the first list and the
4762 first name in the second list were incorrectly joined into one item (with a
4763 colon in the middle).
4769 1. Added -C and -D options to the exinext utility, mainly to make it easier
4770 to include in the automated testing, but these could be helpful when
4771 multiple configurations are in use.
4773 2. The exinext utility was not formatting the output nicely when there was
4774 an alternate port involved in the retry record key, nor when there was a
4775 message id as well (for retries that were specific to a specific message
4776 and a specific host). It was also confused by IPv6 addresses, because of
4777 the additional colons they contain. I have fixed the IPv4 problem, and
4778 patched it up to do a reasonable job for IPv6.
4780 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4781 delivery, the log line now contains "pipelined" if PIPELINING was used.
4783 4. An SMTP transport process used to panic and die if the bind() call to set
4784 an explicit outgoing interface failed. This has been changed; it is now
4785 treated in the same way as a connect() failure.
4787 5. A reference to $sender_host_name in the part of a conditional expansion
4788 that was being skipped was still causing a DNS lookup. This no longer
4791 6. The def: expansion condition was not recognizing references to header lines
4792 that used bh_ and bheader_.
4794 7. Added the _cache feature to named lists.
4796 8. The code for checking quota_filecount in the appendfile transport was
4797 allowing one more file than it should have been.
4799 9. For compatibility with Sendmail, the command line option
4807 and sets the incoming protocol and host name (for trusted callers). The
4808 host name and its colon can be omitted when only the protocol is to be set.
4809 Note the Exim already has two private options, -pd and -ps, that refer to
4810 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4811 "s", but I don't think that's a major issue.
4813 10. A number of refactoring changes to the code, none of which should affect
4816 (a) The number of logging options was getting close to filling up the
4817 32-bit word that was used as a bit map. I have split them into two classes:
4818 those that are passed in the argument to log_write(), and those that are
4819 only ever tested independently outside of that function. These are now in
4820 separate 32-bit words, so there is plenty of room for expansion again.
4821 There is no change in the user interface or the logging behaviour.
4823 (b) When building, for example, log lines, the code previously used a
4824 macro that called string_cat() twice, in order to add two strings. This is
4825 not really sufficiently general. Furthermore, there was one instance where
4826 it was actually wrong because one of the argument was used twice, and in
4827 one call a function was used. (As it happened, calling the function twice
4828 did not affect the overall behaviour.) The macro has been replaced by a
4829 function that can join an arbitrary number of extra strings onto a growing
4832 (c) The code for expansion conditions now uses a table and a binary chop
4833 instead of a serial search (which was left over from when there were very
4834 few conditions). Also, it now recognizes conditions like "pam" even when
4835 the relevant support is not compiled in: a suitably worded error message is
4836 given if an attempt is made to use such a condition.
4838 11. Added ${time_interval:xxxxx}.
4840 12. A bug was causing one of the ddress fields not to be passed back correctly
4841 from remote delivery subprocesses. The field in question was not being
4842 subsequently used, so this caused to problems in practice.
4844 13. Added new log selectors queue_time and deliver_time.
4846 14. Might have fixed a bug in maildirsizefile handling that threw up
4847 "unexpected character" debug warnings, and recalculated the data
4848 unnecessarily. In any case, I expanded the warning message to give more
4851 15. Added the message "Restricted characters in address" to the statements in
4852 the default ACL that block characters like @ and % in local parts.
4854 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4855 Three changes have been made:
4857 (a) There was a serious bug; a negative response to MAIL caused the whole
4858 recipient domain to be cached as invalid, thereby blocking all messages
4859 to all local parts at the same domain, from all senders. This bug has
4860 been fixed. The domain is no longer cached after a negative response to
4861 MAIL if the sender used is not empty.
4863 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4866 (c) A new callout option, "use_sender" has been added for people who want
4867 the modified behaviour.
4873 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4876 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4877 indeed breaks things for older releases.
4879 3. Added additional logging to the case where there is a problem reading data
4880 from a filter that is running in a subprocess using a pipe, in order to
4881 try to track down a specific problem.
4883 4. Testing facility fudge: when running in the test harness and attempting
4884 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4885 getting "No route to host". Convert this to a timeout.
4887 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4890 6. Some OS don't have socklen_t but use size_t instead. This affects the
4891 fifth argument of getsockopt() amongst other things. This is now
4892 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4893 can be set for individual OS. I have set it for SunOS5, OSF1, and
4894 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4895 some earlier ones do not.
4897 7. Change 4.30/15 was not doing the test caselessly.
4899 8. The standard form for an IPv6 address literal was being rejected by address
4900 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4901 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4902 this, as well as the form without the "IPv6" on the front (but only when
4903 address literals are enabled, of course).
4905 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4907 10. Exim crashed if a message with an empty sender address specified by -f
4908 encountered a router with an errors_to setting. This could be provoked only
4909 by a command such as
4913 where an empty string was supplied; "<>" did not hit this bug.
4915 11. Installed PCRE release 4.5.
4917 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4918 remained set. It is now erased.
4920 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4921 times from message ids (which are base 36 rather than the normal 62).
4923 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4924 were being counted as actual protocol errors, and logged if the log
4925 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4926 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4927 and DATA following a set of rejected RCPTs do not count as protocol errors.
4928 In other words, Exim assumes they were pipelined, though this may not
4929 actually be the case. Of course, in all cases the client gets an
4930 appropriate error code.
4932 15. If a lookup fails in an ACL condition, a message about the failure may
4933 be available; it is used if testing the ACL cannot continue, because most
4934 such messages specify what the cause of the deferral is. However, some
4935 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4936 that caused an old message to be retained and used if a later statement
4937 caused a defer, replacing the real cause of the deferral.
4939 16. If an IP address had so many PTR records that the DNS lookup buffer
4940 was not large enough to hold them, Exim could crash while trying to process
4941 the truncated data. It now detects and logs this case.
4943 17. Further to 4.21/58, another change has been made: if (and only if) the
4944 first line of a message (the first header line) ends with CRLF, a bare LF
4945 in a subsequent header line has a space inserted after it, so as not to
4946 terminate the header.
4948 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4949 unnecessarily, used atoi() instead of strtol(), and didn't check the
4950 termination when getting file sizes from file names by regex.
4952 19. Completely re-implemented the support for maildirsize files, in the light
4953 of a number of problems with the previous contributed implementation
4954 (4.30/29). In particular:
4956 . If the quota is zero, the maildirsize file is maintained, but no quota is
4959 . If the maildir directory does not exist, it is created before any attempt
4960 to write a maildirsize file.
4962 . The quota value in the file is just a cache; if the quota is changed in
4963 the transport, the new value overrides.
4965 . A regular expression is available for excluding directories from the
4968 20. The autoreply transport checks the characters in options that define the
4969 message's headers; it allows continued headers, but it was checking with
4970 isspace() after an embedded newline instead of explicitly looking for a
4973 21. If all the "regular" hosts to which an address was routed had passed their
4974 expiry times, and had not reached their retry times, the address was
4975 bounced, even if fallback hosts were defined. Now Exim should go on to try
4978 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4979 equivalent code in the SMTP transport. Some hosts send humungous responses
4980 to HELO/EHLO, more than 1024 it seems.
4982 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4983 gives compiler warnings in some environments. I've now done it "properly",
4986 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4987 (because of problems with the built-in one) was declared to return uschar *
4988 instead of char *, causing compiler failure.
4990 25. Fixed a file descriptor leak when processing alias/forward files.
4992 26. Fixed a minor format string issue in dbfn.c.
4994 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4996 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4997 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5000 29. When a sender address is verified, it is cached, to save repeating the test
5001 when there is more than one recipient in a message. However, when the
5002 verification involves a callout, it is possible for different callout
5003 options to be set for different recipients. It is too complicated to keep
5004 track of this in the cache, so now Exim always runs a verification when a
5005 callout is required, relying on the callout cache for the optimization.
5006 The overhead is duplication of the address routing, but this should not be
5009 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5010 to be verified with callout=postmaster, and the main callout worked but the
5011 postmaster check failed, the verification correctly failed. However, if a
5012 subsequent RCPT command asked for sender verification *without* the
5013 postmaster check, incorrect caching caused this verification also to fail,
5016 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5017 it was not caching the DNS options (qualify_single, search_parents) that
5018 were used when the lookup failed. A subsequent lookup with different
5019 options therefore always gave the same answer, though there were cases
5020 where it should not have. (Example: a "domains = !$mx_any" option on a
5021 dnslookup router: the "domains" option is always processed without any
5022 widening, but the router might have qualify_single set.) Now Exim uses the
5023 cached value only when the same options are set.
5025 32. Added John Jetmore's "exipick" utility to the distribution.
5027 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5028 than a timeout (e.g. a certificate is required, and is not provided), an
5029 Exim server now closes the connection immediately. Previously it waited for
5030 the client to close - but if the client is SSL, it seems that they each
5031 wait for each other, leading to a delay before one of them times out.
5033 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5034 maintained 0.8.x compatibility because I don't think many are using it, and
5035 it is clearly obsolete.
5037 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5040 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5041 client certificate was expired. A simple patch fixes this, though I don't
5042 understand the full logic of why the verify callback is called multiple
5045 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5046 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5047 which causes problems with some clients (such as the Certicom SSL Plus
5048 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5049 disables the coutermeasure allowing Eudora to connect."
5051 38. Exim was not checking that a write() to a log file succeeded. This could
5052 lead to Bad Things if a log got too big, in particular if it hit a file
5053 size limit. Exim now panics and dies if it cannot write to a log file, just
5054 as it does if it cannot open a log file.
5056 39. Modified OS/Makefile-Linux so that it now contains
5058 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5060 The two -D definitions ensure that Exim is compiled with large file
5061 support, which makes it possible to handle log files that are bigger than
5064 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5065 instance) a domain was checked against a named list that involved a lookup,
5066 causing $domain_data to be set, then another domain was checked against the
5067 same list, then the first domain was re-checked, the value of $domain_data
5068 after the final check could be wrong. In particular, if the second check
5069 failed, it could be set empty. This bug probably also applied to
5072 41. The strip_trailing_dot option was not being applied to the address given
5073 with the -f command-line option.
5075 42. The code for reading a message's header from the spool was incrementing
5076 $received_count, but never initializing it. This meant that the value was
5077 incorrect (doubled) while delivering a message in the same process in which
5078 it was received. In the most common configuration of Exim, this never
5079 happens - a fresh exec is done - but it can happen when
5080 deliver_drop_privilege is set.
5082 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5083 it now includes up to 150 characters of the unexpected data in the log
5086 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5087 and building data strings. The size of both of these buffers was 10 000
5088 bytes - far larger than anybody would *ever* want, thought I. Needless to
5089 say, somebody hit the limit. I have increased the maximum line length to
5090 20 000 and the maximum data length of concatenated lines to 100 000. I have
5091 also fixed two bugs, because there was no checking on these buffers. Tsk,
5092 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5093 buffer is too small.
5095 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5096 lsearch lookups. Now it does.
5098 46. When parsing a route_list item in a manualroute router, a fixed-length
5099 buffer was used for the list of hosts. I made this 1024 bytes long,
5100 thinking that nobody would ever have a list of hosts that long. Wrong.
5101 Somebody had a whole pile of complicated expansion conditions, and the
5102 string was silently truncated, leading to an expansion error. It turns out
5103 that it is easier to change to an unlimited length (owing to other changes
5104 that have happened since this code was originally written) than to build
5105 structure for giving a limitation error. The length of the item that
5106 expands into the list of hosts is now unlimited.
5108 47. The lsearch lookup could not handle data where the length of text line was
5109 more than 4095 characters. Such lines were truncated, leading to shortened
5110 data being returned. It should now handle lines of any length.
5112 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5113 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5116 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5117 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5118 no longer happens in comment lines. A list of replacements is now placed
5119 at the head of all of the source files, except those whose only change is
5120 to replace PERL_COMMAND in the very first #! line.
5122 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5123 messages on the queue, with a bottom-up merge sort, using code contributed
5124 by Michael Haardt. This should make operations like -bp somewhat faster on
5125 large queues. It won't affect queue runners, except when queue_run_in_order
5128 51. Installed eximstats 1.31 in the distribution.
5130 52. Added support for SRV lookups to the dnslookup router.
5132 53. If an ACL referred to $message_body or $message_body_end, the value was not
5133 reset for any messages that followed in the same SMTP session.
5135 54. The store-handling optimization for building very long strings was not
5136 differentiating between the different store pools. I don't think this
5137 actually made any difference in practice, but I've tidied it.
5139 55. While running the routers to verify a sender address, $sender_address
5140 was still set to the sender address. This is wrong, because when routing to
5141 send a bounce to the sender, it would be empty. Therefore, I have changed
5142 it so that, while verifying a sender address, $sender_address is set to <>.
5143 (There is no change to what happens when verifying a recipient address.)
5145 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5146 target A or AAAA records (if not already returned) without resetting the
5147 qualify_single or search_parents options of the DNS resolver. These are
5148 inappropriate in this case because the targets of MX and SRV records must
5149 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5150 target that, when qualified, matched something in the local domain. These
5151 two options are now turned off when doing these lookups.
5153 57. It seems that at least some releases of Reiserfs (which does not have the
5154 concept of a fixed number of inodes) returns zero and not -1 for the
5155 number of available inodes. This interacted badly with check_spool_inodes,
5156 which assumed that -1 was the "no such thing" setting. What I have done is
5157 to check that the total number of inodes is greater than zero before doing
5158 the test of how many are available.
5160 58. When a "warn" ACL statement has a log_message modifier, the message is
5161 remembered, and not repeated. This is to avoid a lot of repetition when a
5162 message has many recipients that cause the same warning to be written.
5163 Howewer, Exim was preserving the list of already written lines for an
5164 entire SMTP session, which doesn't seem right. The memory is now reset if a
5165 new message is started.
5167 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5168 debugging output unless log_rewrite was also set.
5170 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5171 of (int)(handle) when we know that handle contains (void *)(-1).
5173 61. The Exim daemon panic-logs an error return when it closes the incoming
5174 connection. However "connection reset by peer" seems to be common, and
5175 isn't really an error worthy of noting specially, so that particular error
5178 62. When Exim is trying to find all the local interfaces, it used to panic and
5179 die if the ioctl to get the interface flags failed. However, it seems that
5180 on at least one OS (Solaris 9) it is possible to have an interface that is
5181 included in the list of interfaces, but for which you get a failure error
5182 for this call. This happens when the interface is not "plumbed" into a
5183 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5184 failure of the "get flags" call assumes that the interface is down.
5186 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5187 makes life easier for people who are doing arithmetic on fields extracted
5188 from dates, where you often get leading zeros that should not be
5189 interpreted as octal.
5191 64. Added qualify_domain to the redirect router, to override the global
5194 65. If a pathologically long header line contained very many addresses (the
5195 report of this problem mentioned 10 000) and each of them was rewritten,
5196 Exim could use up a very large amount of memory. (It kept on making new
5197 copies of the header line as it rewrote, and never released the old ones.)
5198 At the expense of a bit more processing, the header rewriting function has
5199 been changed so that it no longer eats memory in this way.
5201 66. The generation of the Received: header has been moved from the time that a
5202 message starts to be received, to the time that it finishes. The timestamp
5203 in the Received: header should now be very close to that of the <= log
5204 line. There are two side-effects of this change:
5206 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5207 logged header lines no longer include the local Received: line, because
5208 it has not yet been created. The same applies to a copy of the message
5209 that is returned to a non-SMTP sender when a message is rejected.
5211 (b) When a filter file is tested using -bf, no additional Received: header
5212 is added to the test message. After some thought, I decided that this
5215 This change does not affect the value of $received_for. It is still set
5216 after address rewriting, but before local_scan() is called.
5218 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5220 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5221 gave an unhelpful panic error message, and a defer error. I have managed to
5222 change this behaviour so that it now rejects any supplied certificate,
5223 which seems right, as the list of acceptable certificates is empty.
5225 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5226 gave an unhelpful defer error. I have not managed to make this reject any
5227 supplied certificates, but the error message it gives is "no certificate
5228 supplied", which is not helpful.
5230 70. exigrep's output now also includes lines that are not associated with any
5231 message, but which match the given pattern. Implemented by a patch from
5232 Martin Sluka, which also tidied up the Perl a bit.
5234 71. Recipient callout verification, like sender verification, was using <> in
5235 the MAIL FROM command. This isn't really the right thing, since the actual
5236 sender may affect whether the remote host accepts the recipient or not. I
5237 have changed it to use the actual sender in the callout; this means that
5238 the cache record is now keyed on a recipient/sender pair, not just the
5239 recipient address. There doesn't seem to be a real danger of callout loops,
5240 since a callout by the remote host to check the sender would use <>.
5241 [SEE ABOVE: changed after hitting problems.]
5243 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5244 temporary errors. However, in the case of such a code being given after
5245 the end of a data transmission (i.e. after ".") Exim was failing to write
5246 a retry record for the message. (Yes, there was some broken host that was
5247 actually sending 8xx at this point.)
5249 73. An unknown lookup type in a host list could cause Exim to panic-die when
5250 the list was checked. (An example that provoked this was putting <; in the
5251 middle of a list instead of at the start.) If this happened during a DATA
5252 ACL check, a -D file could be left lying around. This kind of configuration
5253 error no longer causes Exim to die; instead it causes a defer errror. The
5254 incident is still logged to the main and panic logs.
5256 74. Buglet left over from Exim 3 conversion. The message "too many messages
5257 in one connection" was written to the rejectlog but not the mainlog, except
5258 when address rewriting (yes!) was being logged.
5260 75. Added write_rejectlog option.
5262 76. When a system filter was run not as root (that is, when system_filter_user
5263 was set), the values of the $n variables were not being returned to the
5264 main process; thus, they were not subsequently available in the $sn
5267 77. Added +return_path_on_delivery log selector.
5269 78. A connection timeout was being treated differently from recipients deferred
5270 when testing hosts_max_try with a message that was older than the host's
5271 retry timeout. (The host should not be counted, thus allowing all hosts to
5272 be tried at least once before bouncing.) This may have been the cause of an
5273 occasionally reported bug whereby a message would remain on the queue
5274 longer than the retry timeout, but would be bounced if a delivery was
5275 forced. I say "may" because I never totally pinned down the problem;
5276 setting up timeout/retry tests is difficult. See also the next item.
5278 79. The ultimate address timeout was not being applied to errors that involved
5279 a combination of host plus message (for example, a timeout on a MAIL
5280 command). When an address resolved to a number of possible hosts, and they
5281 were not all tried for each delivery (e.g. because of hosts_max_try), a
5282 message could remain on the queue longer than the retry timeout.
5284 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5287 81. Fixed an obscure SMTP outgoing bug which required at least the following
5288 conditions: (a) there was another message waiting for the same server;
5289 (b) the server returned 5xx to all RCPT commands in the first message so
5290 that the message was not completed; (c) the server dropped the connection
5291 or gave a negative response to the RSET that Exim sends to abort the
5292 transaction. The observed case was a dropped connection after DATA that had
5293 been sent in pipelining mode. That is, the server had advertised PIPELINING
5294 but was not implementing it correctly. The effect of the bug was incorrect
5295 behaviour, such as trying another host, and this could lead to a crash.
5301 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5302 and daemon.c were passed as pointers to ints; they should have been
5303 pointers to socklen_t variables (which are typically unsigned ints).
5305 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5308 3. Fixed a really odd bug that affected only the testing scheme; patching a
5309 certain fixed string in the binary changed the value of another string that
5310 happened to be identical to the end of the original first string.
5312 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5313 name", it returns that address as the IP address. On some operating
5314 systems (e.g. Solaris), it also passes back the IP address string as the
5315 "host name". However, on others (e.g. Linux), it passes back an empty
5316 string. Exim wasn't checking for this, and was changing the host name to an
5317 empty string, assuming it had been canonicized.
5319 5. Although rare, it is permitted to have more than one PTR record for a given
5320 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5321 all the names associated with an address, because they do in Solaris.
5322 However, it seems that they do not in Linux for data that comes from the
5323 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5324 I found this out when I moved to a new Linux workstation and tried to run
5325 the Exim test suite.
5327 To get round this problem I have changed the code so that it now does its
5328 own call to the DNS to look up PTR records when searching for a host name.
5329 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5330 addresses that are only in /etc/hosts are still found.
5332 This behaviour is, however, controlled by an option called host_lookup_
5333 order, which defaults to "bydns:byaddr". If people want to use the other
5334 order, or indeed, just use one or the other means of lookup, they can
5335 specify it in this variable.
5337 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5338 some operating systems, this comes back from gethostbyaddr() as an empty
5339 string, and this is what Exim used to test for. However, it seems that in
5340 other systems, "." is yielded. Exim now tests for this case too.
5342 7. The values of check_spool_space and check_log_space are now held internally
5343 as a number of kilobytes instead of an absolute number of bytes. If a
5344 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5345 kilobyte. This means that much larger values can be stored.
5347 8. Exim monitor: an attempt to get the action menu when not actually pointing
5348 at a message produces an empty menu entitled "No message selected". This
5349 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5350 no entries in it ("Shell widget menu has zero width and/or height"). So I
5351 have added a single, blank menu entry in this case.
5353 9. Added ${quote_local_part.
5355 10. MIME decoding is now applied to the contents of Subject: header lines when
5358 11. Now that a reference to $sender_host_address automatically causes a reverse
5359 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5360 host lookup before query-style lookups in lists that might use this
5361 variable. This has therefore been abolished, and the "net-" prefix is no
5362 longer necessary for query-style lookups.
5364 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5365 have been a typo for LFLAGS, so it has been changed.
5367 13. The install script calls Exim with "-C /dev/null" in order to find the
5368 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5369 to be output. Howeve, since Exim outputs its version number before the
5370 error, it didn't break the script. It just looked ugly. I fixed this by
5371 always allowing "-C /dev/null" if the caller is root.
5373 14. Ignore overlarge ACL variable number when reading spool file - insurance
5374 against a later release with more variables having written the file.
5376 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5377 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5378 this, as well as the form without the "IPv6" on the front.
5380 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5381 OS/Makefile-Darwin file.
5383 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5384 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5386 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5387 variable. However, some OS use ints for inodes. I've added cast to long int
5388 to get rid of the compiler warning.
5390 19. I had forgotten to lock out "/../" in configuration file names when
5391 ALT_CONFIG_PREFIX was set.
5393 20. Routers used for verification do not need to specify transports. However,
5394 if such a router generated a host list, and callout was configured, Exim
5395 crashed, because it could not find a port number from the (non-existent)
5396 transport. It now assumes port 25 in this circumstance.
5398 21. Added the -t option to exigrep.
5400 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5401 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5402 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5403 them is set, LOOKUP_LSEARCH is forced.
5405 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5407 24. Added sender and host information to the "rejected by local_scan()" log
5408 line; previously there was no indication of these.
5410 25. Added .include_if_exists.
5412 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5413 Linux. It turns out that not all file systems support this. Apparently some
5414 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5415 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5416 not supported on the file descriptor, is now ignored when Exim is trying to
5417 sync a directory. This applies only to Linux.
5419 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5421 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5422 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5425 29. Added support for maildirsize files from supplied patch (modified a bit).
5427 30. The use of :fail: followed by an empty string could lead Exim to respond to
5428 sender verification failures with (e.g.):
5430 550 Verification failed for <xxx>
5431 550 Sender verify failed
5433 where the first response line was missing the '-' that indicates it is not
5434 the final line of the response.
5436 31. The loop for finding the name of the user that called Exim had a hardwired
5437 limit of 10; it now uses the value of finduser_retries, which is used for
5438 all other user lookups.
5440 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5443 33. Exim was neglecting to zero errno before one call of strtol() when
5444 expanding a string and expecting an integer value. On some systems this
5445 resulted in spurious "integer overflow" errors. Also, it was casting the
5446 result into an int without checking.
5448 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5449 did not work. The code looks as if it has *never* worked, though it appears
5450 to have been documented since at least releast 1.62. I have made it work.
5452 35. The "timeout_DNS" error in retry rules, also documented since at least
5453 1.62, also never worked. As it isn't clear exactly what this means, and
5454 clearly it isn't a major issue, I have abolished the feature by treating it
5455 as "timeout", and writing a warning to the main and panic logs.
5457 36. The display of retry rules for -brt wasn't always showing the error code
5460 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5461 timeout_connect_A, timeout_connect_MX.
5463 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5464 to the empty sender.
5466 39. The daemon was not analysing the content of -oX till after it had closed
5467 stderr and disconnected from the controlling terminal. This meant that any
5468 syntax errors were only noted on the panic log, and the return code from
5469 the command was 0. By re-arranging the code a little, I've made the
5470 decoding happen first, so such errors now appear on stderr, and the return
5471 code is 1. However, the actual setting up of the sockets still happens in
5472 the disconnected process, so errors there are still only recorded on the
5475 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5476 connections (as happens on some IP stacks) was logged at start up time as
5477 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5478 it from "IPv6 and IPv4", which means that two separate sockets are being
5481 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5482 says whether AF_INET or AF_INET6 was passed as an argument.
5484 42. Exiwhat output was messed up when time zones were included in log
5487 43. Exiwhat now gives more information about the daemon's listening ports,
5488 and whether -tls-on-connect was used.
5490 44. The "port" option of the smtp transport is now expanded.
5492 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5493 silently ignored. Now an error message is written to the main and panic
5496 46. There's a new ACL modifier called "logwrite" which writes to a log file
5497 as soon as it is encountered.
5499 47. Added $local_user_uid and $local_user_gid at routing time.
5501 48. Exim crashed when trying to verify a sender address that was being
5504 49. Exim was recognizing only a space character after ".include". It now also
5505 recognizes a tab character.
5507 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5508 extracting the relevant information from the specification. The man page no
5509 longer contains scrambled data for the -d option, and I've added a section
5510 at the front about calling Exim under different names.
5512 51. Added "extra_headers" argument to the "mail" command in filter files.
5514 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5517 53. Installed eximstats 1.29.
5519 54. Added transport_filter_timeout as a generic transport option.
5521 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5522 Cc: header lines. This was required by RFC 822, but it not required by RFC
5525 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5526 incoming messages that did not have them. Now it does so only if the
5527 message originates locally, that is, if there is no associated remote host
5528 address. When Resent- header lines are present, this applies to the Resent-
5529 lines rather than the non-Resent- lines.
5531 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5532 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5534 58. Messages for configuration errors now include the name of the main
5535 configuration file - useful now that there may be more than one file in a
5536 list (.included file names were always shown).
5538 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5539 for those rare installations that do not start the daemon as root or run it
5540 setuid root. I've cut out the call to initgroups() if the daemon is not
5543 60. The Exim user and group can now be bound into the binary as text strings
5544 that are looked up at the start of Exim's processing.
5546 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5548 62. Added $mailstore_basename variable.
5550 63. Installed patch to sieve.c from Michael Haardt.
5552 64. When Exim failed to open the panic log after failing to open the main log,
5553 the original message it was trying to log was written to stderr and debug
5554 output, but if they were not available (the usual case in production), it
5555 was lost. Now it is written to syslog before the two lines that record the
5556 failures to open the logs.
5558 65. Users' Exim filters run in subprocesses under the user's uid. It is
5559 possible for a "deliver" command or an alias in a "personal" command to
5560 provoke an address rewrite. If logging of address rewriting is configured,
5561 this fails because the process is not running as root or exim. There may be
5562 a better way of dealing with this, but for the moment (because 4.30 needs
5563 to be released), I have disabled address rewrite logging when running a
5564 filter in a non-root, non-exim process.
5570 1. The buildconfig auxiliary program wasn't quoting the value set for
5571 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5572 not defined. This bug was masked in 4.22 by the effect that was fixed in
5575 2. Some messages that were rejected after a message id was allocated were
5576 shown as "incomplete" by exigrep. It no longer does this for messages that
5577 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5579 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5580 have allow_domain_literals set, the ID did not get logged in the <= line.
5581 Domain literals are now always recognized in Message-ID: header lines.
5583 4. The first argument for a ${extract expansion item is the key name or field
5584 number. Leading and trailing spaces in this item were not being ignored,
5585 causing some misleading effects.
5587 5. When deliver_drop_privilege was set, single queue runner processes started
5588 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5589 same command in the process it spins off) were not dropping privilege.
5591 6. When the daemon running as "exim" started a queue runner, it always
5592 re-executed Exim in the spun-off process. This is a waste of effort when
5593 deliver_drop_privilege is set. The new process now just calls the
5594 queue-runner function directly.
5600 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5603 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5604 ignored. Though the use of -oP was forcing the writing of a pid file, it
5605 was always written to the default place.
5607 3. If the message "no IP address found for host xxxx" is generated during
5608 incoming verification, it is now followed by identification of the incoming
5609 connection (so you can more easily find what provoked it).
5611 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5613 5. Added some features to "harden" Exim a bit more against certain attacks:
5615 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5616 be put in Local/Makefile. This is like the never_users runtime option,
5617 but it cannot be overridden. The default setting is "root".
5619 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5620 prefix string with which any file named in a -C command line option
5623 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5624 is retained for -C and -D only if the caller of Exim is root. Without
5625 it, the exim user may also use -C and -D and retain privilege.
5627 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5628 command line option is disabled.
5630 6. Macro names set by the -D option must start with an upper case letter, just
5631 like macro names defined in the configuration file.
5633 7. Added "dereference=" facility to LDAP.
5635 8. Two instances of the typo "uknown" in the source files are fixed.
5637 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5638 the Configure-Makefile script screwed up while processing it.
5640 10. Incorporated PCRE 4.4.
5642 11. The SMTP synchronization check was not operating right at the start of an
5643 SMTP session. For example, it could not catch a HELO sent before the client
5644 waited for the greeting. There is now a check for outstanding input at the
5645 point when the greeting is written. Because of the duplex, asynchronous
5646 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5647 way, but not yet received, when the check is performed.
5649 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5650 on TCP/IP sockets, because this apparently causes some broken clients to
5653 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5654 unchanged) from the Cygwin maintainer.
5656 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5657 was supported instead of testing for mbx. Effectively a typo.
5659 15. The spa authenticator server code was not checking that the input it
5660 received was valid base64.
5662 16. The debug output line for the "set" modifier in ACLs was not showing the
5663 name of the variable that was being set.
5665 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5667 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5668 lookup on its own. Something else was needed to trigger the lookup. For
5669 example, a match in host_lookup or the need for a host name in a host list.
5670 Now, if $sender_host_name is referenced and the host name has not yet been
5671 looked up, a lookup is performed. If the lookup fails, the variable remains
5672 empty, and $host_lookup_failed is set to "1".
5674 19. Added "eqi" as a case-independent comparison operator.
5676 20. The saslauthd authentication condition could segfault if neither service
5677 nor realm was specified.
5679 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5680 error message that was logged was misleading, and incoming SMTP
5681 connections were dropped. The message is now more accurate, and temporary
5682 errors are given to SMTP connections.
5684 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5685 (except RSET) with a 421 error, until QUIT is received. However, it was
5686 failing to send a response to QUIT.
5688 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5689 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5690 ACL is now run after the helo_try_verify_hosts code.
5692 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5693 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5694 case-independent, so other case variants are also recognized). Apparently
5695 some systems use these upper case variants.
5697 25. If more than two messages were waiting for the same host, and a transport
5698 filter was specified for the transport, Exim sent two messages over the
5699 same TCP/IP connection, and then failed with "socket operation on non-
5700 socket" when it tried to send the third.
5702 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5704 27. The extern definition of crypt16() in expand.c was not being excluded when
5705 the OS had its own crypt16() function.
5707 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5708 as a preferred synonym for return_size_limit, both as an option and as an
5711 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5713 30. Changed the default configuration ACL to relax the local part checking rule
5714 for addresses that are not in any local domains. For these addresses,
5715 slashes and pipe symbols are allowed within local parts, but the sequence
5716 /../ is explicitly forbidden.
5718 31. SPA server authentication was not clearing the challenge buffer before
5721 32. log_message in a "warn" ACL statement was writing to the reject log as
5722 well as to the main log, which contradicts the documentation and doesn't
5723 seem right (because no rejection is happening). So I have stopped it.
5725 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5726 However, I am unable to do any testing of this.
5728 34. Fixed an infelicity in the appendfile transport. When checking directories
5729 for a mailbox, to see if any needed to be created, it was accidentally
5730 using path names with one or more superfluous leading slashes; tracing
5731 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5733 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5734 contents are added to the log line that is written for every discarded
5735 recipient. (Previously a log_message setting was ignored.)
5737 36. The ${quote: operator now quotes the string if it is empty.
5739 37. The install script runs exim in order to find its version number. If for
5740 some reason other than non-existence or emptiness, which it checks, it
5741 could not run './exim', it was installing it with an empty version number,
5742 i.e. as "exim-". This error state is now caught, and the installation is
5745 38. An argument was missing from the function that creates an error message
5746 when Exim fails to connect to the socket for saslauthd authentication.
5747 This could cause Exim to crash, or give a corrupted message.
5749 39. Added isip, isip4, and isip6 to ${if conditions.
5751 40. The ACL variables $acl_xx are now saved with the message, and can be
5752 accessed later in routers, transports, and filters.
5754 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5755 strings in the file are not string-expanded.
5757 42. If a MAIL command specified a SIZE value that was too large to fit into an
5758 int variable, the check against message_size_limit failed. Such values are
5759 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5760 day this will have to be increased, but I don't think I want to be around
5761 when emails are that large.
5768 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5769 iconv() is not standard in FreeBSD.
5771 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5772 IPv6 enabled. The observed symptom was a segmentation fault on return from
5773 the function os_common_find_running_interfaces() in src/os.c.
5775 3. In the check_special_case() function in daemon.c I had used "errno" as an
5776 argument name, which causes warnings on some systems. This was basically a
5777 typo, since it was named "eno" in the comments!
5779 4. The code that waits for the clock to tick (at a resolution of some fraction
5780 of a second) so as to ensure message-id uniqueness was always waiting for
5781 at least one whole tick, when it could have waited for less. [This is
5782 almost certainly not relevant at current processor speeds, where it is
5783 unlikely to ever wait at all. But we try to future-proof.]
5785 5. The function that sleeps for a time interval that includes fractions of a
5786 second contained a race. It did not block SIGALRM between setting the
5787 timer, and suspending (a couple of lines later). If the interval was short
5788 and the sigsuspend() was delayed until after it had expired, the suspension
5789 never ended. On busy systems this could lead to processes getting stuck for
5792 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5793 process, before it forks any delivery processes. The open lookup caching
5794 mechanism meant that the open file or database connection was passed into
5795 the delivery process. The problem was that delivery processes always tidy
5796 up cached lookup data. This could cause a problem for the next delivery
5797 process started by the queue runner, because the external queue runner
5798 process does not know about the closure. So the next delivery process
5799 still has data in the lookup cache. In the case of a file lookup, there was
5800 no problem because closing a file descriptor in a subprocess doesn't affect
5801 the parent. However, if the lookup was caching a connection to a database,
5802 the connection was closed, and the second delivery process was likely to
5803 see errors such as "PGSQL: query failed: server closed the connection
5804 unexpectedly". The problem has been fixed by closing all cached lookups
5805 in a queue runner before running a delivery process.
5807 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5808 seem to have the "const" qualifier which it has on other OS. I've
5811 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5812 *of the same type* (client or server) with the same public name that an
5813 error should be diagnosed.
5815 9. When Exim looked up a host name for an IP address, but failed to find the
5816 original IP address when looking up the host name (a safety check), it
5817 output the message "<ip address> does not match any IP for NULL", which was
5818 confusing, to say the least. The bug was that the host name should have
5819 appeared instead of "NULL".
5821 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5822 user that is built into the binary, and the -C or -D options is used, root
5823 privilege is dropped before the configuration file is read. In addition,
5824 logging is switched to stderr instead of the normal log files. If the
5825 configuration then re-defines the Exim user, the unprivileged environment
5826 is probably not what is expected, so Exim logs a panic warning message (but
5829 However, if deliver_drop_privilege is set, the unprivileged state may well
5830 be exactly what is intended, so the warning has been cut out in that case,
5831 and Exim is allowed to try to write to its normal log files.
5837 1. smtp_return_error_details was not giving details for temporary sender
5838 or receiver verification errors.
5840 2. Diagnose a configuration error if two authenticators have the same public
5843 3. Exim used not to create the message log file for a message until the first
5844 delivery attempt. This could be confusing when incoming messages were held
5845 for policy or load reasons. The message log file is now created at the time
5846 the message is received, and an initial "Received" line is written to it.
5848 4. The automatically generated man page for command line options had a minor
5849 bug that caused no ill effects; however, a more serious problem was that
5850 the procedure for building the man page automatically didn't always
5851 operate. Consequently, release 4.20 contains an out-of-date version. This
5852 shouldn't happen again.
5854 5. When building Exim with embedded Perl support, the script that builds the
5855 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5856 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5858 6. The freeze_tell option was not being used for messages that were frozen on
5859 arrival, either by an ACL or by local_scan().
5861 7. Added the smtp_incomplete_transaction log selector.
5863 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5864 was accepting AUTH without a new EHLO.
5866 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5867 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5870 10. Logging of TCP/IP connections (when configured) now happens in the main
5871 daemon process instead of the child process, so that the TCP/IP connection
5872 count is more accurate (but it can never be perfect).
5874 11. The use of "drop" in a nested ACL was not being handled correctly in the
5875 outer ACL. Now, if condition failure induced by the nested "drop" causes
5876 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5877 or "require"), the connection is dropped.
5879 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5880 that yield "discard" can now be used with an "accept" or a "discard" verb,
5881 but an error is generated for any others (because I can't see a useful way
5882 to define what should happen).
5884 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5885 are now processed in the same way as lines in the Exim configuration file.
5886 In particular, continuation lines are supported.
5888 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5890 15. Added -ti meaning -t -i.
5892 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5893 domains, and warn by logging if others are found.
5895 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5896 returned from GIFCONF when Exim is trying to find the list of interfaces on
5897 a host. The code in os.c has been modified to copy each ifreq to an aligned
5898 structure in all cases.
5900 Also, in some cases, the returned ifreq's were being copied to a 'struct
5901 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5902 means the last couple of bytes of an IPv6 address could be chopped if the
5903 ifreq contained only a normal sockaddr (14 bytes storage).
5905 18. Named domain lists were not supported in the hosts_treat_as_local option.
5906 An entry such as +xxxx was not recognized, and was treated as a literal
5909 19. Ensure that header lines added by a DATA ACL are included in the reject log
5910 if the ACL subsequently rejects the message.
5912 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5913 MD5 (which is deprecated).
5915 21. When testing a filter file using -bf, Exim was writing a message when it
5916 took the sender from a "From " line in the message, but it was not doing so
5917 when it took $return_path from a Return-Path: header line. It now does.
5919 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5920 with a valid header line field name (a series of printing characters
5921 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5923 23. Changed "disc" in the source to "disk" to conform to the documentation and
5924 the book and for uniformity.
5926 24. Ignore Sendmail's -Ooption=value command line item.
5928 25. When execve() failed while trying to run a command in a pipe transport,
5929 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5930 could be confused with a return value of 69 from the command itself. This
5931 has been changed to 127, the value the shell returns if it is asked to run
5932 a non-existent command. The wording for the related log line suggests a
5933 non-existent command as the problem.
5935 26. If received_header_text expands to an empty string, do not add a Received:
5936 header line to the message. (Well, it adds a token one on the spool, but
5937 marks it "old" so that it doesn't get used or transmitted.)
5939 27. Installed eximstats 1.28 (addition of -nt option).
5941 28. There was no check for failure on the call to getsockname() in the daemon
5942 code. This can fail if there is a shortage of resources on the system, with
5943 ENOMEM, for example. A temporary error is now given on failure.
5945 29. Contrary to the C standard, it seems that in some environments, the
5946 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5947 program. Exim now does this explicitly; it affects the formatting of
5948 timestamps using strftime().
5950 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5951 complaints. I've now initialized all the variables, to avoid this.
5953 32. Header lines added by a system filter were not being "seen" during
5954 transport-time rewrites.
5956 33. The info_callback() function passed to OpenSSL is set up with type void
5957 (*)(SSL *, int, int), as described somewhere. However, when calling the
5958 function (actually a macro) that sets it up, the type void(*)() is
5959 expected. I've put in a cast to prevent warnings from picky compilers.
5961 34. If a DNS black list lookup found a CNAME record, but there were no A
5962 records associated with the domain it pointed at, Exim crashed.
5964 35. If a DNS black list lookup returned more than one A record, Exim ignored
5965 all but the first. It now scans all returned addresses if a particular IP
5966 value is being sought. In this situation, the contents of the
5967 $dnslist_value variable are a list of all the addresses, separated by a
5970 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5971 to accept a host name and all its aliases if the forward lookup for any of
5972 them yielded the IP address of the incoming connection. Now it accepts only
5973 those names whose forward lookup yields the correct IP address. Any other
5974 names are discarded. This closes a loophole whereby a rogue DNS
5975 administrator could create reverse DNS records to break through a
5976 wildcarded host restriction in an ACL.
5978 37. If a user filter or a system filter that ran in a subprocess used any of
5979 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5980 the wrong values were passed to the pipe command ($thisaddress had the
5981 value of $0, $0 had the value of $1, etc). This bug was introduced by
5982 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5984 38. Improved the line breaking for long SMTP error messages from ACLs.
5985 Previously, if there was no break point between 40 and 75 characters, Exim
5986 left the rest of the message alone. Two changes have been made: (a) I've
5987 reduced the minimum length to 35 characters; (b) if it can't find a break
5988 point between 35 and 75 characters, it looks ahead and uses the first one
5989 that it finds. This may give the occasional overlong line, but at least the
5990 remaining text gets split now.
5992 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5993 file descriptors might be low, and that setting 1000 would always raise it.
5994 It turns out that in some environments, the limit is already over 1000 and
5995 that lowering it causes trouble. So now Exim takes care not to decrease it.
5997 40. When delivering a message, the value of $return_path is set to $sender_
5998 address at the start of routing (routers may change the value). By an
5999 oversight, this default was not being set up when an address was tested by
6000 -bt or -bv, which affected the outcome if any router or filter referred to
6003 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6004 log only when "message" or "log_message" are set. However, if one of the
6005 conditions was an address verification, or a call to a nested ACL, the
6006 messages generated by the underlying test were being passed through. This
6007 no longer happens. The underlying message is available in $acl_verify_
6008 message for both "message" and "log_message" expansions, so it can be
6009 passed through if needed.
6011 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6012 new expansion $bh_ to give the encoded byte string without charset
6013 translation. Translation happens only if iconv() is available; HAVE_ICONV
6014 indicates this at build time. HEADERS_CHARSET gives the charset to
6015 translate to; headers_charset can change it in the configuration, and
6016 "headers charset" can change it in an individual filter file.
6018 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6019 that creates RFC 2047 encoded "words" labels them as that charset instead
6020 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6021 expansion operator; (ii) when Exim creates a From: line for a local
6022 message; (iii) when a header line is rewritten to include a "phrase" part.
6024 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6025 buggy, causing it to skip the first lines of messages whose message ID
6026 ended in 'D'. This would not have bitten before Exim release 4.14, because
6027 message IDs were unlikely to end in 'D' before then. The effect was to have
6028 incorrect size information for certain domains.
6030 45. #include "config.h" was missing at the start of the crypt16.c module. This
6031 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6034 46. If there was a timeout during a "random" callout check, Exim treated it as
6035 a failure of the random address, and carried on sending RSET and the real
6036 address. If the delay was just some slowness somewhere, the response to the
6037 original RCPT would be taken as a response to RSET and so on, causing
6038 mayhem of various kinds.
6040 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6041 when I implemented it. It didn't allow for the fact that some option values
6042 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6043 the right test for positive values.
6045 48. Domain names in DNS records are case-independent. Exim always looks them up
6046 in lower case. Some resolvers return domain names in exactly the case they
6047 appear in the zone file, that is, they may contain uppercase letters. Not
6048 all resolvers do this - some return always lower case. Exim was treating a
6049 change of case by a resolver as a change of domain, similar to a widening
6050 of a domain abbreviation. This triggered its re-routing code and so it was
6051 trying to route what was effectively the same domain again. This normally
6052 caused routing to fail (because the router wouldn't handle the domain
6053 twice). Now Exim checks for this case specially, and just changes the
6054 casing of the domain that it ultimately uses when it transmits the message
6057 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6060 50. If a filter generated a file delivery with a non-absolute name (possible if
6061 no home directory exists for the router), the forbid_file option was not
6064 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6065 the existing equality matching.
6067 52. Exim was using ints instead of ino_t variables in some places where it was
6068 dealing with inode numbers.
6070 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6071 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6072 called TMPDIR, and if it finds it is different, it changes its value.
6074 54. The smtp_printf() function is now made available to local_scan() so
6075 additional output lines can be written before returning. There is also an
6076 smtp_fflush() function to enable the detection of a dropped connection.
6077 The variables smtp_input and smtp_batched_input are exported to
6080 55. Changed the default runtime configuration: the message "Unknown user"
6081 has been removed from the ACL, and instead placed on the localuser router,
6082 using the cannot_route_message feature. This means that any verification
6083 failures that generate their own messages won't get overridden. Similarly,
6084 the "Unrouteable address" message that was in the ACL for unverifiable
6085 relay addresses has also been removed.
6087 56. Added hosts_avoid_esmtp to the smtp transport.
6089 57. The exicyclog script was not checking for the esoteric option
6090 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6091 will work only if exicyclog is run under the appropriate euid.
6093 58. Following a discussion on the list, the rules by which Exim recognises line
6094 endings on incoming messages have been changed. The -dropcr and drop_cr
6095 options are now no-ops, retained only for backwards compatibility. The
6096 following line terminators are recognized: LF CRLF CR. However, special
6097 processing applies to CR:
6099 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6100 nor a local message in the state where . is a terminator.
6102 (ii) If a bare CR is encountered in a header line, an extra space is added
6103 after the line terminator so as not to end the header. The reasoning
6104 behind this is that bare CRs in header lines are most likely either
6105 to be mistakes, or people trying to play silly games.
6107 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6108 was being incorrectly given as 18 bytes larger than it should have been.
6109 This is a VOB (very old bug).
6111 60. This may never have affected anything current, but just in case it has:
6112 When the local host is found other than at the start of a list of hosts,
6113 the local host, those with the same MX, and any that follow, are discarded.
6114 When the list in question was part of a longer list of hosts, the following
6115 hosts (not currently being processed) were also being discarded. This no
6116 longer happens. I'm not sure if this situation could ever has previously
6119 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6122 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6123 header. This is something that is recommended in a new Internet Draft, and
6124 is something that is documented as being done by Sendmail. There are two
6125 possible values. For messages generated by the autoreply transport, Exim
6128 Auto-Submitted: auto-replied
6130 whereas for all other generated messages (e.g. bounces) it adds
6132 Auto-Submitted: auto-generated
6134 63. The "personal" condition in filters now includes a test for the
6135 Auto-Submitted: header. If it contains the string "auto-" the message it
6136 not considered personal.
6138 64. Added rcpt_include_affixes as a generic transport option.
6140 65. Added queue_only_override (default true).
6142 66. Added the syslog_duplication option.
6144 67. If what should have been the first header line of a message consisted of
6145 a space followed by a colon, Exim was mis-interpreting it as a header line.
6146 It isn't of course - it is syntactically invalid and should therefore be
6147 treated as the start of the message body. The misbehaviour could have
6148 caused a number of strange effects, including loss of data in subsequent
6149 header lines, and spool format errors.
6151 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6152 client host had authenticated. This control can now be exercised by an ACL
6153 for more flexibility.
6155 69. By default, callouts do not happen when testing with -bh. There is now a
6156 variant, -bhc, which does actually run the callout code, including
6157 consulting and updating the callout cache.
6159 70. Added support for saslauthd authentication, courtesy of Alexander
6162 71. If statvfs() failed on the spool or log directories while checking their
6163 size for availability, Exim confusingly gave the error "space shortage".
6164 Furthermore, in debugging mode it crashed with a floating point exception.
6165 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6166 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6167 serious problem, Exim now writes to the main and panic logs when this
6168 happens, with details of the failure. It then refuses to accept the
6169 incoming message, giving the message "spool directory problem" or "log
6170 directory problem" with a 421 code for SMTP messages.
6172 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6173 0, 1, and 2 exist, because some OS complain for execs without them (see
6174 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6175 descriptors. However, the code omitted to check that the open succeeded,
6176 causing mysterious errors if for some reason the permissions on /dev/null
6177 got screwed. Now Exim writes a message to the main and panic logs, and
6178 bombs out if it can't open /dev/null.
6180 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6181 interact so that it is all more flexible. It is supposed to remain
6182 backwards compatible. Also added extra_local_interfaces.
6184 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6185 to bomb out with an assertion failure - to the client this appears as a
6186 connection drop. This problem occurs in the part of the code that was taken
6187 from the Samba project. Fortunately, the assertion is in a very simple
6188 function, so I have fixed this by reproducing the function inline in the
6189 one place where it is called, and arranging for authentication to fail
6190 instead of killing the process with assert().
6192 75. The SPA client code was not working when the server requested OEM rather
6193 than Unicode encoding.
6195 76. Added code to make require_files with a specific uid setting more usable in
6196 the case where statting the file as root fails - usually a non-root-mounted
6197 NFS file system. When this happens and the failure is EACCES, Exim now
6198 forks a subprocess and does the per-uid checking as the relevant uid.
6200 77. Added process_log_path.
6202 78. If log_file_path was not explicitly set, a setting of check_log_space or
6203 check_log_inodes was ignored.
6205 79. If a space check for the spool or log partitions fails, the incident is now
6206 logged. Of course, in the latter case the data may get lost...
6208 80. Added the %p formatting code to string_format() so that it can be used to
6209 print addresses in debug_print(). Adjusted all the address printing in the
6210 debugging in store.c to use %p rather than %d.
6212 81. There was a concern that a line of code in smtp_in.c could overflow a
6213 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6214 initially expressed, the concern was not well-founded, because trailing
6215 spaces are removed early. However, if the trailing spaces were followed by
6216 a NULL, they did not get removed, so the overflow was possible. Two fixes
6219 (a) I re-wrote the offending code in a cleaner fashion.
6220 (b) If an incoming SMTP command contains a NULL character, it is rejected
6223 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6224 runs initgroups(), so that if the Exim user is in any additional groups,
6225 they will be used during message reception.
6231 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob