1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
57 TL/01 When running the test suite, the README says that variables such as
58 no_msglog_check are global and can be placed anywhere in a specific
59 test's script, however it was observed that placement needed to be near
60 the beginning for it to behave that way. Changed the runtest perl
61 script to read through the entire script once to detect and set these
62 variables, reset to the beginning of the script, and then run through
63 the script parsing/test process like normal.
65 TL/02 The BSD's have an arc4random API. One of the functions to induce
66 adding randomness was arc4random_stir(), but it has been removed in
67 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
68 function when detected.
70 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
71 cause callback expansion.
73 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
74 syntax errors in an expansion can be treated as a string instead of
75 logging or causing an error, due to the internal use of bool_lax
76 instead of bool when processing it.
78 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
79 server certificates when making smtp deliveries.
81 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
83 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
85 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
87 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
88 Merged patch from Sebastian Wiedenroth.
90 JH/05 Fix results-pipe from transport process. Several recipients, combined
91 with certificate use, exposed issues where response data items split
92 over buffer boundaries were not parsed properly. This eventually
93 resulted in duplicates being sent. This issue only became common enough
94 to notice due to the introduction of conection certificate information,
95 the item size being so much larger. Found and fixed by Wolfgang Breyha.
97 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
98 size buffer was used, resulting in syntax errors when an expansion
101 JH/07 Add support for directories of certificates when compiled with a GnuTLS
102 version 3.3.6 or later.
104 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
105 is EXPERIMENTAL_EVENT, the main-configuration and transport options
106 both become "event_action", the variables become $event_name, $event_data
107 and $event_defer_errno. There is a new variable $verify_mode, usable in
108 routers, transports and related events. The tls:cert event is now also
109 raised for inbound connections, if the main configuration event_action
112 TL/06 In test suite, disable OCSP for old versions of openssl which contained
113 early OCSP support, but no stapling (appears to be less than 1.0.0).
115 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
116 server certificate names available under the smtp transport option
117 "tls_verify_cert_hostname" now do not permit multi-component wildcard
120 JH/10 Time-related extraction expansions from certificates now use the main
121 option "timezone" setting for output formatting, and are consistent
122 between OpenSSL and GnuTLS compilations. Bug 1541.
124 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
125 encoded parameter in the incoming message. Bug 1558.
127 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
128 include certificate info, eximon was claiming there were spoolfile
131 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
133 JH/14 Log delivery-related information more consistently, using the sequence
134 "H=<name> [<ip>]" wherever possible.
136 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
137 are problematic for Debian distribution, omit them from the release
140 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
142 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
145 JH/17 Fix a null-indirection in certextract expansions when a nondefault
146 output list separator was used.
151 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
152 checkers that were complaining about end of non-void function with no
155 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
156 This was a regression intruduced in 4.83 by another bugfix.
158 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
160 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
161 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
167 TF/01 Correctly close the server side of TLS when forking for delivery.
169 When a message was received over SMTP with TLS, Exim failed to clear up
170 the incoming connection properly after forking off the child process to
171 deliver the message. In some situations the subsequent outgoing
172 delivery connection happened to have the same fd number as the incoming
173 connection previously had. Exim would try to use TLS and fail, logging
174 a "Bad file descriptor" error.
176 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
177 utilities have not been installed.
179 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
180 temporary space as the ACL may create new global variables.
182 TL/01 LDAP support uses per connection or global context settings, depending
183 upon the detected version of the libraries at build time.
185 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
186 to extract and use the src ip:port in logging and expansions as if it
187 were a direct connection from the outside internet. PPv2 support was
188 updated based on HAProxy spec change in May 2014.
190 JH/02 Add ${listextract {number}{list}{success}{fail}}.
192 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
193 Properly escape header and check for NULL return.
195 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
198 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
200 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
201 characters in header names, implemented as a verify condition.
202 Contributed by Michael Fischer v. Mollard.
204 TL/05 Rename SPF condition results err_perm and err_temp to standardized
205 results permerror and temperror. Previous values are deprecated but
206 still accepted. In a future release, err_perm and err_temp will be
207 completely removed, which will be a backward incompatibility if the
208 ACL tests for either of these two old results. Patch contributed by
209 user bes-internal on the mailing list.
211 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
213 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
214 selectors, in both main and reject logs.
216 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
219 JH/07 Add malware type "sock" for talking to simple daemon.
221 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
223 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
224 routers/transports under cutthrough routing.
226 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
227 numbers. Touch up "bool" conditional to keep the same definition.
229 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
231 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
233 JH/12 Expand items in router/transport headers_add or headers_remove lists
234 individually rather than the list as a whole. Bug 1452.
236 Required for reasonable handling of multiple headers_ options when
237 they may be empty; requires that headers_remove items with embedded
238 colons must have them doubled (or the list-separator changed).
240 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
241 view the policy declared in the DMARC record. Currently, $dmarc_status
242 is a combined value of both the record presence and the result of the
245 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
247 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
248 dnslookup router and the smtp transport (applying to the forward
251 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
252 of ldap servers used for a specific lookup. Patch provided by Heiko
255 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
256 New variable $lookup_dnssec_authenticated for observability.
258 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
259 Patch submitted by Lars Timman.
261 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
263 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
264 Requires trusted mode and valid format message id, aborts otherwise.
265 Patch contributed by Heiko Schlichting.
267 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
268 certextract with support for various fields. Bug 1358.
270 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
271 is requested by default, modifiable by smtp transport option
274 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
275 operate on certificate variables to give certificate fingerprints
276 Also new ${sha256:cert_variable}.
278 JH/23 The PRDR feature is moved from being Experimental into the mainline.
280 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
281 Christian Aistleitner.
283 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
285 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
286 file. Patch from Wolfgang Breyha.
288 JH/25 Expand the coverage of the delivery $host and $host_address to
289 client authenticators run in verify callout. Bug 1476.
291 JH/26 Port service names are now accepted for tls_on_connect_ports, to
292 align with daemon_smtp_ports. Bug 72.
294 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
295 support and error reporting did not work properly.
297 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
298 and is readable. Patch from Andrew Colin Kissa.
300 TL/14 Enhance documentation of ${run expansion and how it parses the
301 commandline after expansion, particularly in the case when an
302 unquoted variable expansion results in an empty value.
304 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
306 PP/02 Fix internal collision of T_APL on systems which support RFC3123
307 by renaming away from it. Addresses GH issue 15, reported by
310 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
312 TL/15 SECURITY: prevent double expansion in math comparison functions
313 (can expand unsanitized data). Not remotely exploitable.
320 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
322 PP/02 Make -n do something, by making it not do something.
323 When combined with -bP, the name of an option is not output.
325 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
328 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
329 $sender_host_name and config options to manage this, and basic check
332 PP/05 DSCP support for outbound connections and control modifier for inbound.
334 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
335 (Only plugin which currently uses this is kerberos4, which nobody should
336 be using, but we should make it available and other future plugins might
337 conceivably use it, even though it would break NAT; stuff *should* be
338 using channel bindings instead).
340 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
341 name; added for Sendmail compatibility; requires admin caller.
342 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
343 just ignore it); requires trusted caller.
344 Also parse but ignore: -Ac -Am -X<logfile>
347 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
349 TL/02 Add +smtp_confirmation as a default logging option.
351 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
352 Patch by Magnus Holmgren from 2007-02-20.
354 TL/04 Bugzilla 1281 - Spec typo.
355 Bugzilla 1283 - Spec typo.
356 Bugzilla 1290 - Spec grammar fixes.
358 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
360 TL/06 Add Experimental DMARC support using libopendmarc libraries.
362 TL/07 Fix an out of order global option causing a segfault. Reported to dev
363 mailing list by by Dmitry Isaikin.
365 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
367 JH/02 Support "G" suffix to numbers in ${if comparisons.
369 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
371 NM/01 Bugzilla 1197 - Spec typo
372 Bugzilla 1196 - Spec examples corrections
374 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
376 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
377 gnutls_enable_pkcs11, but renamed to more accurately indicate its
380 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
381 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
383 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
384 "acl {{name}{arg}...}", and optional args on acl condition
387 JH/05 Permit multiple router/transport headers_add/remove lines.
389 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
391 JH/07 Avoid using a waiting database for a single-message-only transport.
392 Performance patch from Paul Fisher. Bugzilla 1262.
394 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
397 JH/09 Add $headers_added variable, with content from use of ACL modifier
398 add_header (but not yet added to the message). Bugzilla 199.
400 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
401 Pulled from Bugzilla 817 by Wolfgang Breyha.
403 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
405 (nb: this is the same fix as in Exim 4.80.1)
407 JH/11 Add A= logging on delivery lines, and a client_set_id option on
410 JH/12 Add optional authenticated_sender logging to A= and a log_selector
413 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
415 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
416 advertise SMTP AUTH mechanism to us, instead of a generic
417 protocol violation error. Also, make Exim more robust to bad
418 data from the Dovecot auth socket.
420 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
422 When a queue runner is handling a message, Exim first routes the
423 recipient addresses, during which it prunes them based on the retry
424 hints database. After that it attempts to deliver the message to
425 any remaining recipients. It then updates the hints database using
428 So if a recipient address works intermittently, it can get repeatedly
429 deferred at routing time. The retry hints record remains fresh so the
430 address never reaches the final cutoff time.
432 This is a fairly common occurrence when a user is bumping up against
433 their storage quota. Exim had some logic in its local delivery code
434 to deal with this. However it did not apply to per-recipient defers
435 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
437 This change adds a proper retry rule check during routing so that the
438 final cutoff time is checked against the message's age. We only do
439 this check if there is an address retry record and there is not a
440 domain retry record; this implies that previous attempts to handle
441 the address had the retry_use_local_parts option turned on. We use
442 this as an approximation for the destination being like a local
443 delivery, as in LMTP.
445 I suspect this new check makes the old local delivery cutoff check
446 redundant, but I have not verified this so I left the code in place.
448 TF/02 Correct gecos expansion when From: is a prefix of the username.
450 Test 0254 submits a message to Exim with the header
454 When I ran the test suite under the user fanf2, Exim expanded
455 the header to contain my full name, whereas it should have added
456 a Resent-Sender: header. It erroneously treats any prefix of the
457 username as equal to the username.
459 This change corrects that bug.
461 GF/01 DCC debug and logging tidyup
462 Error conditions log to paniclog rather than rejectlog.
463 Debug lines prefixed by "DCC: " to remove any ambiguity.
465 TF/03 Avoid unnecessary rebuilds of lookup-related code.
467 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
468 Bug spotted by Jeremy Harris; was flawed since initial commit.
469 Would have resulted in OCSP responses post-SNI triggering an Exim
470 NULL dereference and crash.
472 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
474 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
475 Bug detection, analysis and fix by Samuel Thibault.
476 Bugzilla 1331, Debian bug #698092.
478 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
480 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
481 Server implementation by Todd Lyons, client by JH.
482 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
483 config variable "prdr_enable" controls whether the server
484 advertises the facility. If the client requests PRDR a new
485 acl_data_smtp_prdr ACL is called once for each recipient, after
486 the body content is received and before the acl_smtp_data ACL.
487 The client is controlled by bolth of: a hosts_try_prdr option
488 on the smtp transport, and the server advertisement.
489 Default client logging of deliveries and rejections involving
490 PRDR are flagged with the string "PRDR".
492 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
493 fclose(). Diagnosis by Todd Lyons.
495 PP/17 Update configure.default to handle IPv6 localhost better.
496 Patch by Alain Williams (plus minor tweaks).
499 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
500 This is now consistent with GnuTLS, and is now documented: the
501 previous undocumented portable approach to treating the option as
502 unset was to force an expansion failure. That still works, and
503 an empty string is now equivalent.
505 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
506 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
507 not performing validation itself.
509 PP/20 Added force_command boolean option to pipe transport.
510 Patch from Nick Koston, of cPanel Inc.
512 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
515 TF/04 Added udpsend ACL modifer and hexquote expansion operator
517 PP/21 Fix eximon continuous updating with timestamped log-files.
518 Broken in a format-string cleanup in 4.80, missed when I repaired the
519 other false fix of the same issue.
520 Report and fix from Heiko Schlichting.
523 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
524 Report from Prashanth Katuri.
526 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
527 It's SecureTransport, so affects any MacOS clients which use the
528 system-integrated TLS libraries, including email clients.
530 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
531 using a MIME ACL for non-SMTP local injection.
532 Report and assistance in diagnosis by Warren Baker.
534 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
536 JH/16 Fix comparisons for 64b. Bugzilla 1385.
538 TL/09 Add expansion variable $authenticated_fail_id to keep track of
539 last id that failed so it may be referenced in subsequent ACL's.
541 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
544 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
545 ldap library initialization, allowing self-signed CA's to be
546 used. Also properly sets require_cert option later in code by
547 using NULL (global ldap config) instead of ldap handle (per
548 session). Bug diagnosis and testing by alxgomz.
550 TL/12 Enhanced documentation in the ratelimit.pl script provided in
551 the src/util/ subdirectory.
553 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
554 renamed to Transport Post Delivery Action by Jeremy Harris, as
557 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
558 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
559 redis_servers = needs to be configured which will be used by the redis
560 lookup. Patch from Warren Baker, of The Packet Hub.
562 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
564 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
565 hostname or reverse DNS when processing a host list. Used suggestions
566 from multiple comments on this bug.
568 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
570 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
571 Missed a few lines, added it to make the runtest require no keyboard
574 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
575 contains upper case chars. Make router use caseful_local_part.
577 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
578 support when GnuTLS has been built with p11-kit.
584 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
586 This, or similar/improved, will also be change PP/11 of 4.82.
592 PP/01 Handle short writes when writing local log-files.
593 In practice, only affects FreeBSD (8 onwards).
594 Bugzilla 1053, with thanks to Dmitry Isaikin.
596 NM/01 Bugzilla 949 - Documentation tweak
598 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
601 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
603 PP/02 Implemented gsasl authenticator.
605 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
607 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
608 `pkg-config foo` for cflags/libs.
610 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
611 with rest of GSASL and with heimdal_gssapi.
613 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
614 `pkg-config foo` for cflags/libs for the TLS implementation.
616 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
617 properties get this fed in as external SSF. A number of robustness
618 and debugging improvements to the cyrus_sasl authenticator.
620 PP/08 cyrus_sasl server now expands the server_realm option.
622 PP/09 Bugzilla 1214 - Log authentication information in reject log.
623 Patch by Jeremy Harris.
625 PP/10 Added dbmjz lookup type.
627 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
629 PP/12 MAIL args handles TAB as well as SP, for better interop with
630 non-compliant senders.
631 Analysis and variant patch by Todd Lyons.
633 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
634 Bug report from Lars Müller <lars@samba.org> (via SUSE),
635 Patch from Dirk Mueller <dmueller@suse.com>
637 PP/13 tls_peerdn now print-escaped for spool files.
638 Observed some $tls_peerdn in wild which contained \n, which resulted
639 in spool file corruption.
641 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
642 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
643 or write after TLS renegotiation, which otherwise led to messages
646 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
647 as a tracking header (ie: a signed header comes before the signature).
648 Patch from Wolfgang Breyha.
650 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
651 comma-sep list; embedded commas doubled.
653 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
655 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
657 Report and patch from Dmitry Banschikov.
659 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
660 Removed SSL_clear() after SSL_new() which led to protocol negotiation
661 failures. We appear to now support TLS1.1+ with Exim.
663 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
664 lets Exim select keys and certificates based upon TLS SNI from client.
665 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
666 before an outbound SMTP session. New log_selector, +tls_sni.
668 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
669 NULL dereference. Report and patch from Alun Jones.
671 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
672 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
673 Not seeing resolver debug output on NetBSD, but suspect this is a
674 resolver implementation change.
676 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
677 Left warnings. Added "eximon gdb" invocation mode.
679 PP/21 Defaulting "accept_8bitmime" to true, not false.
681 PP/22 Added -bw for inetd wait mode support.
683 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
684 locate the relevant includes and libraries. Made this the default.
686 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
687 Bugzilla 1246, report and most of solution from Tomasz Kusy.
689 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
690 This may cause build issues on older platforms.
692 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
693 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
694 gnutls_require_mac & gnutls_require_protocols (no longer supported).
695 Added SNI support via GnuTLS too.
696 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
698 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
700 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
701 Applied second patch from Janne, implementing suggestion to default
702 multiple-strings-in-record handling to match SPF spec.
704 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
706 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
707 read-only, out of scope).
708 Patch from Wolfgang Breyha, report from Stuart Northfield.
710 PP/29 Fix three issues highlighted by clang analyser static analysis.
711 Only crash-plausible issue would require the Cambridge-specific
712 iplookup router and a misconfiguration.
713 Report from Marcin Mirosław.
715 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
717 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
718 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
719 As part of this, removing so much warning spew let me fix some minor
720 real issues in debug logging.
722 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
723 assignment on my part. Fixed.
725 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
726 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
727 Janne Snabb (who went above and beyond: thank you).
729 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
730 string otherwise requires a connection and a bunch more work and it's
731 relatively easy to get wrong. Should also expose TLS library linkage
734 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
735 64-bit ${eval} (JH/03).
737 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
738 GNU libc to support some of the 64-bit stuff, should not lead to
739 conflicts. Defined before os.h is pulled in, so if a given platform
740 needs to override this, it can.
742 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
743 protection layer was required, which is not implemented.
744 Bugzilla 1254, patch from Wolfgang Breyha.
746 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
747 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
748 tls_dhparam take prime identifiers. Also unbreak combination of
749 OpenSSL+DH_params+TLSSNI.
751 PP/39 Disable SSLv2 by default in OpenSSL support.
757 PP/01 Solaris build fix for Oracle's LDAP libraries.
758 Bugzilla 1109, patch from Stephen Usher.
760 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
762 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
765 TF/02 Fix a couple more cases where we did not log the error message
766 when unlink() failed. See also change 4.74-TF/03.
768 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
769 lock up or crash if it happened to be inside a call to libc when it
770 got a SIGUSR1 from exiwhat.
772 The SIGUSR1 handler appends the current process status to the process
773 log which is later printed by exiwhat. It used to use the general
774 purpose logging code to do this, but several functions it calls are
775 not safe for signals.
777 The new output code in the SIGUSR1 handler is specific to the process
778 log, and simple enough that it's easy to inspect for signal safety.
779 Removing some special cases also simplifies the general logging code.
780 Removing the spurious timestamps from the process log simplifies
783 TF/04 Improved ratelimit ACL condition.
785 The /noupdate option has been deprecated in favour of /readonly which
786 has clearer semantics. The /leaky, /strict, and /readonly update modes
787 are mutually exclusive. The update mode is no longer included in the
788 database key; it just determines when the database is updated. (This
789 means that when you upgrde Exim will forget old rate measurements.)
791 Exim now checks that the per_* options are used with an update mode that
792 makes sense for the current ACL. For example, when Exim is processing a
793 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
794 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
795 must specify per_mail/readonly. If you omit the update mode it defaults to
796 /leaky where that makes sense (as before) or /readonly where required.
798 The /noupdate option is now undocumented but still supported for
799 backwards compatibility. It is equivalent to /readonly except that in
800 ACLs where /readonly is required you may specify /leaky/noupdate or
801 /strict/noupdate which are treated the same as /readonly.
803 A useful new feature is the /count= option. This is a generalization
804 of the per_byte option, so that you can measure the throughput of other
805 aggregate values. For example, the per_byte option is now equivalent
806 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
808 The per_rcpt option has been generalized using the /count= mechanism
809 (though it's more complicated than the per_byte equivalence). When it is
810 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
811 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
812 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
813 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
814 Note that using per_rcpt with a non-readonly update mode in more than
815 one ACL will cause the recipients to be double-counted. (The per_mail
816 and per_byte options don't have this problem.)
818 The handling of very low rates has changed slightly. If the computed rate
819 is less than the event's count (usually one) then this event is the first
820 after a long gap. In this case the rate is set to the same as this event's
821 count, so that the first message of a spam run is counted properly.
823 The major new feature is a mechanism for counting the rate of unique
824 events. The new per_addr option counts the number of different
825 recipients that someone has sent messages to in the last time period. It
826 behaves like per_rcpt if all the recipient addresses are different, but
827 duplicate recipient addresses do not increase the measured rate. Like
828 the /count= option this is a general mechanism, so the per_addr option
829 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
830 example, measure the rate that a client uses different sender addresses
831 with the options per_mail/unique=$sender_address. There are further
832 details in the main documentation.
834 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
836 TF/06 Removed a few PCRE remnants.
838 TF/07 Automatically extract Exim's version number from tags in the git
839 repository when doing development or release builds.
841 PP/02 Raise smtp_cmd_buffer_size to 16kB.
842 Bugzilla 879. Patch from Paul Fisher.
844 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
845 Heavily based on revision 40f9a89a from Simon Arlott's tree.
848 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
850 PP/05 Variable $av_failed, true if the AV scanner deferred.
851 Bugzilla 1078. Patch from John Horne.
853 PP/06 Stop make process more reliably on build failure.
854 Bugzilla 1087. Patch from Heiko Schlittermann.
856 PP/07 Make maildir_use_size_file an _expandable_ boolean.
857 Bugzilla 1089. Patch from Heiko Schlittermann.
859 PP/08 Handle ${run} returning more data than OS pipe buffer size.
860 Bugzilla 1131. Patch from Holger Weiß.
862 PP/09 Handle IPv6 addresses with SPF.
863 Bugzilla 860. Patch from Wolfgang Breyha.
865 PP/10 GnuTLS: support TLS 1.2 & 1.1.
867 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
870 PP/11 match_* no longer expand right-hand-side by default.
871 New compile-time build option, EXPAND_LISTMATCH_RHS.
872 New expansion conditions, "inlist", "inlisti".
874 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
876 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
878 PP/14 fix log_write() format string regression from TF/03.
879 Bugzilla 1152. Patch from Dmitry Isaikin.
885 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
887 PP/02 Harmonised TLS library version reporting; only show if debugging.
888 Layout now matches that introduced for other libraries in 4.74 PP/03.
890 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
892 PP/04 New "dns_use_edns0" global option.
894 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
897 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
898 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
900 TK/01 Updated PolarSSL code to 0.14.2.
901 Bugzilla 1097. Patch from Andreas Metzler.
903 PP/07 Catch divide-by-zero in ${eval:...}.
906 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
909 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
910 format-string attack -- SECURITY: remote arbitrary code execution.
912 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
913 time unintentionally subject to list matching rules, letting the header
914 cause arbitrary Exim lookups (of items which can occur in lists, *not*
915 arbitrary string expansion). This allowed for information disclosure.
917 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
918 INT_MIN/-1 -- value coerced to INT_MAX.
924 NM/01 Workround for PCRE version dependency in version reporting
927 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
928 This fixes portability to compilers other than gcc, notably
929 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
931 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
932 makefiles for portability to HP-UX and POSIX correctness.
934 PP/01 Permit LOOKUP_foo enabling on the make command-line.
935 Also via indented variable definition in the Makefile.
936 (Debugging by Oliver Heesakkers).
938 PP/02 Restore caching of spamd results with expanded spamd_address.
939 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
941 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
942 Improves build reliability. Fix from: Frank Elsner
944 NM/02 Fix wide character breakage in the rfc2047 coding
945 Fixes bug 1064. Patch from Andrey N. Oktyabrski
947 NM/03 Allow underscore in dnslist lookups
948 Fixes bug 1026. Patch from Graeme Fowler
950 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
951 Code patches from Adam Ciarcinski of NetBSD.
953 NM/04 Fixed exiqgrep to cope with mailq missing size issue
956 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
957 is logged, to avoid truncation. Patch from John Horne.
959 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
960 Patch from Jakob Hirsch.
962 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
963 SQL string expansion failure details.
964 Patch from Andrey Oktyabrski.
966 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
967 Patch from Simon Arlott.
969 PP/09 New lookups functionality failed to compile on old gcc which rejects
970 extern declarations in function scope.
971 Patch from Oliver Fleischmann
973 PP/10 Use sig_atomic_t for flags set from signal handlers.
974 Check getgroups() return and improve debugging.
975 Fixed developed for diagnosis in bug 927 (which turned out to be
978 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
979 Patch from Mark Zealey.
981 PP/12 Bugzilla 1056: Improved spamd server selection.
982 Patch from Mark Zealey.
984 PP/13 Bugzilla 1086: Deal with maildir quota file races.
985 Based on patch from Heiko Schlittermann.
987 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
988 Patch from Uwe Doering, sign-off by Michael Haardt.
990 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
991 variable declaration deep within a block. Bug and patch from
994 PP/15 lookups-Makefile IRIX compatibilty coercion.
996 PP/16 Make DISABLE_DKIM build knob functional.
998 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
999 Patch by Simon Arlott
1001 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1002 variable argument macros. Our copy now differs from upstream.
1008 TF/01 Failure to get a lock on a hints database can have serious
1009 consequences so log it to the panic log.
1011 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1012 controlled using the smtp_confirmation log selector.
1014 TF/03 Include the error message when we fail to unlink a spool file.
1016 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1017 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1018 for maintaining out-of-tree patches for some time.
1020 PP/01 Bugzilla 139: Documentation and portability issues.
1021 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1022 Handle per-OS dynamic-module compilation flags.
1024 PP/02 Let /dev/null have normal permissions.
1025 The 4.73 fixes were a little too stringent and complained about the
1026 permissions on /dev/null. Exempt it from some checks.
1027 Reported by Andreas M. Kirchwitz.
1029 PP/03 Report version information for many libraries, including
1030 Exim version information for dynamically loaded libraries. Created
1031 version.h, now support a version extension string for distributors
1032 who patch heavily. Dynamic module ABI change.
1034 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1035 privilege escalation vulnerability whereby the Exim run-time user
1036 can cause root to append content of the attacker's choosing to
1039 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1042 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1043 If dropping privileges for untrusted macros, we disabled normal logging
1044 on the basis that it would fail; for the Exim run-time user, this is not
1045 the case, and it resulted in successful deliveries going unlogged.
1046 Fixed. Reported by Andreas Metzler.
1052 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1053 only prepend for the Resent-* case. Fixes regression introduced in
1054 Exim 4.70 by NM/22 for Bugzilla 607.
1056 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1057 increasing numbers of administrators be bitten by this.
1059 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1061 PP/03 Bugzilla 994: added openssl_options main configuration option.
1063 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1065 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1067 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1069 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1070 without a peer certificate, leading to a segfault because of an
1071 assumption that peers always have certificates. Be a little more
1072 paranoid. Problem reported by Martin Tscholak.
1074 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1075 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1076 NB: ClamAV planning to remove STREAM in "middle of 2010".
1077 CL also introduces -bmalware, various -d+acl logging additions and
1078 more caution in buffer sizes.
1080 PP/09 Implemented reverse_ip expansion operator.
1082 PP/10 Bugzilla 937: provide a "debug" ACL control.
1084 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1086 PP/12 Bugzilla 973: Implement --version.
1088 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1090 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1092 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1094 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1095 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1096 ignore trailing whitespace.
1098 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1100 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1103 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1104 Notification from Dr Andrew Aitchison.
1106 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1107 ExtendedDetectionInfo response format.
1108 Notification from John Horne.
1110 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1113 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1114 XSL and documented dependency on system catalogs, with examples of how
1117 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1120 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1121 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1122 configuration file which is writeable by the Exim user or group.
1124 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1125 of configuration files to cover files specified with the -C option if
1126 they are going to be used with root privileges, not just the default
1129 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1130 option (effectively making it always true).
1132 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1133 files to be used while preserving root privileges.
1135 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1136 that rogue child processes cannot use them.
1138 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1139 run-time user, instead of root.
1141 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1142 Exim run-time user without dropping privileges.
1144 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1145 result string, instead of calling string_vformat() twice with the same
1148 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1149 for other users. Others should always drop root privileges if they use
1150 -C on the command line, even for a whitelisted configure file.
1152 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1154 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1160 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1161 $data_path, and $header_path variables; fixed documentation bugs and
1164 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1165 exipick to access non-standard spools, including the "frozen" queue
1168 NM/01 Bugzilla 965: Support mysql stored procedures.
1169 Patch from Alain Williams
1171 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1173 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1174 Patch from Andreas Metzler
1176 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1177 Patch from Kirill Miazine
1179 NM/05 Bugzilla 671: Added umask to procmail example.
1181 JJ/03 installed exipick 20100323.0, fixing doc bug
1183 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1184 directory. Notification and patch from Dan Rosenberg.
1186 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1188 TK/02 Improve log output when DKIM signing operation fails.
1190 MH/01 Treat the transport option dkim_domain as a colon separated
1191 list, not as a single string, and sign the message with each element,
1192 omitting multiple occurences of the same signer.
1194 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1195 Bugzilla 985, 986. Patch by Simon Arlott
1197 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1198 Patch by Simon Arlott
1200 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1201 MBX locking. Notification from Dan Rosenberg.
1207 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1209 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1211 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1213 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1215 NM/04 Bugzilla 915: Flag broken perl installation during build.
1221 TK/01 Added patch by Johannes Berg that expands the main option
1222 "spamd_address" if it starts with a dollar sign.
1224 TK/02 Write list of recipients to X-Envelope-Sender header when building
1225 the mbox-format spool file for content scanning (suggested by Jakob
1228 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1229 (http://www.dcc-servers.net/) support via dccifd. Activated by
1230 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1232 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1233 by Mark Daniel Reidel <mr@df.eu>.
1235 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1236 When building exim an external PCRE library is now needed -
1237 PCRE is a system library on the majority of modern systems.
1238 See entry on PCRE_LIBS in EDITME file.
1240 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1241 conversation. Added nologin parameter to request.
1242 Patch contributed by Kirill Miazine.
1244 TF/01 Do not log submission mode rewrites if they do not change the address.
1246 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1248 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1249 log files in place. Contributed by Roberto Lima.
1251 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1253 TF/03 Bugzilla 615: When checking the local_parts router precondition
1254 after a local_part_suffix or local_part_prefix option, Exim now
1255 does not use the address's named list lookup cache, since this
1256 contains cached lookups for the whole local part.
1258 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1259 Robert Millan. Documentation is in experimental-spec.txt.
1261 TF/04 Bugzilla 668: Fix parallel build (make -j).
1263 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1265 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1266 Patch provided by Jan Srzednicki.
1268 TF/05 Leading white space used to be stripped from $spam_report which
1269 wrecked the formatting. Now it is preserved.
1271 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1272 that they are available at delivery time.
1274 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1276 TF/08 TLS error reporting now respects the incoming_interface and
1277 incoming_port log selectors.
1279 TF/09 Produce a more useful error message if an SMTP transport's hosts
1280 setting expands to an empty string.
1282 NM/06 Bugzilla 744: EXPN did not work under TLS.
1283 Patch provided by Phil Pennock.
1285 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1286 Patch provided by Richard Godbee.
1288 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1289 acl_smtp_notquit, added index entry.
1291 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1292 Patch provided by Eugene Bujak.
1294 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1295 accept(). Patch provided by Maxim Dounin.
1297 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1298 Patch provided by Phil Pennock.
1300 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1302 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1303 Patch provided by Brad "anomie" Jorsch.
1305 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1306 Patch provided by Dean Brooks.
1308 TK/05 Add native DKIM support (does not depend on external libraries).
1310 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1311 Patch provided by Graeme Fowler.
1313 NM/16 Bugzilla 851: Documentation example syntax fix.
1315 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1317 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1320 NM/19 Bugzilla 745: TLS version reporting.
1321 Patch provided by Phil Pennock.
1323 NM/20 Bugzilla 167: bool: condition support.
1324 Patch provided by Phil Pennock.
1326 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1327 clients. Patch provided by Phil Pennock.
1329 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1330 Patch provided by Brad "anomie" Jorsch.
1332 NM/23 Bugzilla 687: Fix misparses in eximstats.
1333 Patch provided by Heiko Schlittermann.
1335 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1336 Patch provided by Heiko Schlittermann.
1338 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1339 plus update to original patch.
1341 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1343 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1344 Patch provided by David Brownlee.
1346 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1348 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1350 NM/30 Bugzilla 888: TLS documentation bugfixes.
1352 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1354 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1355 Unlike the original bugzilla I have changed all shell scripts in src tree.
1357 NM/33 Bugzilla 898: Transport filter timeout fix.
1358 Patch by Todd Rinaldo.
1360 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1361 Patch by Serge Demonchaux.
1363 NM/35 Bugzilla 39: Base64 decode bug fixes.
1364 Patch by Jakob Hirsch.
1366 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1368 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1370 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1372 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1378 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1379 ALT-N's libdkim that I have put here:
1380 http://duncanthrax.net/exim-experimental/
1382 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1383 were called 'true' and it seems that C99 defines that as a reserved
1384 keyword to be used with 'bool' variable types. That means you could
1385 not include C99-style headers which use bools without triggering
1386 build errors in sieve.c.
1388 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1389 as mailq or other aliases. Changed the --help handling significantly
1390 to do whats expected. exim_usage() emits usage/help information.
1392 SC/01 Added the -bylocaldomain option to eximstats.
1394 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1396 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1398 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1404 PH/01 Another patch from the Sieve maintainer.
1406 PH/02 When an IPv6 address is converted to a string for single-key lookup
1407 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1408 dots are used instead of colons so that keys in lsearch files need not
1409 contain colons. This was done some time before quoting was made available
1410 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1411 (notated using the quote facility) so as to distinguish them from IPv4
1412 keys. This meant that lookups for IP addresses in host lists did not work
1413 for iplsearch lookups.
1415 This has been fixed by arranging for IPv6 addresses to be expressed with
1416 colons if the lookup type is iplsearch. This is not incompatible, because
1417 previously such lookups could never work.
1419 The situation is now rather anomolous, since one *can* have colons in
1420 ordinary lsearch keys. However, making the change in all cases is
1421 incompatible and would probably break a number of configurations.
1423 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1426 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1427 conversion specification without a maximum field width, thereby enabling
1428 a rogue spamd server to cause a buffer overflow. While nobody in their
1429 right mind would setup Exim to query an untrusted spamd server, an
1430 attacker that gains access to a server running spamd could potentially
1431 exploit this vulnerability to run arbitrary code as the Exim user.
1433 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1434 $primary_hostname instead of what libspf2 thinks the hosts name is.
1436 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1437 a directory entry by the name of the lookup key. Previously, if a
1438 symlink pointed to a non-existing file or a file in a directory that
1439 Exim lacked permissions to read, a lookup for a key matching that
1440 symlink would fail. Now it is enough that a matching directory entry
1441 exists, symlink or not. (Bugzilla 503.)
1443 PH/03 The body_linecount and body_zerocount variables are now exported in the
1446 PH/04 Added the $dnslist_matched variable.
1448 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1449 This means they are set thereafter only if the connection becomes
1452 PH/06 Added the client_condition to authenticators so that some can be skipped
1453 by clients under certain conditions.
1455 PH/07 The error message for a badly-placed control=no_multiline_responses left
1456 "_responses" off the end of the name.
1458 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1460 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1461 (without spaces) instead of just copying the configuration text.
1463 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1465 PH/11 Added $max_received_linelength.
1467 PH/12 Added +ignore_defer and +include_defer to host lists.
1469 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1470 way in which PCRE > 7.0 is built.
1472 PH/14 Implemented queue_only_load_latch.
1474 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1475 MAIL command. The effect was to mangle the value on 64-bit systems.
1477 PH/16 Another patch from the Sieve maintainer.
1479 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1481 PH/18 If a system quota error occurred while trying to create the file for
1482 a maildir delivery, the message "Mailbox is full" was not appended to the
1483 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1484 only to a quota excession during the actual writing of the file.
1486 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1487 characters?) which causes problems in log lines. The DN values are now
1488 passed through string_printing() before being added to log lines.
1490 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1491 and InterBase are left for another time.)
1493 PH/21 Added message_body_newlines option.
1495 PH/22 Guard against possible overflow in moan_check_errorcopy().
1497 PH/23 POSIX allows open() to be a macro; guard against that.
1499 PH/24 If the recipient of an error message contained an @ in the local part
1500 (suitably quoted, of course), incorrect values were put in $domain and
1501 $local_part during the evaluation of errors_copy.
1507 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1508 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1511 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1512 issue a MAIL command.
1514 PH/02 In an ACL statement such as
1516 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1518 if a client was not listed at all, or was listed with a value other than
1519 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1520 the condition was not true (as it should be), so access was not denied.
1521 The bug was that the ! inversion was incorrectly passed on to the second
1522 item. This has been fixed.
1524 PH/03 Added additional dnslists conditions == and =& which are different from
1525 = and & when the dns lookup returns more than one IP address.
1527 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1528 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1530 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1531 FSYNC, which compiles an option called disable_fsync that allows for
1532 bypassing fsync(). The documentation is heavily laced with warnings.
1534 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1536 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1537 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1538 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1539 including adding "make clean"; (3) Added -fPIC when compiling the test
1540 dynamically loaded module, to get rid of a warning.
1542 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1543 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1544 = 0s. The bug is otherwise harmless.
1546 PH/07 There was a bug in the dovecot authenticator such that the value of
1547 $auth1 could be overwritten, and so not correctly preserved, after a
1548 successful authentication. This usually meant that the value preserved by
1549 the server_setid option was incorrect.
1551 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1553 PH/09 Installed PCRE release 7.0.
1555 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1556 run for batched SMTP input. It is now run at the start of every message
1557 in the batch. While fixing this I discovered that the process information
1558 (output by running exiwhat) was not always getting set for -bs and -bS
1559 input. This is fixed, and it now also says "batched" for BSMTP.
1561 PH/11 Added control=no_pipelining.
1563 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1564 patch, slightly modified), and move the expansion of helo_data till after
1565 the connection is made in the smtp transport (so it can use these
1568 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1570 PH/14 Added log_selector = +pid.
1572 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1574 PH/16 Add ${if forany and ${if forall.
1576 PH/17 Added dsn_from option to vary the From: line in DSNs.
1578 PH/18 Flush SMTP output before performing a callout, unless control =
1579 no_callout_flush is set.
1581 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1582 was true (the default) a successful delivery failed to delete the retry
1583 item, thus causing premature timeout of the address. The bug is now
1586 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1588 PH/21 Long custom messages for fakedefer and fakereject are now split up
1589 into multiline reponses in the same way that messages for "deny" and
1590 other ACL rejections are.
1592 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1593 with slight modification.
1595 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1596 draft, changing the syntax and factoring some duplicate code".
1598 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1599 for deliveries of the second and subsequent messages over the same SMTP
1602 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1603 ${reduce, with only minor "tidies".
1605 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1607 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1608 expansion side effects.
1610 PH/27 When a message times out after an over-quota error from an Exim-imposed
1611 quota, the bounce message says "mailbox is full". This message was not
1612 being given when it was a system quota that was exceeded. It now should
1615 MH/03 Made $recipients available in local_scan(). local_scan() already has
1616 better access to the recipient list through recipients_list[], but
1617 $recipients can be useful in postmaster-provided expansion strings.
1619 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1620 in the case of a MAIL command with additional options following the
1621 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1622 were accidentally chopped off.
1624 PH/29 SMTP synchronization checks are implemented when a command is read -
1625 there is a check that no more input is waiting when there shouldn't be
1626 any. However, for some commands, a delay in an ACL can mean that it is
1627 some time before the response is written. In this time, more input might
1628 arrive, invalidly. So now there are extra checks after an ACL has run for
1629 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1630 pipelining has not been advertised.
1632 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1634 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1635 returned in the SMTP response when smtp_return_error_details was false.
1636 This has been fixed.
1638 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1639 instead of the C I/O that was originally supplied, because problems were
1640 reported on Solaris.
1642 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1643 Exim which did not show up earlier: it was assuming that a call to
1644 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1645 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1646 was a macro that became an assignment, so it seemed to work. This has
1647 changed to a proper function call with a void return, hence the compile
1648 error. Exim's code has been fixed.
1650 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1653 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1655 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1657 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1658 because it turns out to be too big or there is a timeout) there is no
1659 "Completed" line in the log. When some messages of this type were
1660 selected by exigrep, they were listed as "not completed". Others were
1661 picked up by some special patterns. I have improved the selection
1662 criteria to be more general.
1664 PH/38 The host_find_failed option in the manualroute router can now be set
1665 to "ignore", to completely ignore a host whose IP address cannot be
1666 found. If all hosts are ignored, the behaviour is controlled by the new
1667 host_all_ignored option.
1669 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1670 homing or because of multiple MX records with /mx) generated more than
1671 one IP address, and the following item turned out to be the local host,
1672 all the secondary addresses of the first item were incorrectly removed
1673 from the list, along with the local host and any following hosts (which
1674 is what is supposed to happen).
1676 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1677 whoever called Exim into the -H file. In the case of the daemon it was
1678 behaving confusingly. When first started, it used values for whoever
1679 started the daemon, but after a SIGHUP it used the Exim user (because it
1680 calls itself on a restart). I have changed the code so that it now always
1683 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1684 message are rejected with the same error (e.g. no authentication or bad
1685 sender address), and a DATA command is nevertheless sent (as can happen
1686 with PIPELINING or a stupid MUA), the error message that was given to the
1687 RCPT commands is included in the rejection of the DATA command. This is
1688 intended to be helpful for MUAs that show only the final error to their
1691 PH/42 Another patch from the Sieve maintainer.
1693 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1694 Eximstats - Fixed some broken HTML links and added missing column headers
1696 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1697 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1699 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1701 PH/43 Yet another patch from the Sieve maintainer.
1703 PH/44 I found a way to check for a TCP/IP connection going away before sending
1704 the response to the final '.' that terminates a message, but only in the
1705 case where the client has not sent further data following the '.'
1706 (unfortunately, this is allowed). However, in many cases there won't be
1707 any further data because there won't be any more messages to send. A call
1708 to select() can be used: if it shows that the input is "ready", there is
1709 either input waiting, or the socket has been closed. An attempt to read
1710 the next input character can distinguish the two cases. Previously, Exim
1711 would have sent an OK response which the client would never have see.
1712 This could lead to message repetition. This fix should cure that, at
1713 least in a lot of common cases.
1715 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1716 advertised in response to EHLO.
1722 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1723 fixed by 4.65/MH/01 (is this a record?) are fixed:
1725 (i) An empty string was always treated as zero by the numeric comparison
1726 operators. This behaviour has been restored.
1728 (ii) It is documented that the numeric comparison operators always treat
1729 their arguments as decimal numbers. This was broken in that numbers
1730 starting with 0 were being interpreted as octal.
1732 While fixing these problems I realized that there was another issue that
1733 hadn't been noticed. Values of message_size_limit (both the global option
1734 and the transport option) were treated as octal if they started with 0.
1735 The documentation was vague. These values are now always treated as
1736 decimal, and I will make that clear in the documentation.
1742 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1743 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1746 MH/01 Don't check that the operands of numeric comparison operators are
1747 integers when their expansion is in "skipping" mode (fixes bug
1748 introduced by 4.64-PH/07).
1750 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1751 child addresses, Exim now panics and dies. Previously, because the count
1752 is held in a short int, deliveries were likely to be lost. As such a
1753 large number of recipients for a single message is ridiculous
1754 (performance will be very, very poor), I have chosen to impose a limit
1755 rather than extend the field.
1761 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1762 leftover -K file (the existence of which was triggered by #402).
1763 While we were at it, introduced process PID as part of the -K
1764 filename. This should rule out race conditions when creating
1767 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1768 processing considerably. Previous code took too long for large mails,
1769 triggering a timeout which in turn triggers #401.
1771 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1772 in the DK code in transports.c. sendfile() is not really portable,
1773 hence the _LINUX specificness.
1775 TF/01 In the add_headers option to the mail command in an Exim filter,
1776 there was a bug that Exim would claim a syntax error in any
1777 header after the first one which had an odd number of characters
1780 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1781 callout verification, Exim cached a "reject" for the entire domain. This
1782 is correct for most verifications, but it is not correct for a recipient
1783 verification with use_sender or use_postmaster set, because in that case
1784 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1785 case of MAIL FROM:<> rejection from other early rejections (e.g.
1786 rejection of HELO). When verifying a recipient using a non-null MAIL
1787 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1788 Whatever the result of the callout, the value of the domain cache is
1789 left unchanged (for any other kind of callout, getting as far as trying
1790 RCPT means that the domain itself is ok).
1792 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1795 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1796 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1797 session, the daemon ignores ECONNECT errors and logs others; it now
1798 ignores EPIPE as well.
1800 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1801 (quoted-printable decoding).
1803 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1804 later the small subsequent patch to fix an introduced bug.
1806 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1808 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1810 PH/08 An error is now given if message_size_limit is specified negative.
1812 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1813 to be given (somewhat) arbitrary names.
1815 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1818 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1819 miscellaneous code fixes
1821 PH/10 Added the log_reject_target ACL modifier to specify where to log
1824 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1825 hostname. This is wrong, because it relates to the incoming message (and
1826 probably the interface on which it is arriving) and not to the outgoing
1827 callout (which could be using a different interface). This has been
1828 changed to use the value of the helo_data option from the smtp transport
1829 instead - this is what is used when a message is actually being sent. If
1830 there is no remote transport (possible with a router that sets up host
1831 addresses), $smtp_active_hostname is used.
1833 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1834 tweaks were necessary in order to get it to work (see also 21 below):
1835 (a) The code assumed that strncpy() returns a negative number on buffer
1836 overflow, which isn't the case. Replaced with Exim's string_format()
1838 (b) There were several signed/unsigned issues. I just did the minimum
1839 hacking in of casts. There is scope for a larger refactoring.
1840 (c) The code used strcasecmp() which is not a standard C function.
1841 Replaced with Exim's strcmpic() function.
1842 (d) The code set only $1; it now sets $auth1 as well.
1843 (e) A simple test gave the error "authentication client didn't specify
1844 service in request". It would seem that Dovecot has changed its
1845 interface. Fortunately there's a specification; I followed it and
1846 changed what the client sends and it appears to be working now.
1848 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1851 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1852 address A is aliased to B and C, where B exists and C does not. Without
1853 -v the output is "A verified" because verification stops after a
1854 successful redirection if more than one address is generated. However,
1855 with -v the child addresses are also verified. Exim was outputting "A
1856 failed to verify" and then showing the successful verification for C,
1857 with its parentage. It now outputs "B failed to verify", showing B's
1858 parentage before showing the successful verification of C.
1860 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1861 look up a TXT record in a specific list after matching in a combined
1864 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1865 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1866 they consult the DNS. I had assumed they would set it the way they
1867 wanted; and indeed my experiments on Linux seem to show that in some
1868 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1869 To be on the safe side, however, I have now made the interface to
1870 host_find_byname() similar to host_find_bydns(), with an argument
1871 containing the DNS resolver options. The host_find_byname() function now
1872 sets these options at its start, just as host_find_bydns() does. The smtp
1873 transport options dns_qualify_single and dns_search_parents are passed to
1874 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1875 of host_find_byname() use the default settings of RES_DEFNAMES
1876 (qualify_single) but not RES_DNSRCH (search_parents).
1878 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1879 spool_read_header() do less string testing, by means of a preliminary
1880 switch on the second character of optional "-foo" lines. (This is
1881 overdue, caused by the large number of possibilities that now exist.
1882 Originally there were few.) While I was there, I also converted the
1883 str(n)cmp tests so they don't re-test the leading "-" and the first
1884 character, in the hope this might squeeze out yet more improvement.
1886 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1887 flag allowing group syntax was set by the header_syntax check but not
1888 turned off, possible causing trouble later; (2) The flag was not being
1889 set at all for the header_verify test, causing "group"-style headers to
1890 be rejected. I have now set it in this case, and also caused header_
1891 verify to ignore an empty address taken from a group. While doing this, I
1892 came across some other cases where the code for allowing group syntax
1893 while scanning a header line wasn't quite right (mostly, not resetting
1894 the flag correctly in the right place). These bugs could have caused
1895 trouble for malformed header lines. I hope it is now all correct.
1897 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1898 with the "reply" argument non-NULL. The code, however (which originally
1899 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1900 but it didn't always do it. This confused somebody who was copying the
1901 code for some other use. I have removed all the tests.
1903 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1904 feature that was used to support insecure browsers during the U.S. crypto
1905 embargo. It requires special client support, and Exim is probably the
1906 only MTA that supported it -- and would never use it because real RSA is
1907 always available. This code has been removed, because it had the bad
1908 effect of slowing Exim down by computing (never used) parameters for the
1909 RSA_EXPORT functionality.
1911 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1912 authenticator to fail if there's a tab character in the incoming data
1913 (there should never be unless someone is messing about, as it's supposed
1914 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1915 if the connection is using TLS or if the remote IP is the same as the
1916 local IP, and the "valid-client-cert option" if a client certificate has
1919 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1920 authenticators. This can be used for authorization after authentication
1921 succeeds. (In the case of plaintext, it servers for both authentication
1924 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1925 if any retry times were supplied.
1927 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1928 connection, where there is no client IP address to check. In this
1929 situation, the verify now always succeeds.
1931 PH/25 Applied John Jetmore's -Mset patch.
1933 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1935 PH/27 In a string expansion for a processed (not raw) header when multiple
1936 headers of the same name were present, leading whitespace was being
1937 removed from all of them, but trailing whitespace was being removed only
1938 from the last one. Now trailing whitespace is removed from each header
1939 before concatenation. Completely empty headers in a concatenation (as
1940 before) are ignored.
1942 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1943 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1945 PH/29 [Removed. This was a change that I later backed out, and forgot to
1946 correct the ChangeLog entry (that I had efficiently created) before
1947 committing the later change.]
1949 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1950 address errors (4xx response to RCPT) over the same connection as other
1951 messages routed to the same hosts. Such deliveries are always "forced",
1952 so retry times are not inspected. This resulted in far too many retries
1953 for the affected addresses. The effect occurred only when there were more
1954 hosts than the hosts_max_try setting in the smtp transport when it had
1955 the 4xx errors. Those hosts that it had tried were not added to the list
1956 of hosts for which the message was waiting, so if all were tried, there
1957 was no problem. Two fixes have been applied:
1959 (i) If there are any address or message errors in an SMTP delivery, none
1960 of the hosts (tried or untried) are now added to the list of hosts
1961 for which the message is waiting, so the message should not be a
1962 candidate for sending over the same connection that was used for a
1963 successful delivery of some other message. This seems entirely
1964 reasonable: after all the message is NOT "waiting for some host".
1965 This is so "obvious" that I'm not sure why it wasn't done
1966 previously. Hope I haven't missed anything, but it can't do any
1967 harm, as the worst effect is to miss an optimization.
1969 (ii) If, despite (i), such a delivery is accidentally attempted, the
1970 routing retry time is respected, so at least it doesn't keep
1971 hammering the server.
1973 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1974 in ${readsocket because some servers need this prod.
1976 PH/32 Added some extra debug output when updating a wait-xxx database.
1978 PH/33 The hint "could be header name not terminated by colon", which has been
1979 given for certain expansion errors for a long time, was not being given
1980 for the ${if def:h_colon_omitted{... case.
1982 PH/34 The spec says: "With one important exception, whenever a domain list is
1983 being scanned, $domain contains the subject domain." There was at least
1984 one case where this was not true.
1986 PH/35 The error "getsockname() failed: connection reset by peer" was being
1987 written to the panic log as well as the main log, but it isn't really
1988 panic-worthy as it just means the connection died rather early on. I have
1989 removed the panic log writing for the ECONNRESET error when getsockname()
1992 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1993 runs only) independently of the message's sender address. This meant
1994 that, if the 4xx error was in fact related to the sender, a different
1995 message to the same recipient with a different sender could confuse
1996 things. In particualar, this can happen when sending to a greylisting
1997 server, but other circumstances could also provoke similar problems.
1998 I have changed the default so that the retry time for these errors is now
1999 based a combination of the sender and recipient addresses. This change
2000 can be overridden by setting address_retry_include_sender=false in the
2003 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2004 remote server are returned as part of bounce messages. This was not
2005 happening for LMTP over a pipe (the lmtp transport), but now it is the
2006 same for both kinds of LMTP.
2008 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2009 in header lines that were in fact CNAMEs. This is no longer the case.
2011 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2012 and queue runs started by the daemon processed all messages. This has
2013 been fixed so that -R and -S can now usefully be given with -q<time>.
2015 PH/40 Import PCRE release 6.7 (fixes some bugs).
2017 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2019 PH/42 Give an error if -q is specified more than once.
2021 PH/43 Renamed the variables $interface_address and $interface_port as
2022 $received_ip_address and $received_port, to make it clear that these
2023 values apply to message reception, and not to the outgoing interface when
2024 a message is delivered. (The old names remain recognized, of course.)
2026 PH/44 There was no timeout on the connect() call when using a Unix domain
2027 socket in the ${readsocket expansion. There now is.
2029 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2030 be meaningful with "accept".
2032 SC/01 Eximstats V1.43
2033 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2035 SC/02 Eximstats V1.44
2036 Use a glob alias rather than an array ref in the generated
2037 parser. This improves both readability and performance.
2039 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2040 Collect SpamAssassin and rejection statistics.
2041 Don't display local sender or destination tables unless
2042 there is data to show.
2043 Added average volumes into the top table text output.
2045 SC/04 Eximstats V1.46
2046 Collect data on the number of addresses (recipients)
2047 as well as the number of messages.
2049 SC/05 Eximstats V1.47
2050 Added 'Message too big' to the list of mail rejection
2051 reasons (thanks to Marco Gaiarin).
2053 SC/06 Eximstats V1.48
2054 Mainlog lines which have GMT offsets and are too short to
2055 have a flag are now skipped.
2057 SC/07 Eximstats V1.49 (Alain Williams)
2058 Added the -emptyok flag.
2060 SC/08 Eximstats V1.50
2061 Fixes for obtaining the IP address from reject messages.
2063 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2064 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2065 whitesspace changes from 4.64-PH/27
2067 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2070 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2071 are found, allow negative numbers in numeric criteria)
2073 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2075 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2076 to match changes made in 4.64-PH/43
2078 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2080 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2081 "domain" is actually an IP address, and force a failure. This locks out
2082 those revolvers/nameservers that support "A-for-A" lookups, in
2083 contravention of the specifications.
2085 PH/48 When a host name was looked up from an IP address, and the subsequent
2086 forward lookup of the name timed out, the host name was left in
2087 $sender_host_name, contrary to the specification.
2089 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2090 restricted to single-key lookups, Exim was not diagnosing an error if
2091 * or *@ was used with a query-style lookup.
2093 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2095 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2096 long ago, but noone interested enough thought of it. Let's just say that
2097 the "1.1" means that there are some new functions that weren't there at
2098 some point in the past.
2100 PH/51 Error processing for expansion failure of helo_data from an smtp
2101 transport during callout processing was broken.
2103 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2104 tested/used via the -bh/-bhc/-bs options.
2106 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2107 bug, fixed in subsequent PCRE releases).
2109 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2110 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2112 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2118 SC/01 Use a glob alias rather than an array ref in eximstats generated
2119 parser. This improves both readability and performance.
2121 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2122 Don't display local sender or destination tables in eximstats unless
2123 there is data to show.
2124 Added average volumes into the eximstats top table text output.
2126 SC/03 Collect data on the number of addresses (recipients) as well
2127 as the number of messages in eximstats.
2129 TF/01 Correct an error in the documentation for the redirect router. Exim
2130 does (usually) call initgroups() when daemonizing.
2132 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2133 with consistent privilege compared to when running as a daemon.
2135 TF/03 Note in the spec that $authenticated_id is not set for local
2136 submissions from trusted users.
2138 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2139 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2141 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2142 by adding some example configuration directives to the default
2143 configuration file. A little bit of work is required to uncomment the
2144 directives and define how usernames and passwords are checked, but
2145 there is now a framework to start from.
2147 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2148 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2149 without this. I don't know how relevant this is to other LDAP libraries.
2151 PH/02 Add the verb name to the "unknown ACL verb" error.
2153 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2155 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2157 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2158 directory not expanded when it should be if an expanded home directory
2159 was set for the address (which is overridden by the transport).
2161 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2164 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2165 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2166 because it is too late at that time, and has no effect.
2168 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2169 security issue with \' (bugzilla #107). I could not use the
2170 PQescapeStringConn() function, because it needs a PGconn value as one of
2173 PH/08 When testing addresses using -bt, indicate those final addresses that
2174 are duplicates that would not cause an additional delivery. At least one
2175 person was confused, thinking that -bt output corresponded to deliveries.
2176 (Suppressing duplicates isn't a good idea as you lose the information
2177 about possibly different redirections that led to the duplicates.)
2179 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2180 systems where poll() doesn't work, in particular OS X.
2182 PH/10 Added more information to debugging output for retry time not reached.
2184 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2185 operations in malware.c.
2187 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2190 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2191 syslog_duplication set false, log lines that would normally be written
2192 both the the main log and to the reject log were not written to syslog at
2195 PH/14 In the default configuration, change the use of "message" in ACL warn
2196 statements to "add_header".
2198 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2199 not followed by a command (e.g. "seen endif").
2201 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2202 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2205 PH/17 Added extra conditions to the default value of delay_warning_condition
2209 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2210 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2211 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2214 The Auto-Submitted: and various List- headers are standardised, whereas I
2215 don't think Precedence: ever was.
2217 PH/18 Refactored debugging code in route_finduser() to show more information,
2218 in particular, the error code if getpwnam() issues one.
2220 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2221 This is apparently needed in addition to the PH/07 change above to avoid
2222 any possible encoding problems.
2224 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2225 but not after initializing Perl.
2227 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2228 output them only if debugging. By default they are written stderr,
2229 apparently, which is not desirable.
2231 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2234 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2237 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2239 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2240 authenticated or an ident call has been made. Suppress the default
2241 values for $authenticated_id and $authenticated_sender (but permit -oMai
2242 and -oMas) when testing with -bh.
2244 PH/24 Re-jigged the order of the tests in the default configuration so that the
2245 tests for valid domains and recipients precede the DNS black list and CSA
2246 tests, on the grounds that those ones are more expensive.
2248 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2249 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2250 This bug exists in every version of Exim that I still have, right back to
2253 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2254 However, an attempt to turn on case-sensitivity in a regex key by
2255 including (?-i) didn't work because the subject string was already
2256 lowercased, and the effects were non-intuitive. It turns out that a
2257 one-line patch can be used to allow (?-i) to work as expected.
2263 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2264 other effects) broke the use of negated acl sub-conditions.
2266 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2269 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2270 "Deny" causes Exim to reject the incoming connection with a 554 error.
2271 Unfortunately, if there is a major crisis, such as a disk failure,
2272 tcp-wrappers gives "deny", whereas what one would like would be some
2273 kind of temporary error. A kludge has been added to help with this.
2274 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2275 554 error is used if errno is still zero or contains ENOENT (which occurs
2276 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2279 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2281 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2282 errors. Otherwise a message that provokes a temporary error (when other
2283 messages do not) can cause a whole host to time out.
2285 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2286 addresses were routed directly to files or pipes from a redirect router.
2287 File deliveries just didn't batch; pipe deliveries might have suffered
2290 PH/06 A failure to get a lock for a hints database would erroneously always say
2291 "Failed to get write lock", even when it was really a read lock.
2293 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2294 of 0600. This has been changed to use the value of the lockfile_mode
2295 option (which defaults to 0600).
2297 PH/08 Applied small patch from the Sieve maintainer.
2299 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2300 folder from quota calculations, a direct delivery into this folder messed
2301 up the contents of the maildirsize file. This was because the regex was
2302 used only to exclude .Trash (or whatever) when the size of the mailbox
2303 was calculated. There was no check that a delivery was happening into an
2304 excluded directory. This bug has been fixed by ignoring all quota
2305 processing for deliveries into excluded directories.
2307 PH/10 Added the maildirfolder_create_regex option to appendfile.
2313 PH/01 The code for finding all the local interface addresses on a FreeBSD
2314 system running IPv6 was broken. This may well have applied to all BSD
2315 systems, as well as to others that have similar system calls. The broken
2316 code found IPv4 interfaces correctly, but gave incorrect values for the
2317 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2318 that it would not match correctly against @[] and not recognize the IPv6
2321 PH/02 The ipliteral router was not recognizing addresses of the form user@
2322 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2324 PH/03 Added disable_ipv6.
2326 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2327 decoded form, because it is most often used to construct To: headers
2328 lines in autoreplies, and the decoded form may well be syntactically
2329 invalid. However, $reply_address has leading white space removed, and all
2330 newlines turned into spaces so that the autoreply transport does not
2333 PH/05 If group was specified without a user on a router, and no group or user
2334 was specified on a transport, the group from the router was ignored.
2336 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2337 for visible compile-time settings that can be used to change these
2338 numbers, for those that want even more. Backwards compatibility with old
2339 spool files has been maintained. However, going back to a previous Exim
2340 release will lost any variables that are in spool files.
2342 PH/07 Two small changes when running in the test harness: increase delay when
2343 passing a TCP/IP connection to a new process, in case the original
2344 process has to generate a bounce, and remove special handling of
2345 127.0.0.2 (sic), which is no longer necessary.
2347 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2348 be the same on different OS.
2350 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2353 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2354 whether --show-vars was specified or not
2356 JJ/02 exipick: Added support for new ACL variable spool format introduced
2359 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2360 syntactically invalid From: or Reply-to: line, and a filter used this to
2361 generate an autoreply, and therefore failed to obtain an address for the
2362 autoreply, Exim could try to deliver to a non-existent relative file
2363 name, causing unrelated and misleading errors. What now happens is that
2364 it logs this as a hard delivery error, but does not attempt to create a
2367 PH/11 The exinext utility has a -C option for testing purposes, but although
2368 the given file was scanned by exinext itself; it wasn't being passed on
2369 when Exim was called.
2371 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2372 an end-of-file indication when reading a command response.
2374 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2375 compiled. In many other places in Exim, IPv6 addresses are always
2376 recognized, so I have changed this. It also means that IPv4 domain
2377 literals of the form [IPV4:n.n.n.n] are now always recognized.
2379 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2380 used if the router is not running as root, for example, when verifying at
2381 ACL time, or when using -bh. The debugging output from this situation was
2382 non-existent - all you got was a failure to exec. I have made two
2385 (a) Failures to set uid/gid, the current directory, or a process leader
2386 in a subprocess such as that created by queryprogram now generate
2387 suitable debugging ouput when -d is set.
2389 (b) The queryprogram router detects when it is not running as root,
2390 outputs suitable debugging information if -d is set, and then runs
2391 the subprocess without attempting to change uid/gid.
2393 PH/15 Minor change to Makefile for building test_host (undocumented testing
2396 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2397 additional section of a DNS packet that returns MX or SRV records.
2398 Instead, it always explicitly searches for A/AAAA records. This avoids
2399 major problems that occur when a DNS server includes only records of one
2400 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2401 fixed another bug: if SRV records were looked up and the corresponding
2402 address records were *not* found in the additional section, the port
2403 values from the SRV records were lost.
2405 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2406 using the correct key (the original address) when searching the retry
2407 rules in order to find which one to use for generating the retry hint.
2409 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2410 adding the default one. Similarly, if it contains a Reply-To: header, the
2411 errors_reply_to option, if set, is not used.
2413 PH/19 When calculating a retry time, Exim used to measure the "time since
2414 failure" by looking at the "first failed" field in the retry record. Now
2415 it does not use this if it is later than than the arrival time of the
2416 message. Instead it uses the arrival time. This makes for better
2417 behaviour in cases where some deliveries succeed, thus re-setting the
2418 "first failed" field. An example is a quota failure for a huge message
2419 when small messages continue to be delivered. Without this change, the
2420 "time since failure" will always be short, possible causing more frequent
2421 delivery attempts for the huge message than are intended.
2422 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2424 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2425 $1, $2, $3) because the numerical variables can be reset during some
2426 expansion items (e.g. "match"), thereby losing the authentication data.
2428 PH/21 Make -bV show the size of off_t variables so that the test suite can
2429 decide whether to run tests for quotas > 2G.
2431 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2432 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2433 filecount value is greater than 2G or if a quota value is greater than 2G
2434 on a system where the size of off_t is not greater than 4, a panic error
2437 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2438 never match. The debug and -bh output now contains an explicit error
2439 message indicating a malformed IPv4 address or mask.
2441 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2442 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2443 PH/23 above applies.
2445 PH/25 Do not write to syslog when running in the test harness. The only
2446 occasion when this arises is a failure to open the main or panic logs
2447 (for which there is an explicit test).
2449 PH/26 Added the /no_tell option to "control=freeze".
2451 PH/27 If a host name lookup failed very early in a connection, for example, if
2452 the IP address matched host_lookup and the reverse lookup yielded a name
2453 that did not have a forward lookup, an error message of the form "no IP
2454 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2455 could be logged. Now it outputs the IP address instead of "NULL".
2457 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2458 allows the sender and the authenticated sender to be set when
2459 submitting a message from within Exim. Since child_open_exim() is
2460 documented for local_scan(), the new function should be too.
2462 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2463 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2464 results in an empty string is now treated as unset.
2466 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2468 PH/31 Added sender_verify_fail logging option.
2470 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2471 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2472 tidied the source and removed it altogether.
2474 PH/33 When a queue run was abandoned because the load average was too high, a
2475 log line was always written; now it is written only if the queue_run log
2476 selector is set. In addition, the log line for abandonment now contains
2477 information about the queue run such as the pid. This is always present
2478 in "start" and "stop" lines but was omitted from the "abandon" line.
2480 PH/34 Omit spaces between a header name and the colon in the error message that
2481 is given when verify = headers_syntax fails (if there are lots of them,
2482 the message gets confusing).
2484 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2485 names, as there are now some PTR records that contain slashes. This check
2486 is only to protect against broken name servers that fall over on strange
2487 characters, so the fact that it applies to all lookups doesn't matter.
2489 PH/36 Now that the new test suite is complete, we can remove some of the
2490 special code in Exim that was needed for the old test suite. For example,
2491 sorting DNS records because real resolvers return them in an arbitrary
2492 order. The new test suite's fake resolver always returns records in the
2495 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2496 bounces) except when queue_only is set, to avoid logging races between
2497 the different processes.
2499 PH/38 Panic-die if .include specifies a non-absolute path.
2501 PH/39 A tweak to the "H" retry rule from its user.
2503 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2504 a label. They prevented compilation on older perls.
2506 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2507 a warning to be raised on newish perls.
2509 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2510 on queue. Changes to match documented behaviour of showing count of
2511 messages matching specified criteria.
2513 PH/40 Changed the default ident timeout from 30s to 5s.
2515 PH/41 Added support for the use of login_cap features, on those BSD systems
2516 that have them, for controlling the resources used by pipe deliveries.
2518 PH/42 The content-scanning code uses fopen() to create files in which to put
2519 message data. Previously it was not paying any attention to the mode of
2520 the files. Exim runs with umask(0) because the rest of the code creates
2521 files with open(), and sets the required mode explicitly. Thus, these
2522 files were ending up world-writeable. This was not a big issue, because,
2523 being within the spool directory, they were not world-accessible. I have
2524 created a function called modefopen, which takes an additional mode
2525 argument. It sets umask(777), creates the file, chmods it to the required
2526 mode, then resets the umask. All the relevant calls to fopen() in the
2527 content scanning code have been changed to use this function.
2529 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2530 to 24 hours. This avoids potential overflow problems when processing G
2531 and H retry rules. I suspect nobody ever tinkers with this value.
2533 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2535 PH/45 When the plaintext authenticator is running as a client, the server's
2536 challenges are checked to ensure they are valid base64 strings. By
2537 default, the authentication attempt is cancelled if an invalid string is
2538 received. Setting client_ignore_invalid_base64 true ignores these errors.
2539 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2540 they are received. Thus, the responses can be made to depend on the
2541 challenges. If an invalid string is ignored, an empty string is placed in
2544 PH/46 Messages that are created by the autoreply transport now contains a
2545 References: header, in accordance with RFCs 2822 and 3834.
2547 PH/47 Added authenticated_sender_force to the smtp transport.
2549 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2551 PH/49 Installed latest patch from the Sieve maintainer.
2553 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2554 was also set, the appendfile transport was unnecessarily scanning a
2555 directory of message files (e.g. for maildir delivery) to find the count
2556 of files (along with the size), even though it did not need this
2557 information. It now does the scan only if it needs to find either the
2558 size of the count of files.
2560 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2562 PH/52 Two bugs concerned with error handling when the smtp transport is
2565 (i) Exim was not creating retry information for temporary errors given
2566 for individual recipients after the DATA command when the smtp transport
2567 was used in LMTP mode. This meant that they could be retried too
2568 frequently, and not timed out correctly.
2570 (ii) Exim was setting the flag that allows error details to be returned
2571 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2572 recipients that were returned after the DATA command.
2574 PH/53 This is related to PH/52, but is more general: for any failing address,
2575 when detailed error information was permitted to be returned to the
2576 sender, but the error was temporary, then after the final timeout, only
2577 "retry timeout exceeded" was returned. Now it returns the full error as
2578 well as "retry timeout exceeded".
2580 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2581 do this, and (what is worse) MTAs that accept it.
2583 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2584 will now be deprecated.
2586 PH/56 New os.c-cygwin from the Cygwin maintainer.
2588 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2589 formats (previously only available in exim formats via -bpr, -bpru,
2590 and -bpra. Now also available in native and exiqgrep formats)
2592 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2593 with very large, slow to parse queues
2595 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2597 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2599 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2600 responses to authentication challenges, though it was showing the
2601 challenges; (ii) I've removed the CR characters from the debug output for
2604 PH/58 Allow for the insertion of a newline as well as a space when a string
2605 is turned into more than one encoded-word during RFC 2047 encoding. The
2606 Sieve code now uses this.
2608 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2609 data_4xx, lost_connection, tls_required.
2611 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2612 message was being sent as an SMTP response.
2614 PH/61 Add -l and -k options to exicyclog.
2616 PH/62 When verifying, if an address was redirected to one new address, so that
2617 verification continued, and the new address failed or deferred after
2618 having set something in $address_data, the value of $address_data was not
2619 passed back to the ACL. This was different to the case when no
2620 redirection occurred. The value is now passed back in both cases.
2622 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2623 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2624 use login_cap.h, so on its own it isn't the distinguishing feature. The
2625 new name refers directly to the setclassresources() function.
2627 PH/65 Added configuration files for NetBSD3.
2629 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2631 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2632 is preferred over IPv4.
2634 PH/68 The bounce_return_message and bounce_return_body options were not being
2635 honoured for bounces generated during the reception of non-SMTP messages.
2636 In particular, this applied to messages rejected by the ACL. This bug has
2637 been fixed. However, if bounce_return_message is true and bounce_return_
2638 body is false, the headers that are returned for a non-SMTP message
2639 include only those that have been read before the error was detected.
2640 (In the case of an ACL rejection, they have all been read.)
2642 PH/69 The HTML version of the specification is now built in a directory called
2643 spec_html instead of spec.html, because the latter looks like a path with
2644 a MIME-type, and this confuses some software.
2646 PH/70 Catch two compiler warnings in sieve.c.
2648 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2649 function verify_get_ident() calls ip_connect() to connect a socket, but
2650 if the "connect()" function timed out, ip_connect() used to close the
2651 socket. However, verify_get_ident() also closes the socket later, and in
2652 between Exim writes to the log, which may get opened at this point. When
2653 the socket was closed in ip_connect(), the log could get the same file
2654 descriptor number as the socket. This naturally causes chaos. The fix is
2655 not to close the socket in ip_connect(); the socket should be closed by
2656 the function that creates it. There was only one place in the code where
2657 this was missing, in the iplookup router, which I don't think anybody now
2658 uses, but I've fixed it anyway.
2660 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2661 well as to direct DNS lookups. Otherwise the handling of names in host
2662 lists is inconsistent and therefore confusing.
2668 PH/01 Two changes to the default runtime configuration:
2670 (1) Move the checks for relay_from_hosts and authenticated clients from
2671 after to before the (commented out) DNS black list checks.
2673 (2) Add control=submission to the relay_from_hosts and authenticated
2674 clients checks, on the grounds that messages accepted by these
2675 statements are most likely to be submissions.
2677 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2679 (1) Generate an error if the third argument for the ${prvs expansion is
2682 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2685 (3) Reset the variables that are obtained from the first argument of
2686 ${prvscheck and used in the second argument before leaving the code,
2687 because their memory is reclaimed, so using them afterwards may do
2690 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2691 one (it's much easier than Tom thought :-).
2693 (5) Because of (4), we can now allow for the use of $prvscheck_result
2694 inside the third argument.
2696 PH/03 For some reason, the default setting of PATH when running a command from
2697 a pipe transport was just "/usr/bin". I have changed it to
2700 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2701 anything to be listed in the output from -bV.
2703 PH/05 When a filter generated an autoreply, the entire To: header line was
2704 quoted in the delivery log line, like this:
2706 => >A.N.Other <ano@some.domain> <original@ddress> ...
2708 This has been changed so that it extracts the operative address. There
2709 may be more than one such address. If so, they are comma-separated, like
2712 => >ano@some.domain,ona@other.domain <original@ddress> ...
2714 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2715 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2716 not being looked up in the rDNS to get a host name, Exim was showing the
2717 IP address twice in Received: lines, even though the IP addresses were
2718 identical. For example:
2720 Received: from [1.2.3.4] (helo=[1.2.3.4])
2722 However, if the real host name was known, it was omitting the HELO data
2723 if it matched the actual IP address. This has been tidied up so that it
2724 doesn't show the same IP address twice.
2726 PH/07 When both +timestamp and +memory debugging was on, the value given by
2727 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2728 called by the debug printing, thereby overwriting the timestamp buffer.
2729 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2731 PH/08 When the original message was included in an autoreply transport, it
2732 always said "this is a copy of the message, including all the headers",
2733 even if body_only or headers_only was set. It now gives an appropriate
2736 PH/09 Applied a patch from the Sieve maintainer which:
2738 o fixes some comments
2739 o adds the (disabled) notify extension core
2740 o adds some debug output for the result of if/elsif tests
2741 o points to the current vacation draft in the documentation
2742 and documents the missing references header update
2746 o fixes a bug in processing the envelope test (when testing
2747 multiple envelope elements, the last element determinted the
2750 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2751 Electronic Mail") by including:
2753 Auto-submitted: auto-generated
2755 in the messages that it generates (bounce messages and others, such as
2756 warnings). In the case of bounce messages for non-SMTP mesages, there was
2757 also a typo: it was using "Auto_submitted" (underscore instead of
2758 hyphen). Since every message generated by Exim is necessarily in response
2759 to another message, thes have all been changed to:
2761 Auto-Submitted: auto-replied
2763 in accordance with these statements in the RFC:
2765 The auto-replied keyword:
2767 - SHOULD be used on messages sent in direct response to another
2768 message by an automatic process,
2770 - MUST NOT be used on manually-generated messages,
2772 - MAY be used on Delivery Status Notifications (DSNs) and Message
2773 Disposition Notifications (MDNs),
2775 - MUST NOT be used on messages generated by automatic or periodic
2776 processes, except for messages which are automatic responses to
2779 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2780 to the default Received: header definition.
2782 PH/12 Added log selector acl_warn_skipped (default on).
2784 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2785 variables because (a) they are in the wrong storage pool and (b) even if
2786 they were copied, it wouldn't work properly because of the caching.
2788 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2789 checking when decoding. Apparently there are clients that generate
2790 overlong encoded strings. Why am I not surprised?
2792 PH/15 If the first argument of "${if match_address" was not empty, but did not
2793 contain an "@" character, Exim crashed. Now it writes a panic log message
2794 and treats the condition as false.
2796 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2798 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2799 extension "envelope-auth". The code is finished and in agreement with
2800 other implementations, but there is no documentation so far and in fact,
2801 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2802 not changing the active code.
2804 Print executed "if" and "elsif" statements when debugging is used. This
2805 helps a great deal to understand what a filter does.
2807 Document more things not specified clearly in RFC3028. I had all this
2808 sorted out, when out of a sudden new issues came to my mind. Oops."
2810 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2813 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2814 canonical form (as documented). However, after a host name lookup from
2815 the IP address, check_host() was doing a simple string comparison with
2816 addresses acquired from the DNS when checking that the found name did
2817 have the original IP as one of its addresses. Since any found IPv6
2818 addresses are likely to be in abbreviated form, the comparison could
2819 fail. Luckily, there already exists a function for doing the comparison
2820 by converting both addresses to binary, so now that is used instead of
2821 the text comparison.
2823 PH/20 There was another similar case to PH/19, when a complete host name was
2824 given in a host list; looking up its IP address could give an abbreviated
2825 form, whereas the current host's name might or might not be abbreviated.
2826 The same fix has been applied.
2832 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2833 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2836 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2837 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2839 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2841 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2842 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2843 Also, for "Auto-Submitted", treat anything other than "no" as
2844 non-personal, in accordance with RFC 3834. (Previously it treated
2845 anything starting "auto-" as non-personal.)
2847 TF/01 The control=submission/name=... option had a problem with syntax
2848 errors if the name included a slash character. The /name= option
2849 now slurps the rest of the string, so it can include any characters
2850 but it must come last in the list of options (after /sender_retain
2853 PH/05 Some modifications to the interface to the fake nameserver for the new
2861 TK/01 Added the "success_on_redirect" address verification option. See
2862 NewStuff for rationale and an example.
2864 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2866 PH/02 Patch to exigrep to allow it to work on syslog lines.
2868 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2869 fread() to skip over the body file's header line, because in Cygwin the
2870 header line is locked and is inaccessible.
2872 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2873 co-exist for some time) to make it clear that it is the Exim ID that is
2874 referenced, not the Message-ID: header line.
2876 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2877 string_format() function, because snprintf() does not exist on all
2880 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2881 ${stat: expansion item.
2883 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2884 protocol synchronization error", to keep the pedants happy.
2886 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2887 well as for IRIX systems, when gcc is being used. See the host.c source
2890 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2892 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2895 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2896 transport and to the smtp transport in LMTP mode.
2898 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2900 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2901 run a filter in a subprocess. This could lead to confusion in subsequent
2902 lookups in the parent process. There should also be a search_tidyup() at
2903 the end of the subprocess.
2905 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2906 only if the host matched helo_try_verify_hosts, which caused the
2907 verification to occur when the EHLO/HELO command was issued. The ACL just
2908 tested the remembered result. Now, if a previous verification attempt has
2909 not happened, "verify = helo" does it there and then.
2911 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2913 TK/03 Fix log output including CR from clamd.
2915 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2916 exist provoked a memory error which could cause a segfault.
2918 PH/15 Installed PCRE 6.2
2920 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2922 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2923 of the problem. Specifically, suggested +O2 rather than +O1 for the
2926 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2928 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2929 record, the port was not correctly logged when the outgoing_port log
2930 selector was set (it logged the transort's default port).
2932 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2933 fallback_hosts, and "hosts" in the smtp transport.
2935 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2936 host errors such as "Connection refused".
2938 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2939 authentication with radiusclient 0.4.9:
2941 - Error returned from rc_read_config was caught wrongly
2942 - Username/password not passed on to radius server due to wrong length.
2944 The presumption is that some radiusclient API changes for 4.51/PH/17
2945 were not taken care of correctly. The code is still untested by me (my
2946 Linux distribution still has 0.3.2 of radiusclient), but it was
2947 contributed by a Radius user.
2949 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2950 expanding the "port" option of the smtp transport.
2952 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2953 while reading a MIME header. Thanks to Tom Hughes for a patch.
2955 PH/24 Include config.h inside local_scan.h so that configuration settings are
2958 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2959 that in an ACL for RCPT (for example), you can examine exactly what was
2962 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2963 commands, but it was not correctly comparing the address with the actual
2964 client host address. Thus, it would show the EHLO address in Received:
2965 header lines when this was not necessary.
2967 PH/27 Added the % operator to ${eval:}.
2969 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2970 it should be ignoring failures (because with -C, for example, it has lost
2971 privilege). It wasn't ignoring creation failures other than "already
2974 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2977 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2978 because an input line was too long, either on its own, or by virtue of
2979 too many continuations, the temporary file was not being removed, and the
2980 return code was incorrect.
2982 PH/31 Missing "BOOL" in function definition in filtertest.c.
2984 PH/32 Applied Sieve patches from the maintainer.
2986 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2988 PH/33 Added "verify = not_blind".
2990 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2991 Local/Makefile (with some defaults set). These are used in built scripts
2992 such as exicyclog, but they have never been used in the exim_install
2993 script (though there are many overriding facilities there). I have
2994 arranged that the exim_install script now takes note of these two
2997 PH/35 Installed configuration files for Dragonfly.
2999 PH/36 When a locally submitted message by a trusted user did not contain a
3000 From: header, and the sender address was obtained from -f or from an SMTP
3001 MAIL command, and the trusted user did not use -F to supply a sender
3002 name, $originator_name was incorrectly used when constructing a From:
3003 header. Furthermore, $originator_name was used for submission mode
3004 messages from external hosts without From: headers in a similar way,
3005 which is clearly wrong.
3007 PH/37 Added control=suppress_local_fixups.
3009 PH/38 When log_selector = +received_sender was set, and the addition of the
3010 sender made the log line's construction buffer exactly full, or one byte
3011 less than full, an overflow happened when the terminating "\n" was
3014 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3015 when the result of a list match is failure because a DNS lookup failed.
3017 PH/40 RM_COMMAND is now used in the building process.
3019 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3020 the "build-* directories that it finds.
3022 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3023 address in a domain literal was a prefix of an interface address.
3025 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3026 when verifying a sender address, unless rewrite_headers is false.
3028 PH/44 Wrote a long comment about why errors_to addresses are verified as
3029 recipients, not senders.
3031 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3032 the ratelimit ACL was added.
3034 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3036 PH/46 Added extra information about PostgreSQL errors to the error string.
3038 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3039 avoiding the need to install special zones in a real server. This is
3040 backwards compatible; if it can't find the fake resolver, it drops back.
3041 Thus, both old and new test suites can be run.
3043 TF/02 Added util/ratelimit.pl
3045 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3046 clock is set back in time.
3048 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3049 Candler <B.Candler@pobox.com>.
3051 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3052 OK for addresses that are the result of redirections.
3054 PH/48 A number of further additions for the benefit of the new test suite,
3055 including a fake gethostbyname() that interfaces to the fake DNS resolver
3058 TF/06 The fix for widen_domains has also been applied to qualify_single and
3059 search_parents which are the other dnslookup options that can cause
3062 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3065 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3067 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3068 service" if it encountered an MX record with an empty target hostname.
3069 The message is now "an MX or SRV record indicated no SMTP service".
3071 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3072 if the DNS of the sending site is misconfigured. This is quite a
3073 common situation. This change restores the behaviour of treating a
3074 helo verification defer as a failure.
3076 PH/51 If self=fail was set on a router, the bounce message did not include the
3077 actual error message.
3083 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3085 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3086 command itself ended in error, the underlying message about the transport
3087 filter timeout was being overwritten with the pipe command error. Now the
3088 underlying error message should be appended to the second error message.
3090 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3092 PH/02 Reduce the amount of output that "make" produces by default. Full output
3093 can still be requested.
3095 PH/03 The warning log line about a condition test deferring for a "warn" verb
3096 was being output only once per connection, rather than after each
3097 occurrence (because it was using the same function as for successful
3098 "warn" verbs). This seems wrong, so I have changed it.
3100 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3101 it should not have, which might have caused a crash in the right
3102 circumstances, but probably never did.
3104 PH/04 Installed a modified version of Tony Finch's patch to make submission
3105 mode fix the return path as well as the Sender: header line, and to
3106 add a /name= option so that you can make the user's friendly name appear
3109 TF/03 Added the control = fakedefer ACL modifier.
3111 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3112 Mark Lowes for thorough testing.
3114 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3116 TK/03 Merged latest SRS patch from Miles Wilton.
3118 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3119 with the definition in sysexits.h (which is #included earlier).
3120 Fortunately, Exim does not actually use EX_OK. The code used to try to
3121 preserve the sysexits.h value, by assumimg that macro definitions were
3122 scanned for macro replacements. I have been disabused of this notion,
3123 so now the code just undefines EX_OK before #including unistd.h.
3125 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3126 in the smtp transport. When a block could not be written in a single
3127 write() function, the timeout was being re-applied to each part-write.
3128 This seems wrong - if the receiver was accepting one byte at a time it
3129 would take for ever. The timeout is now adjusted when this happens. It
3130 doesn't have to be particularly precise.
3132 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3133 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3135 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3136 without a domain if the check to <postmaster@domain> fails.
3138 SC/01 Eximstats: added -xls and the ability to specify output files
3139 (patch written by Frank Heydlauf).
3141 SC/02 Eximstats: use FileHandles for outputing results.
3143 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3145 SC/04 Eximstats: fixed display of large numbers with -nvr option
3147 SC/05 Eximstats: fixed merging of reports with empty tables.
3149 SC/06 Eximstats: added the -include_original_destination flag
3151 SC/07 Eximstats: removed tabs and trailing whitespace.
3153 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3155 TK/06 MBOX spool code: Add real "From " MBOX separator line
3156 so the .eml file is really in mbox format (even though
3157 most programs do not really care). Patch from Alex Miller.
3159 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3160 The latter is generated from $received_to and is only set if the
3161 message has one envelope recipient. SA can use these headers,
3162 obviously out-of-the-box. Patch from Alex Miller.
3164 PH/08 The ${def test on a variable was returning false if the variable's
3165 value was "0", contrary to what the specification has always said!
3166 The result should be true unless the variable is empty.
3168 PH/09 The syntax error of a character other than { following "${if
3169 def:variable_name" (after optional whitespace) was not being diagnosed.
3170 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3171 accidental colon was present, for example, could give incorrect results.
3173 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3174 result is used (not including appendfile, where other changes are about
3177 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3178 This involved changing a lot of size variables from int to off_t. It
3179 should work with maildirs and everything.
3181 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3182 spamd dying while we are connected to it.
3184 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3187 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3188 function for BDB 4.3.
3190 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3192 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3193 the sources for PCRE have been re-arranged and more files are now
3196 PH/15 The code I had for printing potentially long long variables in PH/11
3197 above was not the best (it lost precision). The length of off_t variables
3198 is now inspected at build time, and an appropriate printing format (%ld
3199 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3200 to be "long long int" or "long int". This is needed for the internal
3201 formatting function string_vformat().
3203 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3204 the configuration file to be ":syslog", then the script "guesses" where
3205 the logs files are, rather than using the compiled in default. In our
3206 case the guess is not the same as the compiled default, so the script
3207 suddenly stopped working when I started to use syslog. The patch checks
3208 to see if log_file_path is "". If so, it attempts to read it from exim
3209 with no configuration file to get the compiled in version, before it
3210 falls back to the previous guessing code."
3212 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3213 implementing BATV in an Exim configuration. See NewStuff for the gory
3216 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3217 Makefile that are specific to HP-UX.
3219 PH/18 If the "use_postmaster" option was set for a recipient callout together
3220 with the "random" option, the postmaster address was used as the MAIL
3221 FROM address for the random test, but not for the subsequent recipient
3222 test. It is now used for both.
3224 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3225 patch removes a few documentation additions to RFC 3028, because the
3226 latest draft now contains them. It adds the new en;ascii-case comparator
3227 and a new error check for 8bit text in MIME parts. Comparator and
3228 require names are now matched exactly. I enabled the subaddress
3229 extension, but it is not well tested yet (read: it works for me)."
3231 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3232 rework some of the code of TK/09 above to avoid the hardwired use of
3233 "%lld" and "long long". Replaced the call to snprintf() with a call to
3236 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3237 records point to non-existent hosts", "retry timeout exceeded", and
3238 "retry time not reached for any host after a long failure period".
3240 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3241 experimental DomainKeys support:
3243 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3244 (2) On an error such as an illegally used "control", the wrong name for
3245 the control was given.
3247 These problems did NOT occur unless DomainKeys support was compiled.
3249 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3251 PH/24 Added ${if match_ip condition.
3253 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3254 databases so that it will be absolutely obvious if a crash occurs in the
3255 DB library. This is a regular occurrence (often caused by mis-matched
3258 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3259 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3260 picked up on a user's system that detects such things. There doesn't seem
3261 to be a gcc warning option for this - only an attribute that has to be
3262 put on the function's prototype. It seems that in Fedora Core 4 they have
3263 set this on a number of new functions. No doubt there will be more in due
3266 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3267 specify a transport. However, if an address that was verified by such a
3268 router was the subject of a callout, Exim crashed because it tried to
3269 read the rcpt_include_affixes from the non-existent transport. Now it
3270 just assumes that the setting of that option is false. This bug was
3271 introduced by 4.51/PH/31.
3273 PH/28 Changed -d+all to exclude +memory, because that information is very
3274 rarely of interest, but it makes the output a lot bigger. People tend to
3275 do -d+all out of habit.
3277 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3278 code in os-type was giving problems when libc.so lives in lib64, like on
3281 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3282 aren't the modern standard, and it seems that some systems' include files
3283 don't always have them. Exim was already checking for some of the newer
3284 ones like T_AAAA, and defining it itself. I've added checks for all the
3285 record types that Exim uses.
3287 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3288 not automatically generating a new one, as it is supposed to. This
3289 prevented TLS from working. If the file did exist, but contained invalid
3290 data, a new version was generated, as expected. It was only the case of a
3291 non-existent file that was broken.
3293 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3294 with a change in libdomainkeys > 0.64.
3296 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3297 from DNS. If the selector record carries the flag, it now has
3298 precedence over the domain-wide flag.
3300 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3302 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3303 the use of a port name that isn't defined in /etc/services) occurred, the
3304 message was deferred as in a normal delivery, and thus remained on the
3305 spool, instead of being failed because of the mua_wrapper setting. This
3306 is now fixed, and I tidied up some of the mua_wrapper messages at the
3309 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3310 the messages in a hash of arrays rather than using individual hashes.
3311 This is a bit cleaner and results in dramatic memory savings, albeit
3312 at a slight CPU cost.
3314 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3315 as requested by Marc Sherman.
3317 SC/10 Eximstats: added histograms for user specified patterns as requested
3320 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3322 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3323 fopen() in the content-scanning modules that did not already have it.
3329 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3330 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3332 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3334 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3336 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3337 to test Sieve filters that use "vacation".
3339 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3340 that changes the way the GnuTLS parameters are stored in the cache file.
3341 The new format can be generated externally. For backward compatibility,
3342 if the data in the cache doesn't make sense, Exim assumes it has read an
3343 old-format file, and it generates new data and writes a new file. This
3344 means that you can't go back to an older release without removing the
3347 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3348 work if there are any delivery delays because "one_time" forces the
3349 parent to be marked "delivered", so its unseen clone is never tried
3350 again. For this reason, Exim now forbids the simultaneous setting of
3353 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3354 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3355 entry if you want to know the details. The fix, however, neglected to
3356 consider the case when local delivery batching is involved. The test for
3357 "previously delivered" was not happening when checking to see if an
3358 address could be batched with a previous (undelivered) one; under
3359 certain circumstances this could lead to multiple deliveries to the same
3362 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3363 in its include files, and this causes problems building Exim.
3365 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3366 header_syntax) but Exim was just ignoring anything given after a slash.
3367 In particular, this caused confusion with an attempt to use "verify =
3368 reverse_host_lookup/defer_ok". An error is now given when options are
3369 supplied for verify items that do not have them. (Maybe reverse_host_
3370 lookup should have a defer_ok option, but that's a different point.)
3372 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3373 defined by RFC 821) to 2048, because there were problems with some AUTH
3374 commands, and RFC 1869 says the size should be increased for extended
3375 SMTP commands that take arguments.
3377 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3380 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3381 "unknown" error; now it says that the functionality isn't in the binary.
3383 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3384 an address' error message when a string expansion fails (syntax or
3385 whatever). Otherwise the password may appear in the log. Following change
3386 PH/42 below, there is no longer a chance of it appearing in a bounce
3389 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3391 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3392 of its addresses was ever tried. (Bugzilla bug #2.)
3394 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3395 the result incorrectly in the debug output. (It correctly added a newline
3396 to what was transported.)
3398 TF/01 Added $received_time.
3400 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3401 commented out examples of how to interface to a virus scanner and to
3402 SpamAssassin. Also added commented examples of av_scanner and
3403 spamd_address settings.
3405 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3406 and controls are allowed in which ACLs. There were a couple of minor
3407 errors. Some of the entries in the conditions table (which is a table of
3408 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3409 negation of where the condition IS allowed.
3411 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3413 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3414 header file does not have a version number, so I've had to invent a new
3415 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3416 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3417 radiusclient), but it was contributed by a Radius user.
3419 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3420 files or directories, for OpenSSL.
3422 PH/19 When an Exim process that is running as root has to create an Exim log
3423 file, it does so in a subprocess that runs as exim:exim so as to get the
3424 ownership right at creation (otherwise, other Exim processes might see
3425 the file with the wrong ownership). There was no test for failure of this
3426 fork() call, which would lead to the process getting stuck as it waited
3427 for a non-existent subprocess. Forks do occasionally fail when resources
3428 run out. I reviewed all the other calls to fork(); they all seem to check
3431 PH/20 When checking for unexpected SMTP input at connect time (before writing
3432 the banner), Exim was not dealing correctly with a non-positive return
3433 from the read() function. If the client had disconnected by this time,
3434 the result was a log entry for a synchronization error with an empty
3435 string after "input=" when read() returned zero. If read() returned -1
3436 (an event I could not check), uninitialized data bytes were printed.
3437 There were reports of junk text (parts of files, etc) appearing after
3440 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3442 PH/22 Added support for macro redefinition, and (re)definition in between
3443 driver and ACL definitions.
3445 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3446 forgetting to use the resulting value; it was using the unexpanded value.
3448 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3449 hadn't been configured. The fix is from Juergen Kreileder, who
3450 understands it better than I do:
3452 "Here's what I see happening with three configured cyrus_sasl
3453 authenticators configured (plain, login, cram-md5):
3455 On startup auth_cyrus_sasl_init() gets called for each of these.
3456 This means three calls to sasl_listmech() without a specified mech_list.
3457 => SASL tests which mechs of all available mechs actually work
3458 => three warnings about OTP not working
3459 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3461 With the patch, sasl_listmech() also gets called three times. But now
3462 SASL's mech_list option is set to the server_mech specified in the the
3463 authenticator. Or in other words, the answer from sasl_listmech()
3464 gets limited to just the mech you're testing for (which is different
3466 => the return list contains just 'plain' or 'login', 'cram-md5' or
3467 nothing depending on the value of ob->server_mech.
3469 I've just tested the patch: Authentication still works fine,
3470 unavailable mechs specified in the exim configuration are still
3471 caught, and the auth.log warnings about OTP are gone."
3473 PH/25 When debugging is enabled, the contents of the command line are added
3474 to the debugging output, even when log_selector=+arguments is not
3477 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3478 answer is "GNU", and only if the return is "GNU/something" is the answer
3481 PH/27 $acl_verify_message is now set immediately after the failure of a
3482 verification in an ACL, and so is available in subsequent modifiers. In
3483 particular, the message can be preserved by coding like this:
3485 warn !verify = sender
3486 set acl_m0 = $acl_verify_message
3488 Previously, $acl_verify_message was set only while expanding "message"
3489 and "log_message" when a very denied access.
3491 PH/28 Modified OS/os.c-Linux with
3493 -#ifndef OS_LOAD_AVERAGE
3494 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3496 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3497 nomenclature these days.)
3499 PH/29 Installed patch from the Sieve maintainer that adds the options
3500 sieve_useraddress and sieve_subaddress to the redirect router.
3502 PH/30 In these circumstances:
3503 . Two addresses routed to the same list of hosts;
3504 . First host does not offer TLS;
3505 . First host accepts first address;
3506 . First host gives temporary error to second address;
3507 . Second host offers TLS and a TLS session is established;
3508 . Second host accepts second address.
3509 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3510 and peerdn, if requested) that were in fact used only for the second
3513 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3514 attention to any local part prefix or suffix that was matched by the
3515 router that accepted the address. It now behaves in the same way as it
3516 does for delivery: the affixes are removed from the local part unless
3517 rcpt_include_affixes is set on the transport.
3519 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3520 timeout during the DATA phase of an incoming message.
3522 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3523 applied a patch sanctioned by the Sieve maintainer.
3525 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3526 the uid or gid is negative. A case of a negative gid caused this to be
3527 noticed. The fix allows for either to be negative.
3529 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3530 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3533 PH/36 The change PH/12 above was broken. Fixed it.
3535 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3536 the grounds that routing the same address twice would always produce the
3537 same answer. This might have been true once, but it is certainly no
3538 longer true now. Routing a child address may depend on the previous
3539 routing that produced that child. Some complicated redirection strategies
3540 went wrong when messages had multiple recipients, and made Exim's
3541 behaviour dependent on the order in which the addresses were given.
3543 I have moved the duplicate checking until after the routing is complete.
3544 Exim scans the addresses that are assigned to local and remote
3545 transports, and removes any duplicates. This means that more work will be
3546 done, as duplicates will always all be routed, but duplicates are
3547 presumably rare, so I don't expect this is of any significance.
3549 For deliveries to pipes, files, and autoreplies, the duplicate checking
3550 still happens during the routing process, since they are not going to be
3553 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3554 It corrects a timeout issue with spamd. This is Ian's comment: "The
3555 background is that sometimes spamd either never reads data from a
3556 connection it has accepted, or it never writes response data. The exiscan
3557 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3558 blindly assumes that writes won't block so it may never time out."
3560 PH/39 Allow G after quota size as well as K and M.
3562 PH/40 The value set for $authenticated_id in an authenticator may not contain
3563 binary zeroes or newlines because the value is written to log lines and
3564 to spool files. There was no check on this. Now the value is run through
3565 the string_printing() function so that such characters are converted to
3566 printable escape sequences.
3568 PH/41 $message_linecount is a new variable that contains the total number of
3569 lines in the message. Compare $body_linecount, which is the count for the
3572 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3573 bounce and delay warning messages, except in certain special cases, which
3576 (a) An SMTP error message from a remote host;
3577 (b) A message specified in a :fail: redirection;
3578 (c) A message specified in a "fail" command in a system filter;
3579 (d) A message specified in a FAIL return from the queryprogram router;
3580 (e) A message specified by the cannot_route_message router option.
3582 In these cases only, Exim does include the error details in bounce and
3583 warning messages. There are also a few cases where bland messages such
3584 as "unrouteable address" or "local delivery error" are given.
3586 PH/43 $value is now also set for the "else" part of a ${run expansion.
3588 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3589 being worked on, but at least Exim now implements the latest version to
3592 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3593 process to complete was treated as a delivery failure, a timeout while
3594 writing the message to the pipe was logged, but erroneously treated as a
3595 successful delivery. Such timeouts include transport filter timeouts. For
3596 consistency with the overall process timeout, these timeouts are now
3597 treated as errors, giving rise to delivery failures by default. However,
3598 there is now a new Boolean option for the pipe transport called
3599 timeout_defer, which, if set TRUE, converts the failures into defers for
3600 both kinds of timeout. A transport filter timeout is now identified in
3603 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3604 systems where "make" and "gmake" are different, calling "gmake" at top
3605 level broke things. I've arranged for the value of $(MAKE) to be passed
3606 from the Makefile to this script so that it can call the same version of
3610 A note about Exim versions 4.44 and 4.50
3611 ----------------------------------------
3613 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3614 changes of various kinds. As a consequence, a big documentation update was
3615 needed. This delayed the release for rather longer than seemed good, especially
3616 in the light of a couple of (minor) security issues. Therefore, the changes
3617 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3618 release. So 4.44 and 4.50 are in effect two different branches that both start
3621 I have left the 4.50 change log unchanged; it contains all the changes since
3622 4.43. The change log for 4.44 is below; many of its items are identical to
3623 those for 4.50. This seems to be the most sensible way to preserve the
3624 historical information.
3630 1. Minor wording change to the doc/README.SIEVE file.
3632 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3633 computation of the current number of files was incorrect.
3635 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3636 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3639 4. Give more explanation in the error message when the command for a transport
3640 filter fails to execute.
3642 5. There are several places where Exim runs a non-Exim command in a
3643 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3644 was being done only for the command run by the queryprogram router. It is
3645 now done for all such subprocesses. The other cases are: ${run, transport
3646 filters, and the commands run by the lmtp and pipe transports.
3648 6. Added CONFIGURE_GROUP build-time option.
3650 7. Some older OS have a limit of 256 on the maximum number of file
3651 descriptors. Exim was using setrlimit() to set 1000 as a large value
3652 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3653 systems. I've change it so that if it can't get 1000, it tries for 256.
3655 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3656 was an oversight, and furthermore, ever since the addition of extra
3657 controls (e.g. 4.43/32), the checks on when to allow different forms of
3658 "control" were broken. There should now be diagnostics for all cases when a
3659 control that does not make sense is encountered.
3661 9. Added the /retain_sender option to "control=submission".
3663 10. $recipients is now available in the predata ACL (oversight).
3665 11. Tidy the search cache before the fork to do a delivery from a message
3666 received from the command line. Otherwise the child will trigger a lookup
3667 failure and thereby defer the delivery if it tries to use (for example) a
3668 cached ldap connection that the parent has called unbind on.
3670 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3671 of $address_data from the recipient verification was clobbered by the
3672 sender verification.
3674 13. The value of address_data from a sender verification is now available in
3675 $sender_address_data in subsequent conditions in the ACL statement.
3677 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3679 15. Added a new option "connect=<time>" to callout options, to set a different
3682 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3683 was its contents. (It was OK if the option was not defined at all.)
3685 17. A "Completed" log line is now written for messages that are removed from
3686 the spool by the -Mrm option.
3688 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3689 information about exactly what failed.
3691 19. Added -dd to debug only the daemon process.
3693 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3694 handles timeouts, both on the server side and network timeouts. Renamed the
3695 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3697 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3698 It is now set to "smtps".
3700 22. $host_address is now set to the target address during the checking of
3701 ignore_target_hosts.
3703 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3704 being passed; this would have caused $sender_host_name to have been used if
3705 matching the list had actually called for a host name (not very likely,
3706 since this list is usually IP addresses). A host name is now passed as
3709 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3710 code that specifies a non-restarting handler (typically sigaction() in
3711 modern systems) in an attempt to fix a rare and obscure crash bug.
3713 25. Narrowed the window for a race in the daemon that could cause it to ignore
3714 SIGCHLD signals. This is not a major problem, because they are used only to
3715 wake it up if nothing else does.
3717 26. A malformed maildirsize file could cause Exim to calculate negative values
3718 for the mailbox size or file count. Odd effects could occur as a result.
3719 The maildirsize information is now recalculated if the size or filecount
3722 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3723 support for a long time. Removed HAVE_SYS_VFS_H.
3725 28. Installed the latest version of exipick from John Jetmore.
3727 29. In an address list, if the pattern was not a regular expression, an empty
3728 subject address (from a bounce message) matched only if the pattern was an
3729 empty string. Non-empty patterns were not even tested. This was the wrong
3730 because it is perfectly reasonable to use an empty address as part of a
3731 database query. An empty address is now tested by patterns that are
3732 lookups. However, all the other forms of pattern expect the subject to
3733 contain a local part and a domain, and therefore, for them, an empty
3734 address still always fails if the pattern is not itself empty.
3736 30. Exim went into a mad DNS loop when attempting to do a callout where the
3737 host was specified on an smtp transport, and looking it up yielded more
3738 than one IP address.
3740 31. Re-factored the code for checking spool and log partition space into a
3741 function that finds that data and another that does the check. The former
3742 is then used to implement four new variables: $spool_space, $log_space,
3743 $spool_inodes, and $log_inodes.
3745 32. The RFC2047 encoding function was originally intended for short strings
3746 such as real names; it was not keeping to the 75-character limit for
3747 encoded words that the RFC imposes. It now respects the limit, and
3748 generates multiple encoded words if necessary. To be on the safe side, I
3749 have increased the buffer size for the ${rfc2047: expansion operator from
3752 33. It is now permitted to omit both strings after an "if" condition; if the
3753 condition is true, the result is "true". As before, when the second string
3754 is omitted, a false condition yields an empty string. This makes it less
3755 cumbersome to write custom ACL and router conditions.
3757 34. Failure to deliver a bounce message always caused it to be frozen, even if
3758 there was an errors_to setting on the router. The errors_to setting is now
3761 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3762 canonical form (fully expanded) before being placed in
3763 $sender_host_address.
3765 36. The table in the code that translates DNS record types into text (T_A to
3766 "A" for instance) was missing entries for NS and CNAME. It is just possible
3767 that this could have caused confusion if both these types were looked up
3768 for the same domain, because the text type is used as part of Exim's
3769 per-process caching. But the chance of anyone hitting this buglet seems
3772 37. The dnsdb lookup has been extended in a number of ways.
3774 (1) There is a new type, "zns", which walks up the domain tree until it
3775 finds some nameserver records. It should be used with care.
3777 (2) There is a new type, "mxh", which is like "mx" except that it returns
3778 just the host names, not the priorities.
3780 (3) It is now possible to give a list of domains (or IP addresses) to be
3781 looked up. The behaviour when one of the lookups defers can be
3782 controlled by a keyword.
3784 (4) It is now possible to specify the separator character for use when
3785 multiple records are returned.
3787 38. The dnslists ACL condition has been extended: it is now possible to supply
3788 a list of IP addresses and/or domains to be looked up in a particular DNS
3791 39. Added log_selector=+queue_time_overall.
3793 40. When running the queue in the test harness, wait just a tad after forking a
3794 delivery process, to get repeatability of debugging output.
3796 41. Include certificate and key file names in error message when GnuTLS fails
3797 to set them up, because the GnuTLS error message doesn't include the name
3798 of the failing file when there is a problem reading it.
3800 42. Allow both -bf and -bF in the same test run.
3802 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3804 44. The "Exiscan patch" is now merged into the mainline Exim source.
3806 45. Sometimes the final signoff response after QUIT could fail to get
3807 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3808 < 0 before doing a fflush(). This bug looks as though it goes back to the
3809 introduction of TLS in release 3.20, but "sometimes" must have been rare
3810 because the tests only now provoked it.
3812 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3813 (this can affect the format of dates).
3815 47. exim_tidydb, when checking for the continued existence of a message for
3816 which it has found a message-specific retry record, was not finding
3817 messages that were in split spool directories. Consequently, it was
3818 deleting retry records that should have stayed in existence.
3820 48. Steve fixed some bugs in eximstats.
3822 49. The SPA authentication driver was not abandoning authentication and moving
3823 on to the next authenticator when an expansion was forced to fail,
3824 contradicting the general specification for all authenticators. Instead it
3825 was generating a temporary error. It now behaves as specified.
3827 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3828 (the order specifies the preference for clients). The order is now AES256,
3829 AES128, 3DES, ARCFOUR128.
3831 51. Small patch to Sieve code - explicitly set From: when generating an
3834 52. Exim crashed if a remote delivery caused a very long error message to be
3835 recorded - for instance if somebody sent an entire SpamAssassin report back
3836 as a large number of 550 error lines. This bug was coincidentally fixed by
3837 increasing the size of one of Exim's internal buffers (big_buffer) that
3838 happened as part of the Exiscan merge. However, to be on the safe side, I
3839 have made the code more robust (and fixed the comments that describe what
3842 53. Now that there can be additional text after "Completed" in log lines (if
3843 the queue_time_overall log selector is set), a one-byte patch to exigrep
3844 was needed to allow it to recognize "Completed" as not the last thing in
3847 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3848 patch that reportedly fixes this has been added. I am not expert enough to
3849 create a test for it. This is what the patch creator wrote:
3851 "I found a little strange behaviour of ldap code when working with
3852 Windows 2003 AD Domain, where users was placed in more than one
3853 Organization Units. When I tried to give exim partial DN, the exit code
3854 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3855 But simultaneously result of request was absolutely normal ldap result,
3856 so I produce this patch..."
3858 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3859 REFERENCE, so I have modified the code to exclude the patch when that macro
3862 55. Some experimental protocols are using DNS PTR records for new purposes. The
3863 keys for these records are domain names, not reversed IP addresses. The
3864 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3865 leaves it alone. Component reversal etc. now happens only for IP addresses.
3868 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3870 57. Double the size of the debug message buffer (to 2048) so that more of very
3871 long debug lines gets shown.
3873 58. The exicyclog utility now does better if the number of log files to keep
3874 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3876 59. Two changes related to the smtp_active_hostname option:
3878 (1) $smtp_active_hostname is now available as a variable.
3879 (2) The default for smtp_banner uses $smtp_active_hostname instead
3880 of $primary_hostname.
3882 60. The host_aton() function is supposed to be passed a string that is known
3883 to be a valid IP address. However, in the case of IPv6 addresses, it was
3884 not checking this. This is a hostage to fortune. Exim now panics and dies
3885 if the condition is not met. A case was found where this could be provoked
3886 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3887 components; fortuitously, this particular loophole had already been fixed
3888 by change 4.50/55 above.
3890 If there are any other similar loopholes, the new check in host_aton()
3891 itself should stop them being exploited. The report I received stated that
3892 data on the command line could provoke the exploit when Exim was running as
3893 exim, but did not say which command line option was involved. All I could
3894 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3895 running as the user.
3898 61. There was a buffer overflow vulnerability in the SPA authentication code
3899 (which came originally from the Samba project). I have added a test to the
3900 spa_base64_to_bits() function which I hope fixes it.
3903 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3904 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3906 63. The daemon start-up calls getloadavg() while still root for those OS that
3907 need the first call to be done as root, but it missed one case: when
3908 deliver_queue_load_max is set with deliver_drop_privilege. This is
3909 necessary for the benefit of the queue runner, because there is no re-exec
3910 when deliver_drop_privilege is set.
3912 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3913 This has been fixed.
3915 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3916 was in use, was not putting the data itself into the right store pool;
3917 consequently, it could be overwritten for a subsequent message in the same
3918 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3921 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3923 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3924 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3925 the function were treating the return as a boolean value, which happened to
3926 work because 0=false and not-0=true, but is not correct code.
3928 68. The host_aton() function was not handling scoped IPv6 addresses (those
3929 with, for example, "%eth0" on the end) correctly.
3931 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3932 negated items (that is, ~something) in unsigned ints. Some compilers
3933 apparently mutter when there is no cast.
3935 70. If an address verification called from an ACL failed, and did not produce a
3936 user-specific message (i.e. there was only a "system" message), nothing was
3937 put in $acl_verify_message. In this situation, it now puts the system
3940 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3941 session; change 4.31/43 added the unwanted input to the log line - except
3942 that it did not do this in the start of session case. It now does.
3944 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3945 This is wrong and can cause the other end to generate a synchronization
3946 error if it is another Exim or anything else that does the synchronization
3947 check. A QUIT command is no longer sent after a timeout.
3949 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3950 during host lookups.
3952 74. The defer_ok option of callout verification was not working if it was used
3953 when verifying addresses in header lines, that is, for this case:
3955 verify = header_sender/callout=defer_ok
3957 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3958 those file descriptors could be used for SMTP connections. If anything
3959 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3960 could be sent to the SMTP client, causing chaos. The daemon now opens
3961 stdin, stdout, and stderr to /dev/null when it puts itself into the
3964 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3965 log by default. The user can override this with suitable Perl magic.
3967 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3968 the log message when discard triggers, was not working for the DATA ACL or
3969 for the non-SMTP ACL.
3971 78. Error message wording change in sieve.c.
3973 79. If smtp_accept_max_per_host was set, the number of connections could be
3974 restricted to fewer than expected, because the daemon was trying to set up
3975 a new connection before checking whether the processes handling previous
3976 connections had finished. The check for completed processes is now done
3977 earlier. On busy systems, this bug wouldn't be noticed because something
3978 else would have woken the daemon, and it would have reaped the completed
3981 80. If a message was submitted locally by a user whose login name contained one
3982 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3983 It caused a spool format error. I have fixed the spool reading code. A
3984 related problem was that the "from" clause in the Received: line became
3985 illegal because of the space(s). It is now covered by ${quote_local_part.
3987 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3990 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3992 83. Patch from Sieve maintainer to fix unterminated string problem in
3993 "vacation" handling.
3995 84. Some minor changes to the Linux configuration files to help with other
3996 OS variants using glibc.
3998 85. One more patch for Sieve to update vacation handling to latest spec.
4001 ----------------------------------------------------
4002 See the note above about the 4.44 and 4.50 releases.
4003 ----------------------------------------------------
4009 1. Change 4.43/35 introduced a bug that caused file counts to be
4010 incorrectly computed when quota_filecount was set in an appendfile
4013 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4014 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4017 3. Give more explanation in the error message when the command for a transport
4018 filter fails to execute.
4020 4. There are several places where Exim runs a non-Exim command in a
4021 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4022 was being done only for the command run by the queryprogram router. It is
4023 now done for all such subprocesses. The other cases are: ${run, transport
4024 filters, and the commands run by the lmtp and pipe transports.
4026 5. Some older OS have a limit of 256 on the maximum number of file
4027 descriptors. Exim was using setrlimit() to set 1000 as a large value
4028 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4029 systems. I've change it so that if it can't get 1000, it tries for 256.
4031 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4032 was an oversight, and furthermore, ever since the addition of extra
4033 controls (e.g. 4.43/32), the checks on when to allow different forms of
4034 "control" were broken. There should now be diagnostics for all cases when a
4035 control that does not make sense is encountered.
4037 7. $recipients is now available in the predata ACL (oversight).
4039 8. Tidy the search cache before the fork to do a delivery from a message
4040 received from the command line. Otherwise the child will trigger a lookup
4041 failure and thereby defer the delivery if it tries to use (for example) a
4042 cached ldap connection that the parent has called unbind on.
4044 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4045 of $address_data from the recipient verification was clobbered by the
4046 sender verification.
4048 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4049 was its contents. (It was OK if the option was not defined at all.)
4051 11. A "Completed" log line is now written for messages that are removed from
4052 the spool by the -Mrm option.
4054 12. $host_address is now set to the target address during the checking of
4055 ignore_target_hosts.
4057 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4058 being passed; this would have caused $sender_host_name to have been used if
4059 matching the list had actually called for a host name (not very likely,
4060 since this list is usually IP addresses). A host name is now passed as
4063 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4064 code that specifies a non-restarting handler (typically sigaction() in
4065 modern systems) in an attempt to fix a rare and obscure crash bug.
4067 15. Narrowed the window for a race in the daemon that could cause it to ignore
4068 SIGCHLD signals. This is not a major problem, because they are used only to
4069 wake it up if nothing else does.
4071 16. A malformed maildirsize file could cause Exim to calculate negative values
4072 for the mailbox size or file count. Odd effects could occur as a result.
4073 The maildirsize information is now recalculated if the size or filecount
4076 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4077 support for a long time. Removed HAVE_SYS_VFS_H.
4079 18. Updated exipick to current release from John Jetmore.
4081 19. Allow an empty sender to be matched against a lookup in an address list.
4082 Previously the only cases considered were a regular expression, or an
4085 20. Exim went into a mad DNS lookup loop when doing a callout where the
4086 host was specified on the transport, if the DNS lookup yielded more than
4089 21. The RFC2047 encoding function was originally intended for short strings
4090 such as real names; it was not keeping to the 75-character limit for
4091 encoded words that the RFC imposes. It now respects the limit, and
4092 generates multiple encoded words if necessary. To be on the safe side, I
4093 have increased the buffer size for the ${rfc2047: expansion operator from
4096 22. Failure to deliver a bounce message always caused it to be frozen, even if
4097 there was an errors_to setting on the router. The errors_to setting is now
4100 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4101 canonical form (fully expanded) before being placed in
4102 $sender_host_address.
4104 24. Updated eximstats to version 1.33
4106 25. Include certificate and key file names in error message when GnuTLS fails
4107 to set them up, because the GnuTLS error message doesn't include the name
4108 of the failing file when there is a problem reading it.
4110 26. Expand error message when OpenSSL has problems setting up cert/key files.
4113 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4114 (this can affect the format of dates).
4116 28. exim_tidydb, when checking for the continued existence of a message for
4117 which it has found a message-specific retry record, was not finding
4118 messages that were in split spool directories. Consequently, it was
4119 deleting retry records that should have stayed in existence.
4121 29. eximstats updated to version 1.35
4122 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4123 1.35 - bugfix such that pie charts by volume are generated correctly
4125 30. The SPA authentication driver was not abandoning authentication and moving
4126 on to the next authenticator when an expansion was forced to fail,
4127 contradicting the general specification for all authenticators. Instead it
4128 was generating a temporary error. It now behaves as specified.
4130 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4131 (the order specifies the preference for clients). The order is now AES256,
4132 AES128, 3DES, ARCFOUR128.
4134 31. Small patch to Sieve code - explicitly set From: when generating an
4137 32. Exim crashed if a remote delivery caused a very long error message to be
4138 recorded - for instance if somebody sent an entire SpamAssassin report back
4139 as a large number of 550 error lines. This bug was coincidentally fixed by
4140 increasing the size of one of Exim's internal buffers (big_buffer) that
4141 happened as part of the Exiscan merge. However, to be on the safe side, I
4142 have made the code more robust (and fixed the comments that describe what
4145 33. Some experimental protocols are using DNS PTR records for new purposes. The
4146 keys for these records are domain names, not reversed IP addresses. The
4147 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4148 leaves it alone. Component reversal etc. now happens only for IP addresses.
4151 34. The host_aton() function is supposed to be passed a string that is known
4152 to be a valid IP address. However, in the case of IPv6 addresses, it was
4153 not checking this. This is a hostage to fortune. Exim now panics and dies
4154 if the condition is not met. A case was found where this could be provoked
4155 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4156 components; fortuitously, this particular loophole had already been fixed
4157 by change 4.50/55 or 4.44/33 above.
4159 If there are any other similar loopholes, the new check in host_aton()
4160 itself should stop them being exploited. The report I received stated that
4161 data on the command line could provoke the exploit when Exim was running as
4162 exim, but did not say which command line option was involved. All I could
4163 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4164 running as the user.
4167 35. There was a buffer overflow vulnerability in the SPA authentication code
4168 (which came originally from the Samba project). I have added a test to the
4169 spa_base64_to_bits() function which I hope fixes it.
4172 36. The daemon start-up calls getloadavg() while still root for those OS that
4173 need the first call to be done as root, but it missed one case: when
4174 deliver_queue_load_max is set with deliver_drop_privilege. This is
4175 necessary for the benefit of the queue runner, because there is no re-exec
4176 when deliver_drop_privilege is set.
4178 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4179 was in use, was not putting the data itself into the right store pool;
4180 consequently, it could be overwritten for a subsequent message in the same
4181 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4184 38. Sometimes the final signoff response after QUIT could fail to get
4185 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4186 < 0 before doing a fflush(). This bug looks as though it goes back to the
4187 introduction of TLS in release 3.20, but "sometimes" must have been rare
4188 because the tests only now provoked it.
4194 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4195 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4196 Now it returns an integer. A number of calls were still expecting a T/F
4197 return. Fortuitously, in all cases, the tests worked in OK situations,
4198 which is the norm. However, things would have gone wrong on any write
4199 failures on the smtp file descriptor. This function is used when sending
4200 messages over SMTP and also when doing verify callouts.
4202 2. When Exim is called to do synchronous delivery of a locally submitted
4203 message (the -odf or -odi options), it no longer closes stderr before doing
4206 3. Implemented the mua_wrapper option.
4208 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4210 5. Implemented the functions header_remove(), header_testname(),
4211 header_add_at_position(), and receive_remove_recipient(), and exported them
4214 6. If an ACL "warn" statement specified the addition of headers, Exim already
4215 inserted X-ACL-Warn: at the start if there was no header name. However, it
4216 was not making this test for the second and subsequent header lines if
4217 there were newlines in the string. This meant that an invalid header could
4218 be inserted if Exim was badly configured.
4220 7. Allow an ACL "warn" statement to add header lines at the start or after all
4221 the Received: headers, as well as at the end.
4223 8. Added the rcpt_4xx retry error code.
4225 9. Added postmaster_mailfrom=xxx to callout verification option.
4227 10. Added mailfrom=xxxx to the callout verification option, for verify=
4230 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4231 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4233 12. Inserted some casts to stop certain compilers warning when using pointer
4234 differences as field lengths or precisions in printf-type calls (mostly
4235 affecting debugging statements).
4237 13. Added optional readline() support for -be (dynamically loaded).
4239 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4240 same clock tick as a message's arrival, so that its received time was the
4241 same as the "first fail" time on the retry record, and that message
4242 remained on the queue past the ultimate address timeout, every queue runner
4243 would try a delivery (because it was past the ultimate address timeout) but
4244 after another failure, the ultimate address timeout, which should have then
4245 bounced the address, did not kick in. This was a "< instead of <=" error;
4246 in most cases the first failure would have been in the next clock tick
4247 after the received time, and all would be well.
4249 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4250 being recognized when the domain list was tested by the match_domain
4251 condition in an expansion string.
4253 16. Added the ${str2b64: operator.
4255 17. Exim was always calling setrlimit() to set a large limit for the number of
4256 processes, without checking whether the existing limit was already
4257 adequate. (It did check for the limit on file descriptors.) Furthermore,
4258 errors from getrlimit() and setrlimit() were being ignored. Now they are
4259 logged to the main and panic logs, but Exim does carry on, to try to do its
4260 job under whatever limits there are.
4262 18. Imported PCRE 5.0.
4264 19. Trivial typo in log message " temporarily refused connection" (the leading
4267 20. If the log selector return_path_on_delivery was set and an address was
4268 redirected to /dev/null, the delivery process crashed because it assumed
4269 that a return path would always be set for a "successful" delivery. In this
4270 case, the whole delivery is bypassed as an optimization, and therefore no
4273 21. Internal re-arrangement: the function for sending a challenge and reading
4274 a response while authentication was assuming a zero-terminated challenge
4275 string. It's now changed to take a pointer and a length, to allow for
4276 binary data in such strings.
4278 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4280 23. Exim was not respecting finduser_retries when seeking the login of the
4281 uid under which it was called; it was always trying 10 times. (The default
4282 setting of finduser_retries is zero.) Also, it was sleeping after the final
4283 failure, which is pointless.
4285 24. Implemented tls_on_connect_ports.
4287 25. Implemented acl_smtp_predata.
4289 26. If the domain in control=submission is set empty, Exim assumes that the
4290 authenticated id is a complete email address when it generates From: or
4291 Sender: header lines.
4293 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4294 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4295 chown and chgrp in /bin and hostname in /usr/bin.
4297 28. Exim was keeping the "process log" file open after each use, just as it
4298 does for the main log. This opens the possibility of it remaining open for
4299 long periods when the USR1 signal hits a daemon. Occasional processlog
4300 errors were reported, that could have been caused by this. Anyway, it seems
4301 much more sensible not to leave this file open at all, so that is what now
4304 29. The long-running daemon process does not normally write to the log once it
4305 has entered its main loop, and it closes the log before doing so. This is
4306 so that log files can straightforwardly be renamed and moved. However,
4307 there are a couple of unusual error situations where the daemon does write
4308 log entries, and I had neglected to close the log afterwards.
4310 30. The text of an SMTP error response that was received during a remote
4311 delivery was being truncated at 512 bytes. This is too short for some of
4312 the long messages that one sometimes sees. I've increased the limit to
4315 31. It is now possible to make retry rules that apply only when a message has a
4316 specific sender, in particular, an empty sender.
4318 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4319 it possible to be selective about when SMTP synchronization is enforced.
4321 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4323 32. Implemented hosts_connection_nolog.
4325 33. Added an ACL for QUIT.
4327 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4330 35. Added mailbox_size and mailbox_filecount to appendfile.
4332 36. Added control = no_multiline_responses to ACLs.
4334 37. There was a bug in the logic of the code that waits for the clock to tick
4335 in the case where the clock went backwards by a substantial amount such
4336 that the microsecond fraction of "now" was more than the microsecond
4337 fraction of "then" (but the whole seconds number was less).
4339 38. Added support for the libradius Radius client library this is found on
4340 FreeBSD (previously only the radiusclient library was supported).
4346 1. When certain lookups returned multiple values in the form name=value, the
4347 quoting of the values was not always being done properly. Specifically:
4348 (a) If the value started with a double quote, but contained no whitespace,
4350 (b) If the value contained whitespace other than a space character (i.e.
4351 tabs or newlines or carriage returns) it was not quoted.
4352 This fix has been applied to the mysql and pgsql lookups by writing a
4353 separate quoting function and calling it from the lookup code. The fix
4354 should probably also be applied to nisplus, ibase and oracle lookups, but
4355 since I cannot test any of those, I have not disturbed their existing code.
4357 2. A hit in the callout cache for a specific address caused a log line with no
4358 reason for rejecting RCPT. Now it says "Previous (cached) callout
4359 verification failure".
4361 3. There was an off-by-one bug in the queryprogram router. An over-long
4362 return line was truncated at 256 instead of 255 characters, thereby
4363 overflowing its buffer with the terminating zero. As well as fixing this, I
4364 have increased the buffer size to 1024 (and made a note to document this).
4366 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4367 when Exim is waiting for an SMTP response from a remote server, Exim
4368 restarts its select() call on the socket, thereby resetting its timeout.
4369 This is not a problem when such interrupts are rare. Somebody set up a cron
4370 job to run exiwhat every 2 minutes, which is less than the normal select()
4371 timeout (5 or 10 minutes). This meant that the select() timeout never
4372 kicked in because it was always reset. I have fixed this by comparing the
4373 time when an interrupt arrives with the time at the start of the first call
4374 to select(). If more time than the timeout has elapsed, the interrupt is
4375 treated as a timeout.
4377 5. Some internal re-factoring in preparation for the addition of Sieve
4378 extensions (by MH). In particular, the "personal" test is moved to a
4379 separate function, and given an option for scanning Cc: and Bcc: (which is
4380 not set for Exim filters).
4382 6. When Exim created an email address using the login of the caller as the
4383 local part (e.g. when creating a From: or Sender: header line), it was not
4384 quoting the local part when it contained special characters such as @.
4386 7. Installed new OpenBSD configuration files.
4388 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4389 try to make them clearer.
4391 9. Callout options, other than the timeout value, were being ignored when
4392 verifying sender addresses in header lines. For example, when using
4394 verify = header_sender/callout=no_cache
4396 the cache was (incorrectly) being used.
4398 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4399 only the Cygwin environment.
4401 11. When return_path_on_delivery was set as a log selector, if different remote
4402 addresses in the same message used different return paths and parallel
4403 remote delivery occurred, the wrong values would sometimes be logged.
4404 (Whenever a remote delivery process finished, the return path value from
4405 the most recently started remote delivery process was logged.)
4407 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4408 lines when AUTH and/or TLS are in use. This is the "received protocol"
4409 field. Exim used to use "asmtp" for authenticated SMTP, without any
4410 indication (in the protocol name) for TLS use. Now it follows the RFC and
4411 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4412 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4413 names appear in log lines as well as in Received: header lines.
4415 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4416 extensions, and comparison tests, and to fix some bugs.
4418 14. Changes to the "personal" filter test:
4420 (1) The test was buggy in that it was just doing the equivalent of
4421 "contains" tests on header lines. For example, if a user's address was
4422 anne@some.where, the "personal" test would incorrectly be true for
4424 To: susanne@some.where
4426 This test is now done by extracting each address from the header in turn,
4427 and checking the entire address. Other tests that are part of "personal"
4428 are now done using regular expressions (for example, to check local parts
4429 of addresses in From: header lines).
4431 (2) The list of non-personal local parts in From: addresses has been
4432 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4433 taken from the Sieve specification recommendations.
4435 (3) If the message contains any header line starting with "List-" it is
4436 treated as non-personal.
4438 (4) The test for "circular" in the Subject: header line has been removed
4439 because it now seems ill-conceived.
4441 15. Minor typos in src/EDITME comments corrected.
4443 16. Installed latest exipick from John Jetmore.
4445 17. If headers_add on a router specified a text string that was too long for
4446 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4447 of string_sprintf() is now avoided.
4449 18. $message_body_size was not set (it was always zero) when running the DATA
4450 ACL and the local_scan() function.
4452 19. For the "mail" command in an Exim filter, no default was being set for
4453 the once_repeat time, causing a random time value to be used if "once" was
4454 specified. (If the value happened to be <= 0, no repeat happened.) The
4455 default is now 0s, meaning "never repeat". The "vacation" command was OK
4456 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4457 (I found it when inspecting the code).
4459 20. There is now an overall timeout for performing a callout verification. It
4460 defaults to 4 times the callout timeout, which applies to individual SMTP
4461 commands during the callout. The overall timeout applies when there is more
4462 than one host that can be tried. The timeout is checked before trying the
4463 next host. This prevents very long delays if there are a large number of
4464 hosts and all are timing out (e.g. when the network connections are timing
4465 out). The value of the overall timeout can be changed by specifying an
4466 additional sub-option for "callout", called "maxwait". For example:
4468 verify = sender/callout=5s,maxwait=20s
4470 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4471 to the end before writing, but this should make it even safer).
4473 22. Exim was forgetting that it had advertised PIPELINING for the second and
4474 subsequent messages on an SMTP connection. It was also not resetting its
4475 memory on STARTTLS and an internal HELO.
4477 23. When Exim logs an SMTP synchronization error within a session, it now
4478 records whether PIPELINING has been advertised or not.
4480 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4481 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4482 rather than long int.
4484 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4486 26. Added the never_mail option to autoreply.
4492 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4493 crash if the getsockname() call failed; this can happen if a connection is
4494 closed very soon after it is established. The problem was simply in the
4495 order in which certain operations were done, causing Exim to try to write
4496 to the SMTP stream before it had set up the file descriptor. The bug has
4497 been fixed by making things happen in the correct order.
4503 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4504 before the connection was closed, thus losing the rejection response.
4506 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4507 some early Solaris releases, but causes trouble in current releases where
4508 socklen_t is defined.
4510 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4513 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4516 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4517 that releasing the top of it at the end releases what was used for sub-
4518 expansions (unless the block got too big). However, discard this block if
4519 the first thing is a variable or header, so that we can use its block when
4520 it is dynamic (useful for very large $message_headers, for example).
4522 6. Lookups now cache *every* query, not just the most recent. A new, separate
4523 store pool is used for this. It can be recovered when all lookup caches are
4524 flushed. Lookups now release memory at the end of their result strings.
4525 This has involved some general refactoring of the lookup sources.
4527 7. Some code has been added to the store_xxx() functions to reduce the amount
4528 of flapping under certain conditions.
4530 8. log_incoming_interface used to affect only the <= reception log lines. Now
4531 it causes the local interface and port to be added to several more SMTP log
4532 lines, for example "SMTP connection from", and rejection lines.
4534 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4536 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4538 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4539 could be overwritten at the end of the current message (or the start of a
4540 new message if it was set in a HELO ACL). The value is now preserved for
4541 the duration of the SMTP connection.
4543 12. If a transport had a headers_rewrite setting, and a matching header line
4544 contained an unqualified address, that address was qualified, even if it
4545 did not match any rewriting rules. The underlying bug was that the values
4546 of the flags that permit the existence of unqualified sender and recipient
4547 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4548 non-local messages, and by -bnq for local messages) were not being
4549 preserved with the message after it was received.
4551 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4552 "next input=" as part of the text comprising the host identity instead of
4553 the correct text. The code was using the same buffer for two different
4554 strings. However, depending on which order the printing function evaluated
4555 its arguments, the bug did not always show up. Under Linux, for example, my
4556 test suite worked just fine.
4558 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4559 doesn't work with some older versions of Perl. It has been changed to "my",
4560 which in any case is probably the better facility to use.
4562 15. A really picky compiler found some instances of statements for creating
4563 error messages that either had too many or two few arguments for the format
4566 16. The size of the buffer for calls to the DNS resolver has been increased
4567 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4568 for addresses that have a lot of PTR records. This alleviates a problem; it
4569 does not fully solve it.
4571 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4572 buffer now truncates the list and logs the incident, which is the same
4573 action as happens when Exim is looking up a host name and its aliases.
4574 Previously in this situation something unpredictable would happen;
4575 sometimes it was "internal error: store_reset failed".
4577 18. If a server dropped the connection unexpectedly when an Exim client was
4578 using GnuTLS and trying to read a response, the client delivery process
4579 crashed while trying to generate an error log message.
4581 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4582 string, for example:
4584 warn message = H1: something\nH2: something
4586 the text was added as a single header line from Exim's point of view
4587 though it ended up OK in the delivered message. However, searching for the
4588 second and subsequent header lines using $h_h2: did not work. This has been
4589 fixed. Similarly, if a system filter added multiple headers in this way,
4590 the routers could not see them.
4592 20. Expanded the error message when iplsearch is called with an invalid key to
4593 suggest using net-iplsearch in a host list.
4595 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4596 ACLs are no longer actually imposed (and a message to that effect is
4599 22. If a "gecos" field in a passwd entry contained escaped characters, in
4600 particular, if it contained a \" sequence, Exim got it wrong when building
4601 a From: or a Sender: header from that name. A second bug also caused
4602 incorrect handling when an unquoted " was present following a character
4603 that needed quoting.
4605 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4606 was not being matched caselessly.
4608 24. Arranged for all hyphens in the exim.8 source to be escaped with
4611 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4612 properly. Recipient callout cache records were still being keyed to include
4613 the sender, even when use_sender was set false. This led to far more
4614 callouts that were necessary. The sender is no longer included in the key
4615 when use_sender is false.
4617 26. Added "control = submission" modifier to ACLs.
4619 27. Added the ${base62d: operator to decode base 62 numbers.
4621 28. dnsdb lookups can now access SRV records.
4623 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4624 the configuration file.
4626 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4627 (-v) mode. This makes the output for a verbose queue run more intelligible.
4629 31. Added a use_postmaster feature to recipient callouts.
4631 32. Added the $body_zerocount variable, containing the number of binary zero
4632 bytes in the message body.
4634 33. The time of last modification of the "new" subdirectory is now used as the
4635 "mailbox time last read" when there is a quota error for a maildir
4638 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4640 35. Added +ignore_unknown as a special item in host lists.
4642 36. Code for decoding IPv6 addresses in host lists is now included, even if
4643 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4644 address was recognized as an IP address, but was then not correctly decoded
4645 into binary, causing unexpected and incorrect effects when compared with
4652 1. Very minor rewording of debugging text in manualroute to say "list of
4653 hosts" instead of "hostlist".
4655 2. If verify=header_syntax was set, and a header line with an unqualified
4656 address (no domain) and a large number of spaces between the end of the
4657 name and the colon was received, the reception process suffered a buffer
4658 overflow, and (when I tested it) crashed. This was caused by some obsolete
4659 code that should have been removed. The fix is to remove it!
4661 3. When running in the test harness, delay a bit after writing a bounce
4662 message to get a bit more predictability in the log output.
4664 4. Added a call to search_tidyup() just before forking a reception process. In
4665 theory, someone could use a lookup in the expansion of smtp_accept_max_
4666 per_host which, without the tidyup, could leave open a database connection.
4668 5. Added the variables $recipient_data and $sender_data which get set from a
4669 lookup success in an ACL "recipients" or "senders" condition, or a router
4670 "senders" option, similar to $domain_data and $local_part_data.
4672 6. Moved the writing of debug_print from before to after the "senders" test
4675 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4676 problems for message scanning, either using a data ACL, or using
4677 local_scan() because the Received: header was not generated till after they
4678 were called (in order to set the time as the time of reception completion).
4679 I have revised the way this works. The header is now generated after the
4680 body is received, but before the ACL or local_scan() are called. After they
4681 are run, the timestamp in the header is updated.
4687 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4688 before starting a queue runner without re-exec. This happened only when
4689 deliver_drop_privilege was set or when the Exim user was set to root. The
4690 effect of the bug was that timeouts during subsequent deliveries caused
4691 crashes instead of being properly handled. The handler is now left at its
4692 default (and expected) setting.
4694 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4695 message, again when deliver_drop_privilege is set or Exim is run as root.
4696 The bug described in (1) was not present in this case, but the tidying up
4697 of the other signals was missing. I have made the two cases consistent.
4699 3. The ignore_target_hosts setting on a manualroute router was being ignored
4700 for hosts that were looked up using the /MX notation.
4702 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4705 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4706 operated on the sender address. After changing the $sender_address to <>
4707 for the sender address verify, Exim was re-instated it as the original
4708 (before rewriting) address, but remembering that it had rewritten it, so it
4709 wasn't rewriting it again. This bug also had the effect of breaking the
4710 sender address verification caching when the sender address was rewritten.
4712 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4713 This has been changed so that if the ip literal address matches
4714 ignore_target_hosts, the router declines.
4716 7. Added expansion conditions match_domain, match_address, and match_local_
4717 part (NOT match_host).
4719 8. The placeholder for the Received: header didn't have a length field set.
4721 9. Added code to Exim itself and to exim_lock to test for a specific race
4722 condition that could lead to file corruption when using MBX delivery. The
4723 issue is with the lockfile that is created in /tmp. If this file is removed
4724 after a process has opened it but before that process has acquired a lock,
4725 there is the potential for a second process to recreate the file and also
4726 acquire a lock. This could lead to two Exim processes writing to the file
4727 at the same time. The added code performs the same test as UW imapd; it
4728 checks after acquiring the lock that its file descriptor still refers to
4729 the same named file.
4731 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4732 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4735 11. Added the smtp_active_hostname option. If used, this will typically be made
4736 to depend on the incoming interface address. Because $interface_address is
4737 not set up until the daemon has forked a reception process, error responses
4738 that can happen earlier (such as "too many connections") no longer contain
4741 12. If an expansion in a condition on a "warn" statement fails because a lookup
4742 defers, the "warn" statement is abandoned, and the next ACL statement is
4743 processed. Previously this caused the whole ACL to be aborted.
4745 13. Added the iplsearch lookup type.
4747 14. Added ident_timeout as a log selector.
4749 15. Added tls_certificate_verified as a log selector.
4751 16. Added a global option tls_require_ciphers (compare the smtp transport
4752 option of the same name). This controls incoming TLS connections.
4754 17. I finally figured out how to make tls_require_ciphers do a similar thing
4755 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4756 before starting the TLS session.
4758 18. Tabs are now shown as \t in -bP output.
4760 19. If the log selector return_path_on_delivery was set, Exim crashed when
4761 bouncing a message because it had too many Received: header lines.
4763 20. If two routers both had headers_remove settings, and the first one included
4764 a superfluous trailing colon, the final name in the first list and the
4765 first name in the second list were incorrectly joined into one item (with a
4766 colon in the middle).
4772 1. Added -C and -D options to the exinext utility, mainly to make it easier
4773 to include in the automated testing, but these could be helpful when
4774 multiple configurations are in use.
4776 2. The exinext utility was not formatting the output nicely when there was
4777 an alternate port involved in the retry record key, nor when there was a
4778 message id as well (for retries that were specific to a specific message
4779 and a specific host). It was also confused by IPv6 addresses, because of
4780 the additional colons they contain. I have fixed the IPv4 problem, and
4781 patched it up to do a reasonable job for IPv6.
4783 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4784 delivery, the log line now contains "pipelined" if PIPELINING was used.
4786 4. An SMTP transport process used to panic and die if the bind() call to set
4787 an explicit outgoing interface failed. This has been changed; it is now
4788 treated in the same way as a connect() failure.
4790 5. A reference to $sender_host_name in the part of a conditional expansion
4791 that was being skipped was still causing a DNS lookup. This no longer
4794 6. The def: expansion condition was not recognizing references to header lines
4795 that used bh_ and bheader_.
4797 7. Added the _cache feature to named lists.
4799 8. The code for checking quota_filecount in the appendfile transport was
4800 allowing one more file than it should have been.
4802 9. For compatibility with Sendmail, the command line option
4810 and sets the incoming protocol and host name (for trusted callers). The
4811 host name and its colon can be omitted when only the protocol is to be set.
4812 Note the Exim already has two private options, -pd and -ps, that refer to
4813 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4814 "s", but I don't think that's a major issue.
4816 10. A number of refactoring changes to the code, none of which should affect
4819 (a) The number of logging options was getting close to filling up the
4820 32-bit word that was used as a bit map. I have split them into two classes:
4821 those that are passed in the argument to log_write(), and those that are
4822 only ever tested independently outside of that function. These are now in
4823 separate 32-bit words, so there is plenty of room for expansion again.
4824 There is no change in the user interface or the logging behaviour.
4826 (b) When building, for example, log lines, the code previously used a
4827 macro that called string_cat() twice, in order to add two strings. This is
4828 not really sufficiently general. Furthermore, there was one instance where
4829 it was actually wrong because one of the argument was used twice, and in
4830 one call a function was used. (As it happened, calling the function twice
4831 did not affect the overall behaviour.) The macro has been replaced by a
4832 function that can join an arbitrary number of extra strings onto a growing
4835 (c) The code for expansion conditions now uses a table and a binary chop
4836 instead of a serial search (which was left over from when there were very
4837 few conditions). Also, it now recognizes conditions like "pam" even when
4838 the relevant support is not compiled in: a suitably worded error message is
4839 given if an attempt is made to use such a condition.
4841 11. Added ${time_interval:xxxxx}.
4843 12. A bug was causing one of the ddress fields not to be passed back correctly
4844 from remote delivery subprocesses. The field in question was not being
4845 subsequently used, so this caused to problems in practice.
4847 13. Added new log selectors queue_time and deliver_time.
4849 14. Might have fixed a bug in maildirsizefile handling that threw up
4850 "unexpected character" debug warnings, and recalculated the data
4851 unnecessarily. In any case, I expanded the warning message to give more
4854 15. Added the message "Restricted characters in address" to the statements in
4855 the default ACL that block characters like @ and % in local parts.
4857 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4858 Three changes have been made:
4860 (a) There was a serious bug; a negative response to MAIL caused the whole
4861 recipient domain to be cached as invalid, thereby blocking all messages
4862 to all local parts at the same domain, from all senders. This bug has
4863 been fixed. The domain is no longer cached after a negative response to
4864 MAIL if the sender used is not empty.
4866 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4869 (c) A new callout option, "use_sender" has been added for people who want
4870 the modified behaviour.
4876 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4879 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4880 indeed breaks things for older releases.
4882 3. Added additional logging to the case where there is a problem reading data
4883 from a filter that is running in a subprocess using a pipe, in order to
4884 try to track down a specific problem.
4886 4. Testing facility fudge: when running in the test harness and attempting
4887 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4888 getting "No route to host". Convert this to a timeout.
4890 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4893 6. Some OS don't have socklen_t but use size_t instead. This affects the
4894 fifth argument of getsockopt() amongst other things. This is now
4895 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4896 can be set for individual OS. I have set it for SunOS5, OSF1, and
4897 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4898 some earlier ones do not.
4900 7. Change 4.30/15 was not doing the test caselessly.
4902 8. The standard form for an IPv6 address literal was being rejected by address
4903 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4904 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4905 this, as well as the form without the "IPv6" on the front (but only when
4906 address literals are enabled, of course).
4908 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4910 10. Exim crashed if a message with an empty sender address specified by -f
4911 encountered a router with an errors_to setting. This could be provoked only
4912 by a command such as
4916 where an empty string was supplied; "<>" did not hit this bug.
4918 11. Installed PCRE release 4.5.
4920 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4921 remained set. It is now erased.
4923 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4924 times from message ids (which are base 36 rather than the normal 62).
4926 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4927 were being counted as actual protocol errors, and logged if the log
4928 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4929 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4930 and DATA following a set of rejected RCPTs do not count as protocol errors.
4931 In other words, Exim assumes they were pipelined, though this may not
4932 actually be the case. Of course, in all cases the client gets an
4933 appropriate error code.
4935 15. If a lookup fails in an ACL condition, a message about the failure may
4936 be available; it is used if testing the ACL cannot continue, because most
4937 such messages specify what the cause of the deferral is. However, some
4938 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4939 that caused an old message to be retained and used if a later statement
4940 caused a defer, replacing the real cause of the deferral.
4942 16. If an IP address had so many PTR records that the DNS lookup buffer
4943 was not large enough to hold them, Exim could crash while trying to process
4944 the truncated data. It now detects and logs this case.
4946 17. Further to 4.21/58, another change has been made: if (and only if) the
4947 first line of a message (the first header line) ends with CRLF, a bare LF
4948 in a subsequent header line has a space inserted after it, so as not to
4949 terminate the header.
4951 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4952 unnecessarily, used atoi() instead of strtol(), and didn't check the
4953 termination when getting file sizes from file names by regex.
4955 19. Completely re-implemented the support for maildirsize files, in the light
4956 of a number of problems with the previous contributed implementation
4957 (4.30/29). In particular:
4959 . If the quota is zero, the maildirsize file is maintained, but no quota is
4962 . If the maildir directory does not exist, it is created before any attempt
4963 to write a maildirsize file.
4965 . The quota value in the file is just a cache; if the quota is changed in
4966 the transport, the new value overrides.
4968 . A regular expression is available for excluding directories from the
4971 20. The autoreply transport checks the characters in options that define the
4972 message's headers; it allows continued headers, but it was checking with
4973 isspace() after an embedded newline instead of explicitly looking for a
4976 21. If all the "regular" hosts to which an address was routed had passed their
4977 expiry times, and had not reached their retry times, the address was
4978 bounced, even if fallback hosts were defined. Now Exim should go on to try
4981 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4982 equivalent code in the SMTP transport. Some hosts send humungous responses
4983 to HELO/EHLO, more than 1024 it seems.
4985 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4986 gives compiler warnings in some environments. I've now done it "properly",
4989 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4990 (because of problems with the built-in one) was declared to return uschar *
4991 instead of char *, causing compiler failure.
4993 25. Fixed a file descriptor leak when processing alias/forward files.
4995 26. Fixed a minor format string issue in dbfn.c.
4997 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4999 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5000 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5003 29. When a sender address is verified, it is cached, to save repeating the test
5004 when there is more than one recipient in a message. However, when the
5005 verification involves a callout, it is possible for different callout
5006 options to be set for different recipients. It is too complicated to keep
5007 track of this in the cache, so now Exim always runs a verification when a
5008 callout is required, relying on the callout cache for the optimization.
5009 The overhead is duplication of the address routing, but this should not be
5012 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5013 to be verified with callout=postmaster, and the main callout worked but the
5014 postmaster check failed, the verification correctly failed. However, if a
5015 subsequent RCPT command asked for sender verification *without* the
5016 postmaster check, incorrect caching caused this verification also to fail,
5019 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5020 it was not caching the DNS options (qualify_single, search_parents) that
5021 were used when the lookup failed. A subsequent lookup with different
5022 options therefore always gave the same answer, though there were cases
5023 where it should not have. (Example: a "domains = !$mx_any" option on a
5024 dnslookup router: the "domains" option is always processed without any
5025 widening, but the router might have qualify_single set.) Now Exim uses the
5026 cached value only when the same options are set.
5028 32. Added John Jetmore's "exipick" utility to the distribution.
5030 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5031 than a timeout (e.g. a certificate is required, and is not provided), an
5032 Exim server now closes the connection immediately. Previously it waited for
5033 the client to close - but if the client is SSL, it seems that they each
5034 wait for each other, leading to a delay before one of them times out.
5036 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5037 maintained 0.8.x compatibility because I don't think many are using it, and
5038 it is clearly obsolete.
5040 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5043 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5044 client certificate was expired. A simple patch fixes this, though I don't
5045 understand the full logic of why the verify callback is called multiple
5048 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5049 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5050 which causes problems with some clients (such as the Certicom SSL Plus
5051 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5052 disables the coutermeasure allowing Eudora to connect."
5054 38. Exim was not checking that a write() to a log file succeeded. This could
5055 lead to Bad Things if a log got too big, in particular if it hit a file
5056 size limit. Exim now panics and dies if it cannot write to a log file, just
5057 as it does if it cannot open a log file.
5059 39. Modified OS/Makefile-Linux so that it now contains
5061 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5063 The two -D definitions ensure that Exim is compiled with large file
5064 support, which makes it possible to handle log files that are bigger than
5067 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5068 instance) a domain was checked against a named list that involved a lookup,
5069 causing $domain_data to be set, then another domain was checked against the
5070 same list, then the first domain was re-checked, the value of $domain_data
5071 after the final check could be wrong. In particular, if the second check
5072 failed, it could be set empty. This bug probably also applied to
5075 41. The strip_trailing_dot option was not being applied to the address given
5076 with the -f command-line option.
5078 42. The code for reading a message's header from the spool was incrementing
5079 $received_count, but never initializing it. This meant that the value was
5080 incorrect (doubled) while delivering a message in the same process in which
5081 it was received. In the most common configuration of Exim, this never
5082 happens - a fresh exec is done - but it can happen when
5083 deliver_drop_privilege is set.
5085 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5086 it now includes up to 150 characters of the unexpected data in the log
5089 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5090 and building data strings. The size of both of these buffers was 10 000
5091 bytes - far larger than anybody would *ever* want, thought I. Needless to
5092 say, somebody hit the limit. I have increased the maximum line length to
5093 20 000 and the maximum data length of concatenated lines to 100 000. I have
5094 also fixed two bugs, because there was no checking on these buffers. Tsk,
5095 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5096 buffer is too small.
5098 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5099 lsearch lookups. Now it does.
5101 46. When parsing a route_list item in a manualroute router, a fixed-length
5102 buffer was used for the list of hosts. I made this 1024 bytes long,
5103 thinking that nobody would ever have a list of hosts that long. Wrong.
5104 Somebody had a whole pile of complicated expansion conditions, and the
5105 string was silently truncated, leading to an expansion error. It turns out
5106 that it is easier to change to an unlimited length (owing to other changes
5107 that have happened since this code was originally written) than to build
5108 structure for giving a limitation error. The length of the item that
5109 expands into the list of hosts is now unlimited.
5111 47. The lsearch lookup could not handle data where the length of text line was
5112 more than 4095 characters. Such lines were truncated, leading to shortened
5113 data being returned. It should now handle lines of any length.
5115 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5116 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5119 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5120 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5121 no longer happens in comment lines. A list of replacements is now placed
5122 at the head of all of the source files, except those whose only change is
5123 to replace PERL_COMMAND in the very first #! line.
5125 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5126 messages on the queue, with a bottom-up merge sort, using code contributed
5127 by Michael Haardt. This should make operations like -bp somewhat faster on
5128 large queues. It won't affect queue runners, except when queue_run_in_order
5131 51. Installed eximstats 1.31 in the distribution.
5133 52. Added support for SRV lookups to the dnslookup router.
5135 53. If an ACL referred to $message_body or $message_body_end, the value was not
5136 reset for any messages that followed in the same SMTP session.
5138 54. The store-handling optimization for building very long strings was not
5139 differentiating between the different store pools. I don't think this
5140 actually made any difference in practice, but I've tidied it.
5142 55. While running the routers to verify a sender address, $sender_address
5143 was still set to the sender address. This is wrong, because when routing to
5144 send a bounce to the sender, it would be empty. Therefore, I have changed
5145 it so that, while verifying a sender address, $sender_address is set to <>.
5146 (There is no change to what happens when verifying a recipient address.)
5148 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5149 target A or AAAA records (if not already returned) without resetting the
5150 qualify_single or search_parents options of the DNS resolver. These are
5151 inappropriate in this case because the targets of MX and SRV records must
5152 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5153 target that, when qualified, matched something in the local domain. These
5154 two options are now turned off when doing these lookups.
5156 57. It seems that at least some releases of Reiserfs (which does not have the
5157 concept of a fixed number of inodes) returns zero and not -1 for the
5158 number of available inodes. This interacted badly with check_spool_inodes,
5159 which assumed that -1 was the "no such thing" setting. What I have done is
5160 to check that the total number of inodes is greater than zero before doing
5161 the test of how many are available.
5163 58. When a "warn" ACL statement has a log_message modifier, the message is
5164 remembered, and not repeated. This is to avoid a lot of repetition when a
5165 message has many recipients that cause the same warning to be written.
5166 Howewer, Exim was preserving the list of already written lines for an
5167 entire SMTP session, which doesn't seem right. The memory is now reset if a
5168 new message is started.
5170 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5171 debugging output unless log_rewrite was also set.
5173 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5174 of (int)(handle) when we know that handle contains (void *)(-1).
5176 61. The Exim daemon panic-logs an error return when it closes the incoming
5177 connection. However "connection reset by peer" seems to be common, and
5178 isn't really an error worthy of noting specially, so that particular error
5181 62. When Exim is trying to find all the local interfaces, it used to panic and
5182 die if the ioctl to get the interface flags failed. However, it seems that
5183 on at least one OS (Solaris 9) it is possible to have an interface that is
5184 included in the list of interfaces, but for which you get a failure error
5185 for this call. This happens when the interface is not "plumbed" into a
5186 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5187 failure of the "get flags" call assumes that the interface is down.
5189 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5190 makes life easier for people who are doing arithmetic on fields extracted
5191 from dates, where you often get leading zeros that should not be
5192 interpreted as octal.
5194 64. Added qualify_domain to the redirect router, to override the global
5197 65. If a pathologically long header line contained very many addresses (the
5198 report of this problem mentioned 10 000) and each of them was rewritten,
5199 Exim could use up a very large amount of memory. (It kept on making new
5200 copies of the header line as it rewrote, and never released the old ones.)
5201 At the expense of a bit more processing, the header rewriting function has
5202 been changed so that it no longer eats memory in this way.
5204 66. The generation of the Received: header has been moved from the time that a
5205 message starts to be received, to the time that it finishes. The timestamp
5206 in the Received: header should now be very close to that of the <= log
5207 line. There are two side-effects of this change:
5209 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5210 logged header lines no longer include the local Received: line, because
5211 it has not yet been created. The same applies to a copy of the message
5212 that is returned to a non-SMTP sender when a message is rejected.
5214 (b) When a filter file is tested using -bf, no additional Received: header
5215 is added to the test message. After some thought, I decided that this
5218 This change does not affect the value of $received_for. It is still set
5219 after address rewriting, but before local_scan() is called.
5221 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5223 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5224 gave an unhelpful panic error message, and a defer error. I have managed to
5225 change this behaviour so that it now rejects any supplied certificate,
5226 which seems right, as the list of acceptable certificates is empty.
5228 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5229 gave an unhelpful defer error. I have not managed to make this reject any
5230 supplied certificates, but the error message it gives is "no certificate
5231 supplied", which is not helpful.
5233 70. exigrep's output now also includes lines that are not associated with any
5234 message, but which match the given pattern. Implemented by a patch from
5235 Martin Sluka, which also tidied up the Perl a bit.
5237 71. Recipient callout verification, like sender verification, was using <> in
5238 the MAIL FROM command. This isn't really the right thing, since the actual
5239 sender may affect whether the remote host accepts the recipient or not. I
5240 have changed it to use the actual sender in the callout; this means that
5241 the cache record is now keyed on a recipient/sender pair, not just the
5242 recipient address. There doesn't seem to be a real danger of callout loops,
5243 since a callout by the remote host to check the sender would use <>.
5244 [SEE ABOVE: changed after hitting problems.]
5246 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5247 temporary errors. However, in the case of such a code being given after
5248 the end of a data transmission (i.e. after ".") Exim was failing to write
5249 a retry record for the message. (Yes, there was some broken host that was
5250 actually sending 8xx at this point.)
5252 73. An unknown lookup type in a host list could cause Exim to panic-die when
5253 the list was checked. (An example that provoked this was putting <; in the
5254 middle of a list instead of at the start.) If this happened during a DATA
5255 ACL check, a -D file could be left lying around. This kind of configuration
5256 error no longer causes Exim to die; instead it causes a defer errror. The
5257 incident is still logged to the main and panic logs.
5259 74. Buglet left over from Exim 3 conversion. The message "too many messages
5260 in one connection" was written to the rejectlog but not the mainlog, except
5261 when address rewriting (yes!) was being logged.
5263 75. Added write_rejectlog option.
5265 76. When a system filter was run not as root (that is, when system_filter_user
5266 was set), the values of the $n variables were not being returned to the
5267 main process; thus, they were not subsequently available in the $sn
5270 77. Added +return_path_on_delivery log selector.
5272 78. A connection timeout was being treated differently from recipients deferred
5273 when testing hosts_max_try with a message that was older than the host's
5274 retry timeout. (The host should not be counted, thus allowing all hosts to
5275 be tried at least once before bouncing.) This may have been the cause of an
5276 occasionally reported bug whereby a message would remain on the queue
5277 longer than the retry timeout, but would be bounced if a delivery was
5278 forced. I say "may" because I never totally pinned down the problem;
5279 setting up timeout/retry tests is difficult. See also the next item.
5281 79. The ultimate address timeout was not being applied to errors that involved
5282 a combination of host plus message (for example, a timeout on a MAIL
5283 command). When an address resolved to a number of possible hosts, and they
5284 were not all tried for each delivery (e.g. because of hosts_max_try), a
5285 message could remain on the queue longer than the retry timeout.
5287 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5290 81. Fixed an obscure SMTP outgoing bug which required at least the following
5291 conditions: (a) there was another message waiting for the same server;
5292 (b) the server returned 5xx to all RCPT commands in the first message so
5293 that the message was not completed; (c) the server dropped the connection
5294 or gave a negative response to the RSET that Exim sends to abort the
5295 transaction. The observed case was a dropped connection after DATA that had
5296 been sent in pipelining mode. That is, the server had advertised PIPELINING
5297 but was not implementing it correctly. The effect of the bug was incorrect
5298 behaviour, such as trying another host, and this could lead to a crash.
5304 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5305 and daemon.c were passed as pointers to ints; they should have been
5306 pointers to socklen_t variables (which are typically unsigned ints).
5308 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5311 3. Fixed a really odd bug that affected only the testing scheme; patching a
5312 certain fixed string in the binary changed the value of another string that
5313 happened to be identical to the end of the original first string.
5315 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5316 name", it returns that address as the IP address. On some operating
5317 systems (e.g. Solaris), it also passes back the IP address string as the
5318 "host name". However, on others (e.g. Linux), it passes back an empty
5319 string. Exim wasn't checking for this, and was changing the host name to an
5320 empty string, assuming it had been canonicized.
5322 5. Although rare, it is permitted to have more than one PTR record for a given
5323 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5324 all the names associated with an address, because they do in Solaris.
5325 However, it seems that they do not in Linux for data that comes from the
5326 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5327 I found this out when I moved to a new Linux workstation and tried to run
5328 the Exim test suite.
5330 To get round this problem I have changed the code so that it now does its
5331 own call to the DNS to look up PTR records when searching for a host name.
5332 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5333 addresses that are only in /etc/hosts are still found.
5335 This behaviour is, however, controlled by an option called host_lookup_
5336 order, which defaults to "bydns:byaddr". If people want to use the other
5337 order, or indeed, just use one or the other means of lookup, they can
5338 specify it in this variable.
5340 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5341 some operating systems, this comes back from gethostbyaddr() as an empty
5342 string, and this is what Exim used to test for. However, it seems that in
5343 other systems, "." is yielded. Exim now tests for this case too.
5345 7. The values of check_spool_space and check_log_space are now held internally
5346 as a number of kilobytes instead of an absolute number of bytes. If a
5347 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5348 kilobyte. This means that much larger values can be stored.
5350 8. Exim monitor: an attempt to get the action menu when not actually pointing
5351 at a message produces an empty menu entitled "No message selected". This
5352 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5353 no entries in it ("Shell widget menu has zero width and/or height"). So I
5354 have added a single, blank menu entry in this case.
5356 9. Added ${quote_local_part.
5358 10. MIME decoding is now applied to the contents of Subject: header lines when
5361 11. Now that a reference to $sender_host_address automatically causes a reverse
5362 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5363 host lookup before query-style lookups in lists that might use this
5364 variable. This has therefore been abolished, and the "net-" prefix is no
5365 longer necessary for query-style lookups.
5367 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5368 have been a typo for LFLAGS, so it has been changed.
5370 13. The install script calls Exim with "-C /dev/null" in order to find the
5371 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5372 to be output. Howeve, since Exim outputs its version number before the
5373 error, it didn't break the script. It just looked ugly. I fixed this by
5374 always allowing "-C /dev/null" if the caller is root.
5376 14. Ignore overlarge ACL variable number when reading spool file - insurance
5377 against a later release with more variables having written the file.
5379 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5380 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5381 this, as well as the form without the "IPv6" on the front.
5383 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5384 OS/Makefile-Darwin file.
5386 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5387 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5389 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5390 variable. However, some OS use ints for inodes. I've added cast to long int
5391 to get rid of the compiler warning.
5393 19. I had forgotten to lock out "/../" in configuration file names when
5394 ALT_CONFIG_PREFIX was set.
5396 20. Routers used for verification do not need to specify transports. However,
5397 if such a router generated a host list, and callout was configured, Exim
5398 crashed, because it could not find a port number from the (non-existent)
5399 transport. It now assumes port 25 in this circumstance.
5401 21. Added the -t option to exigrep.
5403 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5404 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5405 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5406 them is set, LOOKUP_LSEARCH is forced.
5408 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5410 24. Added sender and host information to the "rejected by local_scan()" log
5411 line; previously there was no indication of these.
5413 25. Added .include_if_exists.
5415 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5416 Linux. It turns out that not all file systems support this. Apparently some
5417 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5418 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5419 not supported on the file descriptor, is now ignored when Exim is trying to
5420 sync a directory. This applies only to Linux.
5422 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5424 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5425 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5428 29. Added support for maildirsize files from supplied patch (modified a bit).
5430 30. The use of :fail: followed by an empty string could lead Exim to respond to
5431 sender verification failures with (e.g.):
5433 550 Verification failed for <xxx>
5434 550 Sender verify failed
5436 where the first response line was missing the '-' that indicates it is not
5437 the final line of the response.
5439 31. The loop for finding the name of the user that called Exim had a hardwired
5440 limit of 10; it now uses the value of finduser_retries, which is used for
5441 all other user lookups.
5443 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5446 33. Exim was neglecting to zero errno before one call of strtol() when
5447 expanding a string and expecting an integer value. On some systems this
5448 resulted in spurious "integer overflow" errors. Also, it was casting the
5449 result into an int without checking.
5451 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5452 did not work. The code looks as if it has *never* worked, though it appears
5453 to have been documented since at least releast 1.62. I have made it work.
5455 35. The "timeout_DNS" error in retry rules, also documented since at least
5456 1.62, also never worked. As it isn't clear exactly what this means, and
5457 clearly it isn't a major issue, I have abolished the feature by treating it
5458 as "timeout", and writing a warning to the main and panic logs.
5460 36. The display of retry rules for -brt wasn't always showing the error code
5463 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5464 timeout_connect_A, timeout_connect_MX.
5466 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5467 to the empty sender.
5469 39. The daemon was not analysing the content of -oX till after it had closed
5470 stderr and disconnected from the controlling terminal. This meant that any
5471 syntax errors were only noted on the panic log, and the return code from
5472 the command was 0. By re-arranging the code a little, I've made the
5473 decoding happen first, so such errors now appear on stderr, and the return
5474 code is 1. However, the actual setting up of the sockets still happens in
5475 the disconnected process, so errors there are still only recorded on the
5478 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5479 connections (as happens on some IP stacks) was logged at start up time as
5480 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5481 it from "IPv6 and IPv4", which means that two separate sockets are being
5484 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5485 says whether AF_INET or AF_INET6 was passed as an argument.
5487 42. Exiwhat output was messed up when time zones were included in log
5490 43. Exiwhat now gives more information about the daemon's listening ports,
5491 and whether -tls-on-connect was used.
5493 44. The "port" option of the smtp transport is now expanded.
5495 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5496 silently ignored. Now an error message is written to the main and panic
5499 46. There's a new ACL modifier called "logwrite" which writes to a log file
5500 as soon as it is encountered.
5502 47. Added $local_user_uid and $local_user_gid at routing time.
5504 48. Exim crashed when trying to verify a sender address that was being
5507 49. Exim was recognizing only a space character after ".include". It now also
5508 recognizes a tab character.
5510 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5511 extracting the relevant information from the specification. The man page no
5512 longer contains scrambled data for the -d option, and I've added a section
5513 at the front about calling Exim under different names.
5515 51. Added "extra_headers" argument to the "mail" command in filter files.
5517 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5520 53. Installed eximstats 1.29.
5522 54. Added transport_filter_timeout as a generic transport option.
5524 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5525 Cc: header lines. This was required by RFC 822, but it not required by RFC
5528 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5529 incoming messages that did not have them. Now it does so only if the
5530 message originates locally, that is, if there is no associated remote host
5531 address. When Resent- header lines are present, this applies to the Resent-
5532 lines rather than the non-Resent- lines.
5534 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5535 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5537 58. Messages for configuration errors now include the name of the main
5538 configuration file - useful now that there may be more than one file in a
5539 list (.included file names were always shown).
5541 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5542 for those rare installations that do not start the daemon as root or run it
5543 setuid root. I've cut out the call to initgroups() if the daemon is not
5546 60. The Exim user and group can now be bound into the binary as text strings
5547 that are looked up at the start of Exim's processing.
5549 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5551 62. Added $mailstore_basename variable.
5553 63. Installed patch to sieve.c from Michael Haardt.
5555 64. When Exim failed to open the panic log after failing to open the main log,
5556 the original message it was trying to log was written to stderr and debug
5557 output, but if they were not available (the usual case in production), it
5558 was lost. Now it is written to syslog before the two lines that record the
5559 failures to open the logs.
5561 65. Users' Exim filters run in subprocesses under the user's uid. It is
5562 possible for a "deliver" command or an alias in a "personal" command to
5563 provoke an address rewrite. If logging of address rewriting is configured,
5564 this fails because the process is not running as root or exim. There may be
5565 a better way of dealing with this, but for the moment (because 4.30 needs
5566 to be released), I have disabled address rewrite logging when running a
5567 filter in a non-root, non-exim process.
5573 1. The buildconfig auxiliary program wasn't quoting the value set for
5574 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5575 not defined. This bug was masked in 4.22 by the effect that was fixed in
5578 2. Some messages that were rejected after a message id was allocated were
5579 shown as "incomplete" by exigrep. It no longer does this for messages that
5580 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5582 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5583 have allow_domain_literals set, the ID did not get logged in the <= line.
5584 Domain literals are now always recognized in Message-ID: header lines.
5586 4. The first argument for a ${extract expansion item is the key name or field
5587 number. Leading and trailing spaces in this item were not being ignored,
5588 causing some misleading effects.
5590 5. When deliver_drop_privilege was set, single queue runner processes started
5591 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5592 same command in the process it spins off) were not dropping privilege.
5594 6. When the daemon running as "exim" started a queue runner, it always
5595 re-executed Exim in the spun-off process. This is a waste of effort when
5596 deliver_drop_privilege is set. The new process now just calls the
5597 queue-runner function directly.
5603 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5606 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5607 ignored. Though the use of -oP was forcing the writing of a pid file, it
5608 was always written to the default place.
5610 3. If the message "no IP address found for host xxxx" is generated during
5611 incoming verification, it is now followed by identification of the incoming
5612 connection (so you can more easily find what provoked it).
5614 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5616 5. Added some features to "harden" Exim a bit more against certain attacks:
5618 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5619 be put in Local/Makefile. This is like the never_users runtime option,
5620 but it cannot be overridden. The default setting is "root".
5622 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5623 prefix string with which any file named in a -C command line option
5626 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5627 is retained for -C and -D only if the caller of Exim is root. Without
5628 it, the exim user may also use -C and -D and retain privilege.
5630 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5631 command line option is disabled.
5633 6. Macro names set by the -D option must start with an upper case letter, just
5634 like macro names defined in the configuration file.
5636 7. Added "dereference=" facility to LDAP.
5638 8. Two instances of the typo "uknown" in the source files are fixed.
5640 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5641 the Configure-Makefile script screwed up while processing it.
5643 10. Incorporated PCRE 4.4.
5645 11. The SMTP synchronization check was not operating right at the start of an
5646 SMTP session. For example, it could not catch a HELO sent before the client
5647 waited for the greeting. There is now a check for outstanding input at the
5648 point when the greeting is written. Because of the duplex, asynchronous
5649 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5650 way, but not yet received, when the check is performed.
5652 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5653 on TCP/IP sockets, because this apparently causes some broken clients to
5656 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5657 unchanged) from the Cygwin maintainer.
5659 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5660 was supported instead of testing for mbx. Effectively a typo.
5662 15. The spa authenticator server code was not checking that the input it
5663 received was valid base64.
5665 16. The debug output line for the "set" modifier in ACLs was not showing the
5666 name of the variable that was being set.
5668 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5670 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5671 lookup on its own. Something else was needed to trigger the lookup. For
5672 example, a match in host_lookup or the need for a host name in a host list.
5673 Now, if $sender_host_name is referenced and the host name has not yet been
5674 looked up, a lookup is performed. If the lookup fails, the variable remains
5675 empty, and $host_lookup_failed is set to "1".
5677 19. Added "eqi" as a case-independent comparison operator.
5679 20. The saslauthd authentication condition could segfault if neither service
5680 nor realm was specified.
5682 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5683 error message that was logged was misleading, and incoming SMTP
5684 connections were dropped. The message is now more accurate, and temporary
5685 errors are given to SMTP connections.
5687 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5688 (except RSET) with a 421 error, until QUIT is received. However, it was
5689 failing to send a response to QUIT.
5691 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5692 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5693 ACL is now run after the helo_try_verify_hosts code.
5695 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5696 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5697 case-independent, so other case variants are also recognized). Apparently
5698 some systems use these upper case variants.
5700 25. If more than two messages were waiting for the same host, and a transport
5701 filter was specified for the transport, Exim sent two messages over the
5702 same TCP/IP connection, and then failed with "socket operation on non-
5703 socket" when it tried to send the third.
5705 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5707 27. The extern definition of crypt16() in expand.c was not being excluded when
5708 the OS had its own crypt16() function.
5710 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5711 as a preferred synonym for return_size_limit, both as an option and as an
5714 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5716 30. Changed the default configuration ACL to relax the local part checking rule
5717 for addresses that are not in any local domains. For these addresses,
5718 slashes and pipe symbols are allowed within local parts, but the sequence
5719 /../ is explicitly forbidden.
5721 31. SPA server authentication was not clearing the challenge buffer before
5724 32. log_message in a "warn" ACL statement was writing to the reject log as
5725 well as to the main log, which contradicts the documentation and doesn't
5726 seem right (because no rejection is happening). So I have stopped it.
5728 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5729 However, I am unable to do any testing of this.
5731 34. Fixed an infelicity in the appendfile transport. When checking directories
5732 for a mailbox, to see if any needed to be created, it was accidentally
5733 using path names with one or more superfluous leading slashes; tracing
5734 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5736 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5737 contents are added to the log line that is written for every discarded
5738 recipient. (Previously a log_message setting was ignored.)
5740 36. The ${quote: operator now quotes the string if it is empty.
5742 37. The install script runs exim in order to find its version number. If for
5743 some reason other than non-existence or emptiness, which it checks, it
5744 could not run './exim', it was installing it with an empty version number,
5745 i.e. as "exim-". This error state is now caught, and the installation is
5748 38. An argument was missing from the function that creates an error message
5749 when Exim fails to connect to the socket for saslauthd authentication.
5750 This could cause Exim to crash, or give a corrupted message.
5752 39. Added isip, isip4, and isip6 to ${if conditions.
5754 40. The ACL variables $acl_xx are now saved with the message, and can be
5755 accessed later in routers, transports, and filters.
5757 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5758 strings in the file are not string-expanded.
5760 42. If a MAIL command specified a SIZE value that was too large to fit into an
5761 int variable, the check against message_size_limit failed. Such values are
5762 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5763 day this will have to be increased, but I don't think I want to be around
5764 when emails are that large.
5771 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5772 iconv() is not standard in FreeBSD.
5774 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5775 IPv6 enabled. The observed symptom was a segmentation fault on return from
5776 the function os_common_find_running_interfaces() in src/os.c.
5778 3. In the check_special_case() function in daemon.c I had used "errno" as an
5779 argument name, which causes warnings on some systems. This was basically a
5780 typo, since it was named "eno" in the comments!
5782 4. The code that waits for the clock to tick (at a resolution of some fraction
5783 of a second) so as to ensure message-id uniqueness was always waiting for
5784 at least one whole tick, when it could have waited for less. [This is
5785 almost certainly not relevant at current processor speeds, where it is
5786 unlikely to ever wait at all. But we try to future-proof.]
5788 5. The function that sleeps for a time interval that includes fractions of a
5789 second contained a race. It did not block SIGALRM between setting the
5790 timer, and suspending (a couple of lines later). If the interval was short
5791 and the sigsuspend() was delayed until after it had expired, the suspension
5792 never ended. On busy systems this could lead to processes getting stuck for
5795 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5796 process, before it forks any delivery processes. The open lookup caching
5797 mechanism meant that the open file or database connection was passed into
5798 the delivery process. The problem was that delivery processes always tidy
5799 up cached lookup data. This could cause a problem for the next delivery
5800 process started by the queue runner, because the external queue runner
5801 process does not know about the closure. So the next delivery process
5802 still has data in the lookup cache. In the case of a file lookup, there was
5803 no problem because closing a file descriptor in a subprocess doesn't affect
5804 the parent. However, if the lookup was caching a connection to a database,
5805 the connection was closed, and the second delivery process was likely to
5806 see errors such as "PGSQL: query failed: server closed the connection
5807 unexpectedly". The problem has been fixed by closing all cached lookups
5808 in a queue runner before running a delivery process.
5810 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5811 seem to have the "const" qualifier which it has on other OS. I've
5814 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5815 *of the same type* (client or server) with the same public name that an
5816 error should be diagnosed.
5818 9. When Exim looked up a host name for an IP address, but failed to find the
5819 original IP address when looking up the host name (a safety check), it
5820 output the message "<ip address> does not match any IP for NULL", which was
5821 confusing, to say the least. The bug was that the host name should have
5822 appeared instead of "NULL".
5824 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5825 user that is built into the binary, and the -C or -D options is used, root
5826 privilege is dropped before the configuration file is read. In addition,
5827 logging is switched to stderr instead of the normal log files. If the
5828 configuration then re-defines the Exim user, the unprivileged environment
5829 is probably not what is expected, so Exim logs a panic warning message (but
5832 However, if deliver_drop_privilege is set, the unprivileged state may well
5833 be exactly what is intended, so the warning has been cut out in that case,
5834 and Exim is allowed to try to write to its normal log files.
5840 1. smtp_return_error_details was not giving details for temporary sender
5841 or receiver verification errors.
5843 2. Diagnose a configuration error if two authenticators have the same public
5846 3. Exim used not to create the message log file for a message until the first
5847 delivery attempt. This could be confusing when incoming messages were held
5848 for policy or load reasons. The message log file is now created at the time
5849 the message is received, and an initial "Received" line is written to it.
5851 4. The automatically generated man page for command line options had a minor
5852 bug that caused no ill effects; however, a more serious problem was that
5853 the procedure for building the man page automatically didn't always
5854 operate. Consequently, release 4.20 contains an out-of-date version. This
5855 shouldn't happen again.
5857 5. When building Exim with embedded Perl support, the script that builds the
5858 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5859 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5861 6. The freeze_tell option was not being used for messages that were frozen on
5862 arrival, either by an ACL or by local_scan().
5864 7. Added the smtp_incomplete_transaction log selector.
5866 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5867 was accepting AUTH without a new EHLO.
5869 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5870 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5873 10. Logging of TCP/IP connections (when configured) now happens in the main
5874 daemon process instead of the child process, so that the TCP/IP connection
5875 count is more accurate (but it can never be perfect).
5877 11. The use of "drop" in a nested ACL was not being handled correctly in the
5878 outer ACL. Now, if condition failure induced by the nested "drop" causes
5879 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5880 or "require"), the connection is dropped.
5882 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5883 that yield "discard" can now be used with an "accept" or a "discard" verb,
5884 but an error is generated for any others (because I can't see a useful way
5885 to define what should happen).
5887 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5888 are now processed in the same way as lines in the Exim configuration file.
5889 In particular, continuation lines are supported.
5891 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5893 15. Added -ti meaning -t -i.
5895 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5896 domains, and warn by logging if others are found.
5898 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5899 returned from GIFCONF when Exim is trying to find the list of interfaces on
5900 a host. The code in os.c has been modified to copy each ifreq to an aligned
5901 structure in all cases.
5903 Also, in some cases, the returned ifreq's were being copied to a 'struct
5904 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5905 means the last couple of bytes of an IPv6 address could be chopped if the
5906 ifreq contained only a normal sockaddr (14 bytes storage).
5908 18. Named domain lists were not supported in the hosts_treat_as_local option.
5909 An entry such as +xxxx was not recognized, and was treated as a literal
5912 19. Ensure that header lines added by a DATA ACL are included in the reject log
5913 if the ACL subsequently rejects the message.
5915 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5916 MD5 (which is deprecated).
5918 21. When testing a filter file using -bf, Exim was writing a message when it
5919 took the sender from a "From " line in the message, but it was not doing so
5920 when it took $return_path from a Return-Path: header line. It now does.
5922 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5923 with a valid header line field name (a series of printing characters
5924 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5926 23. Changed "disc" in the source to "disk" to conform to the documentation and
5927 the book and for uniformity.
5929 24. Ignore Sendmail's -Ooption=value command line item.
5931 25. When execve() failed while trying to run a command in a pipe transport,
5932 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5933 could be confused with a return value of 69 from the command itself. This
5934 has been changed to 127, the value the shell returns if it is asked to run
5935 a non-existent command. The wording for the related log line suggests a
5936 non-existent command as the problem.
5938 26. If received_header_text expands to an empty string, do not add a Received:
5939 header line to the message. (Well, it adds a token one on the spool, but
5940 marks it "old" so that it doesn't get used or transmitted.)
5942 27. Installed eximstats 1.28 (addition of -nt option).
5944 28. There was no check for failure on the call to getsockname() in the daemon
5945 code. This can fail if there is a shortage of resources on the system, with
5946 ENOMEM, for example. A temporary error is now given on failure.
5948 29. Contrary to the C standard, it seems that in some environments, the
5949 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5950 program. Exim now does this explicitly; it affects the formatting of
5951 timestamps using strftime().
5953 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5954 complaints. I've now initialized all the variables, to avoid this.
5956 32. Header lines added by a system filter were not being "seen" during
5957 transport-time rewrites.
5959 33. The info_callback() function passed to OpenSSL is set up with type void
5960 (*)(SSL *, int, int), as described somewhere. However, when calling the
5961 function (actually a macro) that sets it up, the type void(*)() is
5962 expected. I've put in a cast to prevent warnings from picky compilers.
5964 34. If a DNS black list lookup found a CNAME record, but there were no A
5965 records associated with the domain it pointed at, Exim crashed.
5967 35. If a DNS black list lookup returned more than one A record, Exim ignored
5968 all but the first. It now scans all returned addresses if a particular IP
5969 value is being sought. In this situation, the contents of the
5970 $dnslist_value variable are a list of all the addresses, separated by a
5973 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5974 to accept a host name and all its aliases if the forward lookup for any of
5975 them yielded the IP address of the incoming connection. Now it accepts only
5976 those names whose forward lookup yields the correct IP address. Any other
5977 names are discarded. This closes a loophole whereby a rogue DNS
5978 administrator could create reverse DNS records to break through a
5979 wildcarded host restriction in an ACL.
5981 37. If a user filter or a system filter that ran in a subprocess used any of
5982 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5983 the wrong values were passed to the pipe command ($thisaddress had the
5984 value of $0, $0 had the value of $1, etc). This bug was introduced by
5985 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5987 38. Improved the line breaking for long SMTP error messages from ACLs.
5988 Previously, if there was no break point between 40 and 75 characters, Exim
5989 left the rest of the message alone. Two changes have been made: (a) I've
5990 reduced the minimum length to 35 characters; (b) if it can't find a break
5991 point between 35 and 75 characters, it looks ahead and uses the first one
5992 that it finds. This may give the occasional overlong line, but at least the
5993 remaining text gets split now.
5995 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5996 file descriptors might be low, and that setting 1000 would always raise it.
5997 It turns out that in some environments, the limit is already over 1000 and
5998 that lowering it causes trouble. So now Exim takes care not to decrease it.
6000 40. When delivering a message, the value of $return_path is set to $sender_
6001 address at the start of routing (routers may change the value). By an
6002 oversight, this default was not being set up when an address was tested by
6003 -bt or -bv, which affected the outcome if any router or filter referred to
6006 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6007 log only when "message" or "log_message" are set. However, if one of the
6008 conditions was an address verification, or a call to a nested ACL, the
6009 messages generated by the underlying test were being passed through. This
6010 no longer happens. The underlying message is available in $acl_verify_
6011 message for both "message" and "log_message" expansions, so it can be
6012 passed through if needed.
6014 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6015 new expansion $bh_ to give the encoded byte string without charset
6016 translation. Translation happens only if iconv() is available; HAVE_ICONV
6017 indicates this at build time. HEADERS_CHARSET gives the charset to
6018 translate to; headers_charset can change it in the configuration, and
6019 "headers charset" can change it in an individual filter file.
6021 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6022 that creates RFC 2047 encoded "words" labels them as that charset instead
6023 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6024 expansion operator; (ii) when Exim creates a From: line for a local
6025 message; (iii) when a header line is rewritten to include a "phrase" part.
6027 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6028 buggy, causing it to skip the first lines of messages whose message ID
6029 ended in 'D'. This would not have bitten before Exim release 4.14, because
6030 message IDs were unlikely to end in 'D' before then. The effect was to have
6031 incorrect size information for certain domains.
6033 45. #include "config.h" was missing at the start of the crypt16.c module. This
6034 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6037 46. If there was a timeout during a "random" callout check, Exim treated it as
6038 a failure of the random address, and carried on sending RSET and the real
6039 address. If the delay was just some slowness somewhere, the response to the
6040 original RCPT would be taken as a response to RSET and so on, causing
6041 mayhem of various kinds.
6043 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6044 when I implemented it. It didn't allow for the fact that some option values
6045 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6046 the right test for positive values.
6048 48. Domain names in DNS records are case-independent. Exim always looks them up
6049 in lower case. Some resolvers return domain names in exactly the case they
6050 appear in the zone file, that is, they may contain uppercase letters. Not
6051 all resolvers do this - some return always lower case. Exim was treating a
6052 change of case by a resolver as a change of domain, similar to a widening
6053 of a domain abbreviation. This triggered its re-routing code and so it was
6054 trying to route what was effectively the same domain again. This normally
6055 caused routing to fail (because the router wouldn't handle the domain
6056 twice). Now Exim checks for this case specially, and just changes the
6057 casing of the domain that it ultimately uses when it transmits the message
6060 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6063 50. If a filter generated a file delivery with a non-absolute name (possible if
6064 no home directory exists for the router), the forbid_file option was not
6067 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6068 the existing equality matching.
6070 52. Exim was using ints instead of ino_t variables in some places where it was
6071 dealing with inode numbers.
6073 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6074 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6075 called TMPDIR, and if it finds it is different, it changes its value.
6077 54. The smtp_printf() function is now made available to local_scan() so
6078 additional output lines can be written before returning. There is also an
6079 smtp_fflush() function to enable the detection of a dropped connection.
6080 The variables smtp_input and smtp_batched_input are exported to
6083 55. Changed the default runtime configuration: the message "Unknown user"
6084 has been removed from the ACL, and instead placed on the localuser router,
6085 using the cannot_route_message feature. This means that any verification
6086 failures that generate their own messages won't get overridden. Similarly,
6087 the "Unrouteable address" message that was in the ACL for unverifiable
6088 relay addresses has also been removed.
6090 56. Added hosts_avoid_esmtp to the smtp transport.
6092 57. The exicyclog script was not checking for the esoteric option
6093 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6094 will work only if exicyclog is run under the appropriate euid.
6096 58. Following a discussion on the list, the rules by which Exim recognises line
6097 endings on incoming messages have been changed. The -dropcr and drop_cr
6098 options are now no-ops, retained only for backwards compatibility. The
6099 following line terminators are recognized: LF CRLF CR. However, special
6100 processing applies to CR:
6102 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6103 nor a local message in the state where . is a terminator.
6105 (ii) If a bare CR is encountered in a header line, an extra space is added
6106 after the line terminator so as not to end the header. The reasoning
6107 behind this is that bare CRs in header lines are most likely either
6108 to be mistakes, or people trying to play silly games.
6110 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6111 was being incorrectly given as 18 bytes larger than it should have been.
6112 This is a VOB (very old bug).
6114 60. This may never have affected anything current, but just in case it has:
6115 When the local host is found other than at the start of a list of hosts,
6116 the local host, those with the same MX, and any that follow, are discarded.
6117 When the list in question was part of a longer list of hosts, the following
6118 hosts (not currently being processed) were also being discarded. This no
6119 longer happens. I'm not sure if this situation could ever has previously
6122 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6125 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6126 header. This is something that is recommended in a new Internet Draft, and
6127 is something that is documented as being done by Sendmail. There are two
6128 possible values. For messages generated by the autoreply transport, Exim
6131 Auto-Submitted: auto-replied
6133 whereas for all other generated messages (e.g. bounces) it adds
6135 Auto-Submitted: auto-generated
6137 63. The "personal" condition in filters now includes a test for the
6138 Auto-Submitted: header. If it contains the string "auto-" the message it
6139 not considered personal.
6141 64. Added rcpt_include_affixes as a generic transport option.
6143 65. Added queue_only_override (default true).
6145 66. Added the syslog_duplication option.
6147 67. If what should have been the first header line of a message consisted of
6148 a space followed by a colon, Exim was mis-interpreting it as a header line.
6149 It isn't of course - it is syntactically invalid and should therefore be
6150 treated as the start of the message body. The misbehaviour could have
6151 caused a number of strange effects, including loss of data in subsequent
6152 header lines, and spool format errors.
6154 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6155 client host had authenticated. This control can now be exercised by an ACL
6156 for more flexibility.
6158 69. By default, callouts do not happen when testing with -bh. There is now a
6159 variant, -bhc, which does actually run the callout code, including
6160 consulting and updating the callout cache.
6162 70. Added support for saslauthd authentication, courtesy of Alexander
6165 71. If statvfs() failed on the spool or log directories while checking their
6166 size for availability, Exim confusingly gave the error "space shortage".
6167 Furthermore, in debugging mode it crashed with a floating point exception.
6168 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6169 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6170 serious problem, Exim now writes to the main and panic logs when this
6171 happens, with details of the failure. It then refuses to accept the
6172 incoming message, giving the message "spool directory problem" or "log
6173 directory problem" with a 421 code for SMTP messages.
6175 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6176 0, 1, and 2 exist, because some OS complain for execs without them (see
6177 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6178 descriptors. However, the code omitted to check that the open succeeded,
6179 causing mysterious errors if for some reason the permissions on /dev/null
6180 got screwed. Now Exim writes a message to the main and panic logs, and
6181 bombs out if it can't open /dev/null.
6183 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6184 interact so that it is all more flexible. It is supposed to remain
6185 backwards compatible. Also added extra_local_interfaces.
6187 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6188 to bomb out with an assertion failure - to the client this appears as a
6189 connection drop. This problem occurs in the part of the code that was taken
6190 from the Samba project. Fortunately, the assertion is in a very simple
6191 function, so I have fixed this by reproducing the function inline in the
6192 one place where it is called, and arranging for authentication to fail
6193 instead of killing the process with assert().
6195 75. The SPA client code was not working when the server requested OEM rather
6196 than Unicode encoding.
6198 76. Added code to make require_files with a specific uid setting more usable in
6199 the case where statting the file as root fails - usually a non-root-mounted
6200 NFS file system. When this happens and the failure is EACCES, Exim now
6201 forks a subprocess and does the per-uid checking as the relevant uid.
6203 77. Added process_log_path.
6205 78. If log_file_path was not explicitly set, a setting of check_log_space or
6206 check_log_inodes was ignored.
6208 79. If a space check for the spool or log partitions fails, the incident is now
6209 logged. Of course, in the latter case the data may get lost...
6211 80. Added the %p formatting code to string_format() so that it can be used to
6212 print addresses in debug_print(). Adjusted all the address printing in the
6213 debugging in store.c to use %p rather than %d.
6215 81. There was a concern that a line of code in smtp_in.c could overflow a
6216 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6217 initially expressed, the concern was not well-founded, because trailing
6218 spaces are removed early. However, if the trailing spaces were followed by
6219 a NULL, they did not get removed, so the overflow was possible. Two fixes
6222 (a) I re-wrote the offending code in a cleaner fashion.
6223 (b) If an incoming SMTP command contains a NULL character, it is rejected
6226 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6227 runs initgroups(), so that if the Exim user is in any additional groups,
6228 they will be used during message reception.
6234 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob