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18 EXPERIMENTAL_CERTNAMES is withdrawn.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
35 TL/01 When running the test suite, the README says that variables such as
36 no_msglog_check are global and can be placed anywhere in a specific
37 test's script, however it was observed that placement needed to be near
38 the beginning for it to behave that way. Changed the runtest perl
39 script to read through the entire script once to detect and set these
40 variables, reset to the beginning of the script, and then run through
41 the script parsing/test process like normal.
43 TL/02 The BSD's have an arc4random API. One of the functions to induce
44 adding randomness was arc4random_stir(), but it has been removed in
45 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
46 function when detected.
48 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
49 cause callback expansion.
51 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
52 syntax errors in an expansion can be treated as a string instead of
53 logging or causing an error, due to the internal use of bool_lax
54 instead of bool when processing it.
56 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
57 server certificates when making smtp deliveries.
59 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
61 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
63 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
65 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
66 Merged patch from Sebastian Wiedenroth.
68 JH/05 Fix results-pipe from transport process. Several recipients, combined
69 with certificate use, exposed issues where response data items split
70 over buffer boundaries were not parsed properly. This eventually
71 resulted in duplicates being sent. This issue only became common enough
72 to notice due to the introduction of conection certificate information,
73 the item size being so much larger. Found and fixed by Wolfgang Breyha.
75 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
76 size buffer was used, resulting in syntax errors when an expansion
79 JH/07 Add support for directories of certificates when compiled with a GnuTLS
80 version 3.3.6 or later.
82 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
83 is EXPERIMENTAL_EVENT, the main-configuration and transport options
84 both become "event_action", the variables become $event_name, $event_data
85 and $event_defer_errno. There is a new variable $verify_mode, usable in
86 routers, transports and related events. The tls:cert event is now also
87 raised for inbound connections, if the main configuration event_action
90 TL/06 In test suite, disable OCSP for old versions of openssl which contained
91 early OCSP support, but no stapling (appears to be less than 1.0.0).
93 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
94 server certificate names available under the smtp transport option
95 "tls_verify_cert_hostname" now do not permit multi-component wildcard
98 JH/10 Time-related extraction expansions from certificates now use the main
99 option "timezone" setting for output formatting, and are consistent
100 between OpenSSL and GnuTLS compilations. Bug 1541.
102 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
103 encoded parameter in the incoming message. Bug 1558.
105 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
106 include certificate info, eximon was claiming there were spoolfile
109 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
111 JH/14 Log delivery-related information more consistently, using the sequence
112 "H=<name> [<ip>]" wherever possible.
114 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
115 are problematic for Debian distribution, omit them from the release
118 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
120 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
123 JH/17 Fix a null-indirection in certextract expansions when a nondefault
124 output list separator was used.
129 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
130 checkers that were complaining about end of non-void function with no
133 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
134 This was a regression intruduced in 4.83 by another bugfix.
136 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
138 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
139 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
145 TF/01 Correctly close the server side of TLS when forking for delivery.
147 When a message was received over SMTP with TLS, Exim failed to clear up
148 the incoming connection properly after forking off the child process to
149 deliver the message. In some situations the subsequent outgoing
150 delivery connection happened to have the same fd number as the incoming
151 connection previously had. Exim would try to use TLS and fail, logging
152 a "Bad file descriptor" error.
154 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
155 utilities have not been installed.
157 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
158 temporary space as the ACL may create new global variables.
160 TL/01 LDAP support uses per connection or global context settings, depending
161 upon the detected version of the libraries at build time.
163 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
164 to extract and use the src ip:port in logging and expansions as if it
165 were a direct connection from the outside internet. PPv2 support was
166 updated based on HAProxy spec change in May 2014.
168 JH/02 Add ${listextract {number}{list}{success}{fail}}.
170 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
171 Properly escape header and check for NULL return.
173 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
176 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
178 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
179 characters in header names, implemented as a verify condition.
180 Contributed by Michael Fischer v. Mollard.
182 TL/05 Rename SPF condition results err_perm and err_temp to standardized
183 results permerror and temperror. Previous values are deprecated but
184 still accepted. In a future release, err_perm and err_temp will be
185 completely removed, which will be a backward incompatibility if the
186 ACL tests for either of these two old results. Patch contributed by
187 user bes-internal on the mailing list.
189 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
191 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
192 selectors, in both main and reject logs.
194 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
197 JH/07 Add malware type "sock" for talking to simple daemon.
199 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
201 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
202 routers/transports under cutthrough routing.
204 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
205 numbers. Touch up "bool" conditional to keep the same definition.
207 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
209 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
211 JH/12 Expand items in router/transport headers_add or headers_remove lists
212 individually rather than the list as a whole. Bug 1452.
214 Required for reasonable handling of multiple headers_ options when
215 they may be empty; requires that headers_remove items with embedded
216 colons must have them doubled (or the list-separator changed).
218 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
219 view the policy declared in the DMARC record. Currently, $dmarc_status
220 is a combined value of both the record presence and the result of the
223 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
225 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
226 dnslookup router and the smtp transport (applying to the forward
229 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
230 of ldap servers used for a specific lookup. Patch provided by Heiko
233 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
234 New variable $lookup_dnssec_authenticated for observability.
236 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
237 Patch submitted by Lars Timman.
239 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
241 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
242 Requires trusted mode and valid format message id, aborts otherwise.
243 Patch contributed by Heiko Schlichting.
245 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
246 certextract with support for various fields. Bug 1358.
248 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
249 is requested by default, modifiable by smtp transport option
252 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
253 operate on certificate variables to give certificate fingerprints
254 Also new ${sha256:cert_variable}.
256 JH/23 The PRDR feature is moved from being Experimental into the mainline.
258 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
259 Christian Aistleitner.
261 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
263 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
264 file. Patch from Wolfgang Breyha.
266 JH/25 Expand the coverage of the delivery $host and $host_address to
267 client authenticators run in verify callout. Bug 1476.
269 JH/26 Port service names are now accepted for tls_on_connect_ports, to
270 align with daemon_smtp_ports. Bug 72.
272 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
273 support and error reporting did not work properly.
275 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
276 and is readable. Patch from Andrew Colin Kissa.
278 TL/14 Enhance documentation of ${run expansion and how it parses the
279 commandline after expansion, particularly in the case when an
280 unquoted variable expansion results in an empty value.
282 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
284 PP/02 Fix internal collision of T_APL on systems which support RFC3123
285 by renaming away from it. Addresses GH issue 15, reported by
288 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
290 TL/15 SECURITY: prevent double expansion in math comparison functions
291 (can expand unsanitized data). Not remotely exploitable.
298 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
300 PP/02 Make -n do something, by making it not do something.
301 When combined with -bP, the name of an option is not output.
303 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
306 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
307 $sender_host_name and config options to manage this, and basic check
310 PP/05 DSCP support for outbound connections and control modifier for inbound.
312 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
313 (Only plugin which currently uses this is kerberos4, which nobody should
314 be using, but we should make it available and other future plugins might
315 conceivably use it, even though it would break NAT; stuff *should* be
316 using channel bindings instead).
318 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
319 name; added for Sendmail compatibility; requires admin caller.
320 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
321 just ignore it); requires trusted caller.
322 Also parse but ignore: -Ac -Am -X<logfile>
325 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
327 TL/02 Add +smtp_confirmation as a default logging option.
329 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
330 Patch by Magnus Holmgren from 2007-02-20.
332 TL/04 Bugzilla 1281 - Spec typo.
333 Bugzilla 1283 - Spec typo.
334 Bugzilla 1290 - Spec grammar fixes.
336 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
338 TL/06 Add Experimental DMARC support using libopendmarc libraries.
340 TL/07 Fix an out of order global option causing a segfault. Reported to dev
341 mailing list by by Dmitry Isaikin.
343 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
345 JH/02 Support "G" suffix to numbers in ${if comparisons.
347 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
349 NM/01 Bugzilla 1197 - Spec typo
350 Bugzilla 1196 - Spec examples corrections
352 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
354 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
355 gnutls_enable_pkcs11, but renamed to more accurately indicate its
358 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
359 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
361 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
362 "acl {{name}{arg}...}", and optional args on acl condition
365 JH/05 Permit multiple router/transport headers_add/remove lines.
367 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
369 JH/07 Avoid using a waiting database for a single-message-only transport.
370 Performance patch from Paul Fisher. Bugzilla 1262.
372 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
375 JH/09 Add $headers_added variable, with content from use of ACL modifier
376 add_header (but not yet added to the message). Bugzilla 199.
378 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
379 Pulled from Bugzilla 817 by Wolfgang Breyha.
381 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
383 (nb: this is the same fix as in Exim 4.80.1)
385 JH/11 Add A= logging on delivery lines, and a client_set_id option on
388 JH/12 Add optional authenticated_sender logging to A= and a log_selector
391 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
393 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
394 advertise SMTP AUTH mechanism to us, instead of a generic
395 protocol violation error. Also, make Exim more robust to bad
396 data from the Dovecot auth socket.
398 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
400 When a queue runner is handling a message, Exim first routes the
401 recipient addresses, during which it prunes them based on the retry
402 hints database. After that it attempts to deliver the message to
403 any remaining recipients. It then updates the hints database using
406 So if a recipient address works intermittently, it can get repeatedly
407 deferred at routing time. The retry hints record remains fresh so the
408 address never reaches the final cutoff time.
410 This is a fairly common occurrence when a user is bumping up against
411 their storage quota. Exim had some logic in its local delivery code
412 to deal with this. However it did not apply to per-recipient defers
413 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
415 This change adds a proper retry rule check during routing so that the
416 final cutoff time is checked against the message's age. We only do
417 this check if there is an address retry record and there is not a
418 domain retry record; this implies that previous attempts to handle
419 the address had the retry_use_local_parts option turned on. We use
420 this as an approximation for the destination being like a local
421 delivery, as in LMTP.
423 I suspect this new check makes the old local delivery cutoff check
424 redundant, but I have not verified this so I left the code in place.
426 TF/02 Correct gecos expansion when From: is a prefix of the username.
428 Test 0254 submits a message to Exim with the header
432 When I ran the test suite under the user fanf2, Exim expanded
433 the header to contain my full name, whereas it should have added
434 a Resent-Sender: header. It erroneously treats any prefix of the
435 username as equal to the username.
437 This change corrects that bug.
439 GF/01 DCC debug and logging tidyup
440 Error conditions log to paniclog rather than rejectlog.
441 Debug lines prefixed by "DCC: " to remove any ambiguity.
443 TF/03 Avoid unnecessary rebuilds of lookup-related code.
445 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
446 Bug spotted by Jeremy Harris; was flawed since initial commit.
447 Would have resulted in OCSP responses post-SNI triggering an Exim
448 NULL dereference and crash.
450 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
452 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
453 Bug detection, analysis and fix by Samuel Thibault.
454 Bugzilla 1331, Debian bug #698092.
456 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
458 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
459 Server implementation by Todd Lyons, client by JH.
460 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
461 config variable "prdr_enable" controls whether the server
462 advertises the facility. If the client requests PRDR a new
463 acl_data_smtp_prdr ACL is called once for each recipient, after
464 the body content is received and before the acl_smtp_data ACL.
465 The client is controlled by bolth of: a hosts_try_prdr option
466 on the smtp transport, and the server advertisement.
467 Default client logging of deliveries and rejections involving
468 PRDR are flagged with the string "PRDR".
470 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
471 fclose(). Diagnosis by Todd Lyons.
473 PP/17 Update configure.default to handle IPv6 localhost better.
474 Patch by Alain Williams (plus minor tweaks).
477 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
478 This is now consistent with GnuTLS, and is now documented: the
479 previous undocumented portable approach to treating the option as
480 unset was to force an expansion failure. That still works, and
481 an empty string is now equivalent.
483 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
484 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
485 not performing validation itself.
487 PP/20 Added force_command boolean option to pipe transport.
488 Patch from Nick Koston, of cPanel Inc.
490 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
493 TF/04 Added udpsend ACL modifer and hexquote expansion operator
495 PP/21 Fix eximon continuous updating with timestamped log-files.
496 Broken in a format-string cleanup in 4.80, missed when I repaired the
497 other false fix of the same issue.
498 Report and fix from Heiko Schlichting.
501 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
502 Report from Prashanth Katuri.
504 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
505 It's SecureTransport, so affects any MacOS clients which use the
506 system-integrated TLS libraries, including email clients.
508 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
509 using a MIME ACL for non-SMTP local injection.
510 Report and assistance in diagnosis by Warren Baker.
512 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
514 JH/16 Fix comparisons for 64b. Bugzilla 1385.
516 TL/09 Add expansion variable $authenticated_fail_id to keep track of
517 last id that failed so it may be referenced in subsequent ACL's.
519 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
522 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
523 ldap library initialization, allowing self-signed CA's to be
524 used. Also properly sets require_cert option later in code by
525 using NULL (global ldap config) instead of ldap handle (per
526 session). Bug diagnosis and testing by alxgomz.
528 TL/12 Enhanced documentation in the ratelimit.pl script provided in
529 the src/util/ subdirectory.
531 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
532 renamed to Transport Post Delivery Action by Jeremy Harris, as
535 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
536 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
537 redis_servers = needs to be configured which will be used by the redis
538 lookup. Patch from Warren Baker, of The Packet Hub.
540 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
542 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
543 hostname or reverse DNS when processing a host list. Used suggestions
544 from multiple comments on this bug.
546 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
548 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
549 Missed a few lines, added it to make the runtest require no keyboard
552 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
553 contains upper case chars. Make router use caseful_local_part.
555 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
556 support when GnuTLS has been built with p11-kit.
562 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
564 This, or similar/improved, will also be change PP/11 of 4.82.
570 PP/01 Handle short writes when writing local log-files.
571 In practice, only affects FreeBSD (8 onwards).
572 Bugzilla 1053, with thanks to Dmitry Isaikin.
574 NM/01 Bugzilla 949 - Documentation tweak
576 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
579 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
581 PP/02 Implemented gsasl authenticator.
583 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
585 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
586 `pkg-config foo` for cflags/libs.
588 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
589 with rest of GSASL and with heimdal_gssapi.
591 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
592 `pkg-config foo` for cflags/libs for the TLS implementation.
594 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
595 properties get this fed in as external SSF. A number of robustness
596 and debugging improvements to the cyrus_sasl authenticator.
598 PP/08 cyrus_sasl server now expands the server_realm option.
600 PP/09 Bugzilla 1214 - Log authentication information in reject log.
601 Patch by Jeremy Harris.
603 PP/10 Added dbmjz lookup type.
605 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
607 PP/12 MAIL args handles TAB as well as SP, for better interop with
608 non-compliant senders.
609 Analysis and variant patch by Todd Lyons.
611 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
612 Bug report from Lars Müller <lars@samba.org> (via SUSE),
613 Patch from Dirk Mueller <dmueller@suse.com>
615 PP/13 tls_peerdn now print-escaped for spool files.
616 Observed some $tls_peerdn in wild which contained \n, which resulted
617 in spool file corruption.
619 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
620 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
621 or write after TLS renegotiation, which otherwise led to messages
624 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
625 as a tracking header (ie: a signed header comes before the signature).
626 Patch from Wolfgang Breyha.
628 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
629 comma-sep list; embedded commas doubled.
631 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
633 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
635 Report and patch from Dmitry Banschikov.
637 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
638 Removed SSL_clear() after SSL_new() which led to protocol negotiation
639 failures. We appear to now support TLS1.1+ with Exim.
641 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
642 lets Exim select keys and certificates based upon TLS SNI from client.
643 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
644 before an outbound SMTP session. New log_selector, +tls_sni.
646 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
647 NULL dereference. Report and patch from Alun Jones.
649 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
650 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
651 Not seeing resolver debug output on NetBSD, but suspect this is a
652 resolver implementation change.
654 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
655 Left warnings. Added "eximon gdb" invocation mode.
657 PP/21 Defaulting "accept_8bitmime" to true, not false.
659 PP/22 Added -bw for inetd wait mode support.
661 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
662 locate the relevant includes and libraries. Made this the default.
664 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
665 Bugzilla 1246, report and most of solution from Tomasz Kusy.
667 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
668 This may cause build issues on older platforms.
670 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
671 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
672 gnutls_require_mac & gnutls_require_protocols (no longer supported).
673 Added SNI support via GnuTLS too.
674 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
676 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
678 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
679 Applied second patch from Janne, implementing suggestion to default
680 multiple-strings-in-record handling to match SPF spec.
682 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
684 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
685 read-only, out of scope).
686 Patch from Wolfgang Breyha, report from Stuart Northfield.
688 PP/29 Fix three issues highlighted by clang analyser static analysis.
689 Only crash-plausible issue would require the Cambridge-specific
690 iplookup router and a misconfiguration.
691 Report from Marcin Mirosław.
693 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
695 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
696 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
697 As part of this, removing so much warning spew let me fix some minor
698 real issues in debug logging.
700 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
701 assignment on my part. Fixed.
703 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
704 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
705 Janne Snabb (who went above and beyond: thank you).
707 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
708 string otherwise requires a connection and a bunch more work and it's
709 relatively easy to get wrong. Should also expose TLS library linkage
712 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
713 64-bit ${eval} (JH/03).
715 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
716 GNU libc to support some of the 64-bit stuff, should not lead to
717 conflicts. Defined before os.h is pulled in, so if a given platform
718 needs to override this, it can.
720 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
721 protection layer was required, which is not implemented.
722 Bugzilla 1254, patch from Wolfgang Breyha.
724 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
725 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
726 tls_dhparam take prime identifiers. Also unbreak combination of
727 OpenSSL+DH_params+TLSSNI.
729 PP/39 Disable SSLv2 by default in OpenSSL support.
735 PP/01 Solaris build fix for Oracle's LDAP libraries.
736 Bugzilla 1109, patch from Stephen Usher.
738 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
740 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
743 TF/02 Fix a couple more cases where we did not log the error message
744 when unlink() failed. See also change 4.74-TF/03.
746 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
747 lock up or crash if it happened to be inside a call to libc when it
748 got a SIGUSR1 from exiwhat.
750 The SIGUSR1 handler appends the current process status to the process
751 log which is later printed by exiwhat. It used to use the general
752 purpose logging code to do this, but several functions it calls are
753 not safe for signals.
755 The new output code in the SIGUSR1 handler is specific to the process
756 log, and simple enough that it's easy to inspect for signal safety.
757 Removing some special cases also simplifies the general logging code.
758 Removing the spurious timestamps from the process log simplifies
761 TF/04 Improved ratelimit ACL condition.
763 The /noupdate option has been deprecated in favour of /readonly which
764 has clearer semantics. The /leaky, /strict, and /readonly update modes
765 are mutually exclusive. The update mode is no longer included in the
766 database key; it just determines when the database is updated. (This
767 means that when you upgrde Exim will forget old rate measurements.)
769 Exim now checks that the per_* options are used with an update mode that
770 makes sense for the current ACL. For example, when Exim is processing a
771 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
772 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
773 must specify per_mail/readonly. If you omit the update mode it defaults to
774 /leaky where that makes sense (as before) or /readonly where required.
776 The /noupdate option is now undocumented but still supported for
777 backwards compatibility. It is equivalent to /readonly except that in
778 ACLs where /readonly is required you may specify /leaky/noupdate or
779 /strict/noupdate which are treated the same as /readonly.
781 A useful new feature is the /count= option. This is a generalization
782 of the per_byte option, so that you can measure the throughput of other
783 aggregate values. For example, the per_byte option is now equivalent
784 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
786 The per_rcpt option has been generalized using the /count= mechanism
787 (though it's more complicated than the per_byte equivalence). When it is
788 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
789 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
790 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
791 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
792 Note that using per_rcpt with a non-readonly update mode in more than
793 one ACL will cause the recipients to be double-counted. (The per_mail
794 and per_byte options don't have this problem.)
796 The handling of very low rates has changed slightly. If the computed rate
797 is less than the event's count (usually one) then this event is the first
798 after a long gap. In this case the rate is set to the same as this event's
799 count, so that the first message of a spam run is counted properly.
801 The major new feature is a mechanism for counting the rate of unique
802 events. The new per_addr option counts the number of different
803 recipients that someone has sent messages to in the last time period. It
804 behaves like per_rcpt if all the recipient addresses are different, but
805 duplicate recipient addresses do not increase the measured rate. Like
806 the /count= option this is a general mechanism, so the per_addr option
807 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
808 example, measure the rate that a client uses different sender addresses
809 with the options per_mail/unique=$sender_address. There are further
810 details in the main documentation.
812 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
814 TF/06 Removed a few PCRE remnants.
816 TF/07 Automatically extract Exim's version number from tags in the git
817 repository when doing development or release builds.
819 PP/02 Raise smtp_cmd_buffer_size to 16kB.
820 Bugzilla 879. Patch from Paul Fisher.
822 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
823 Heavily based on revision 40f9a89a from Simon Arlott's tree.
826 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
828 PP/05 Variable $av_failed, true if the AV scanner deferred.
829 Bugzilla 1078. Patch from John Horne.
831 PP/06 Stop make process more reliably on build failure.
832 Bugzilla 1087. Patch from Heiko Schlittermann.
834 PP/07 Make maildir_use_size_file an _expandable_ boolean.
835 Bugzilla 1089. Patch from Heiko Schlittermann.
837 PP/08 Handle ${run} returning more data than OS pipe buffer size.
838 Bugzilla 1131. Patch from Holger Weiß.
840 PP/09 Handle IPv6 addresses with SPF.
841 Bugzilla 860. Patch from Wolfgang Breyha.
843 PP/10 GnuTLS: support TLS 1.2 & 1.1.
845 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
848 PP/11 match_* no longer expand right-hand-side by default.
849 New compile-time build option, EXPAND_LISTMATCH_RHS.
850 New expansion conditions, "inlist", "inlisti".
852 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
854 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
856 PP/14 fix log_write() format string regression from TF/03.
857 Bugzilla 1152. Patch from Dmitry Isaikin.
863 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
865 PP/02 Harmonised TLS library version reporting; only show if debugging.
866 Layout now matches that introduced for other libraries in 4.74 PP/03.
868 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
870 PP/04 New "dns_use_edns0" global option.
872 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
875 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
876 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
878 TK/01 Updated PolarSSL code to 0.14.2.
879 Bugzilla 1097. Patch from Andreas Metzler.
881 PP/07 Catch divide-by-zero in ${eval:...}.
884 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
887 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
888 format-string attack -- SECURITY: remote arbitrary code execution.
890 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
891 time unintentionally subject to list matching rules, letting the header
892 cause arbitrary Exim lookups (of items which can occur in lists, *not*
893 arbitrary string expansion). This allowed for information disclosure.
895 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
896 INT_MIN/-1 -- value coerced to INT_MAX.
902 NM/01 Workround for PCRE version dependency in version reporting
905 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
906 This fixes portability to compilers other than gcc, notably
907 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
909 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
910 makefiles for portability to HP-UX and POSIX correctness.
912 PP/01 Permit LOOKUP_foo enabling on the make command-line.
913 Also via indented variable definition in the Makefile.
914 (Debugging by Oliver Heesakkers).
916 PP/02 Restore caching of spamd results with expanded spamd_address.
917 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
919 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
920 Improves build reliability. Fix from: Frank Elsner
922 NM/02 Fix wide character breakage in the rfc2047 coding
923 Fixes bug 1064. Patch from Andrey N. Oktyabrski
925 NM/03 Allow underscore in dnslist lookups
926 Fixes bug 1026. Patch from Graeme Fowler
928 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
929 Code patches from Adam Ciarcinski of NetBSD.
931 NM/04 Fixed exiqgrep to cope with mailq missing size issue
934 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
935 is logged, to avoid truncation. Patch from John Horne.
937 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
938 Patch from Jakob Hirsch.
940 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
941 SQL string expansion failure details.
942 Patch from Andrey Oktyabrski.
944 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
945 Patch from Simon Arlott.
947 PP/09 New lookups functionality failed to compile on old gcc which rejects
948 extern declarations in function scope.
949 Patch from Oliver Fleischmann
951 PP/10 Use sig_atomic_t for flags set from signal handlers.
952 Check getgroups() return and improve debugging.
953 Fixed developed for diagnosis in bug 927 (which turned out to be
956 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
957 Patch from Mark Zealey.
959 PP/12 Bugzilla 1056: Improved spamd server selection.
960 Patch from Mark Zealey.
962 PP/13 Bugzilla 1086: Deal with maildir quota file races.
963 Based on patch from Heiko Schlittermann.
965 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
966 Patch from Uwe Doering, sign-off by Michael Haardt.
968 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
969 variable declaration deep within a block. Bug and patch from
972 PP/15 lookups-Makefile IRIX compatibilty coercion.
974 PP/16 Make DISABLE_DKIM build knob functional.
976 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
977 Patch by Simon Arlott
979 TF/03 Fix valgrind.h portability to C89 compilers that do not support
980 variable argument macros. Our copy now differs from upstream.
986 TF/01 Failure to get a lock on a hints database can have serious
987 consequences so log it to the panic log.
989 TF/02 Log LMTP confirmation messages in the same way as SMTP,
990 controlled using the smtp_confirmation log selector.
992 TF/03 Include the error message when we fail to unlink a spool file.
994 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
995 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
996 for maintaining out-of-tree patches for some time.
998 PP/01 Bugzilla 139: Documentation and portability issues.
999 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1000 Handle per-OS dynamic-module compilation flags.
1002 PP/02 Let /dev/null have normal permissions.
1003 The 4.73 fixes were a little too stringent and complained about the
1004 permissions on /dev/null. Exempt it from some checks.
1005 Reported by Andreas M. Kirchwitz.
1007 PP/03 Report version information for many libraries, including
1008 Exim version information for dynamically loaded libraries. Created
1009 version.h, now support a version extension string for distributors
1010 who patch heavily. Dynamic module ABI change.
1012 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1013 privilege escalation vulnerability whereby the Exim run-time user
1014 can cause root to append content of the attacker's choosing to
1017 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1020 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1021 If dropping privileges for untrusted macros, we disabled normal logging
1022 on the basis that it would fail; for the Exim run-time user, this is not
1023 the case, and it resulted in successful deliveries going unlogged.
1024 Fixed. Reported by Andreas Metzler.
1030 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1031 only prepend for the Resent-* case. Fixes regression introduced in
1032 Exim 4.70 by NM/22 for Bugzilla 607.
1034 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1035 increasing numbers of administrators be bitten by this.
1037 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1039 PP/03 Bugzilla 994: added openssl_options main configuration option.
1041 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1043 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1045 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1047 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1048 without a peer certificate, leading to a segfault because of an
1049 assumption that peers always have certificates. Be a little more
1050 paranoid. Problem reported by Martin Tscholak.
1052 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1053 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1054 NB: ClamAV planning to remove STREAM in "middle of 2010".
1055 CL also introduces -bmalware, various -d+acl logging additions and
1056 more caution in buffer sizes.
1058 PP/09 Implemented reverse_ip expansion operator.
1060 PP/10 Bugzilla 937: provide a "debug" ACL control.
1062 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1064 PP/12 Bugzilla 973: Implement --version.
1066 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1068 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1070 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1072 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1073 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1074 ignore trailing whitespace.
1076 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1078 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1081 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1082 Notification from Dr Andrew Aitchison.
1084 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1085 ExtendedDetectionInfo response format.
1086 Notification from John Horne.
1088 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1091 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1092 XSL and documented dependency on system catalogs, with examples of how
1095 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1098 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1099 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1100 configuration file which is writeable by the Exim user or group.
1102 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1103 of configuration files to cover files specified with the -C option if
1104 they are going to be used with root privileges, not just the default
1107 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1108 option (effectively making it always true).
1110 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1111 files to be used while preserving root privileges.
1113 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1114 that rogue child processes cannot use them.
1116 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1117 run-time user, instead of root.
1119 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1120 Exim run-time user without dropping privileges.
1122 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1123 result string, instead of calling string_vformat() twice with the same
1126 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1127 for other users. Others should always drop root privileges if they use
1128 -C on the command line, even for a whitelisted configure file.
1130 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1132 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1138 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1139 $data_path, and $header_path variables; fixed documentation bugs and
1142 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1143 exipick to access non-standard spools, including the "frozen" queue
1146 NM/01 Bugzilla 965: Support mysql stored procedures.
1147 Patch from Alain Williams
1149 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1151 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1152 Patch from Andreas Metzler
1154 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1155 Patch from Kirill Miazine
1157 NM/05 Bugzilla 671: Added umask to procmail example.
1159 JJ/03 installed exipick 20100323.0, fixing doc bug
1161 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1162 directory. Notification and patch from Dan Rosenberg.
1164 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1166 TK/02 Improve log output when DKIM signing operation fails.
1168 MH/01 Treat the transport option dkim_domain as a colon separated
1169 list, not as a single string, and sign the message with each element,
1170 omitting multiple occurences of the same signer.
1172 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1173 Bugzilla 985, 986. Patch by Simon Arlott
1175 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1176 Patch by Simon Arlott
1178 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1179 MBX locking. Notification from Dan Rosenberg.
1185 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1187 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1189 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1191 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1193 NM/04 Bugzilla 915: Flag broken perl installation during build.
1199 TK/01 Added patch by Johannes Berg that expands the main option
1200 "spamd_address" if it starts with a dollar sign.
1202 TK/02 Write list of recipients to X-Envelope-Sender header when building
1203 the mbox-format spool file for content scanning (suggested by Jakob
1206 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1207 (http://www.dcc-servers.net/) support via dccifd. Activated by
1208 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1210 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1211 by Mark Daniel Reidel <mr@df.eu>.
1213 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1214 When building exim an external PCRE library is now needed -
1215 PCRE is a system library on the majority of modern systems.
1216 See entry on PCRE_LIBS in EDITME file.
1218 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1219 conversation. Added nologin parameter to request.
1220 Patch contributed by Kirill Miazine.
1222 TF/01 Do not log submission mode rewrites if they do not change the address.
1224 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1226 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1227 log files in place. Contributed by Roberto Lima.
1229 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1231 TF/03 Bugzilla 615: When checking the local_parts router precondition
1232 after a local_part_suffix or local_part_prefix option, Exim now
1233 does not use the address's named list lookup cache, since this
1234 contains cached lookups for the whole local part.
1236 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1237 Robert Millan. Documentation is in experimental-spec.txt.
1239 TF/04 Bugzilla 668: Fix parallel build (make -j).
1241 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1243 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1244 Patch provided by Jan Srzednicki.
1246 TF/05 Leading white space used to be stripped from $spam_report which
1247 wrecked the formatting. Now it is preserved.
1249 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1250 that they are available at delivery time.
1252 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1254 TF/08 TLS error reporting now respects the incoming_interface and
1255 incoming_port log selectors.
1257 TF/09 Produce a more useful error message if an SMTP transport's hosts
1258 setting expands to an empty string.
1260 NM/06 Bugzilla 744: EXPN did not work under TLS.
1261 Patch provided by Phil Pennock.
1263 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1264 Patch provided by Richard Godbee.
1266 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1267 acl_smtp_notquit, added index entry.
1269 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1270 Patch provided by Eugene Bujak.
1272 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1273 accept(). Patch provided by Maxim Dounin.
1275 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1276 Patch provided by Phil Pennock.
1278 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1280 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1281 Patch provided by Brad "anomie" Jorsch.
1283 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1284 Patch provided by Dean Brooks.
1286 TK/05 Add native DKIM support (does not depend on external libraries).
1288 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1289 Patch provided by Graeme Fowler.
1291 NM/16 Bugzilla 851: Documentation example syntax fix.
1293 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1295 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1298 NM/19 Bugzilla 745: TLS version reporting.
1299 Patch provided by Phil Pennock.
1301 NM/20 Bugzilla 167: bool: condition support.
1302 Patch provided by Phil Pennock.
1304 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1305 clients. Patch provided by Phil Pennock.
1307 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1308 Patch provided by Brad "anomie" Jorsch.
1310 NM/23 Bugzilla 687: Fix misparses in eximstats.
1311 Patch provided by Heiko Schlittermann.
1313 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1314 Patch provided by Heiko Schlittermann.
1316 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1317 plus update to original patch.
1319 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1321 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1322 Patch provided by David Brownlee.
1324 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1326 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1328 NM/30 Bugzilla 888: TLS documentation bugfixes.
1330 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1332 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1333 Unlike the original bugzilla I have changed all shell scripts in src tree.
1335 NM/33 Bugzilla 898: Transport filter timeout fix.
1336 Patch by Todd Rinaldo.
1338 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1339 Patch by Serge Demonchaux.
1341 NM/35 Bugzilla 39: Base64 decode bug fixes.
1342 Patch by Jakob Hirsch.
1344 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1346 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1348 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1350 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1356 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1357 ALT-N's libdkim that I have put here:
1358 http://duncanthrax.net/exim-experimental/
1360 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1361 were called 'true' and it seems that C99 defines that as a reserved
1362 keyword to be used with 'bool' variable types. That means you could
1363 not include C99-style headers which use bools without triggering
1364 build errors in sieve.c.
1366 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1367 as mailq or other aliases. Changed the --help handling significantly
1368 to do whats expected. exim_usage() emits usage/help information.
1370 SC/01 Added the -bylocaldomain option to eximstats.
1372 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1374 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1376 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1382 PH/01 Another patch from the Sieve maintainer.
1384 PH/02 When an IPv6 address is converted to a string for single-key lookup
1385 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1386 dots are used instead of colons so that keys in lsearch files need not
1387 contain colons. This was done some time before quoting was made available
1388 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1389 (notated using the quote facility) so as to distinguish them from IPv4
1390 keys. This meant that lookups for IP addresses in host lists did not work
1391 for iplsearch lookups.
1393 This has been fixed by arranging for IPv6 addresses to be expressed with
1394 colons if the lookup type is iplsearch. This is not incompatible, because
1395 previously such lookups could never work.
1397 The situation is now rather anomolous, since one *can* have colons in
1398 ordinary lsearch keys. However, making the change in all cases is
1399 incompatible and would probably break a number of configurations.
1401 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1404 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1405 conversion specification without a maximum field width, thereby enabling
1406 a rogue spamd server to cause a buffer overflow. While nobody in their
1407 right mind would setup Exim to query an untrusted spamd server, an
1408 attacker that gains access to a server running spamd could potentially
1409 exploit this vulnerability to run arbitrary code as the Exim user.
1411 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1412 $primary_hostname instead of what libspf2 thinks the hosts name is.
1414 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1415 a directory entry by the name of the lookup key. Previously, if a
1416 symlink pointed to a non-existing file or a file in a directory that
1417 Exim lacked permissions to read, a lookup for a key matching that
1418 symlink would fail. Now it is enough that a matching directory entry
1419 exists, symlink or not. (Bugzilla 503.)
1421 PH/03 The body_linecount and body_zerocount variables are now exported in the
1424 PH/04 Added the $dnslist_matched variable.
1426 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1427 This means they are set thereafter only if the connection becomes
1430 PH/06 Added the client_condition to authenticators so that some can be skipped
1431 by clients under certain conditions.
1433 PH/07 The error message for a badly-placed control=no_multiline_responses left
1434 "_responses" off the end of the name.
1436 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1438 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1439 (without spaces) instead of just copying the configuration text.
1441 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1443 PH/11 Added $max_received_linelength.
1445 PH/12 Added +ignore_defer and +include_defer to host lists.
1447 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1448 way in which PCRE > 7.0 is built.
1450 PH/14 Implemented queue_only_load_latch.
1452 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1453 MAIL command. The effect was to mangle the value on 64-bit systems.
1455 PH/16 Another patch from the Sieve maintainer.
1457 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1459 PH/18 If a system quota error occurred while trying to create the file for
1460 a maildir delivery, the message "Mailbox is full" was not appended to the
1461 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1462 only to a quota excession during the actual writing of the file.
1464 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1465 characters?) which causes problems in log lines. The DN values are now
1466 passed through string_printing() before being added to log lines.
1468 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1469 and InterBase are left for another time.)
1471 PH/21 Added message_body_newlines option.
1473 PH/22 Guard against possible overflow in moan_check_errorcopy().
1475 PH/23 POSIX allows open() to be a macro; guard against that.
1477 PH/24 If the recipient of an error message contained an @ in the local part
1478 (suitably quoted, of course), incorrect values were put in $domain and
1479 $local_part during the evaluation of errors_copy.
1485 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1486 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1489 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1490 issue a MAIL command.
1492 PH/02 In an ACL statement such as
1494 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1496 if a client was not listed at all, or was listed with a value other than
1497 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1498 the condition was not true (as it should be), so access was not denied.
1499 The bug was that the ! inversion was incorrectly passed on to the second
1500 item. This has been fixed.
1502 PH/03 Added additional dnslists conditions == and =& which are different from
1503 = and & when the dns lookup returns more than one IP address.
1505 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1506 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1508 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1509 FSYNC, which compiles an option called disable_fsync that allows for
1510 bypassing fsync(). The documentation is heavily laced with warnings.
1512 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1514 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1515 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1516 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1517 including adding "make clean"; (3) Added -fPIC when compiling the test
1518 dynamically loaded module, to get rid of a warning.
1520 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1521 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1522 = 0s. The bug is otherwise harmless.
1524 PH/07 There was a bug in the dovecot authenticator such that the value of
1525 $auth1 could be overwritten, and so not correctly preserved, after a
1526 successful authentication. This usually meant that the value preserved by
1527 the server_setid option was incorrect.
1529 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1531 PH/09 Installed PCRE release 7.0.
1533 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1534 run for batched SMTP input. It is now run at the start of every message
1535 in the batch. While fixing this I discovered that the process information
1536 (output by running exiwhat) was not always getting set for -bs and -bS
1537 input. This is fixed, and it now also says "batched" for BSMTP.
1539 PH/11 Added control=no_pipelining.
1541 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1542 patch, slightly modified), and move the expansion of helo_data till after
1543 the connection is made in the smtp transport (so it can use these
1546 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1548 PH/14 Added log_selector = +pid.
1550 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1552 PH/16 Add ${if forany and ${if forall.
1554 PH/17 Added dsn_from option to vary the From: line in DSNs.
1556 PH/18 Flush SMTP output before performing a callout, unless control =
1557 no_callout_flush is set.
1559 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1560 was true (the default) a successful delivery failed to delete the retry
1561 item, thus causing premature timeout of the address. The bug is now
1564 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1566 PH/21 Long custom messages for fakedefer and fakereject are now split up
1567 into multiline reponses in the same way that messages for "deny" and
1568 other ACL rejections are.
1570 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1571 with slight modification.
1573 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1574 draft, changing the syntax and factoring some duplicate code".
1576 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1577 for deliveries of the second and subsequent messages over the same SMTP
1580 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1581 ${reduce, with only minor "tidies".
1583 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1585 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1586 expansion side effects.
1588 PH/27 When a message times out after an over-quota error from an Exim-imposed
1589 quota, the bounce message says "mailbox is full". This message was not
1590 being given when it was a system quota that was exceeded. It now should
1593 MH/03 Made $recipients available in local_scan(). local_scan() already has
1594 better access to the recipient list through recipients_list[], but
1595 $recipients can be useful in postmaster-provided expansion strings.
1597 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1598 in the case of a MAIL command with additional options following the
1599 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1600 were accidentally chopped off.
1602 PH/29 SMTP synchronization checks are implemented when a command is read -
1603 there is a check that no more input is waiting when there shouldn't be
1604 any. However, for some commands, a delay in an ACL can mean that it is
1605 some time before the response is written. In this time, more input might
1606 arrive, invalidly. So now there are extra checks after an ACL has run for
1607 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1608 pipelining has not been advertised.
1610 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1612 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1613 returned in the SMTP response when smtp_return_error_details was false.
1614 This has been fixed.
1616 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1617 instead of the C I/O that was originally supplied, because problems were
1618 reported on Solaris.
1620 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1621 Exim which did not show up earlier: it was assuming that a call to
1622 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1623 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1624 was a macro that became an assignment, so it seemed to work. This has
1625 changed to a proper function call with a void return, hence the compile
1626 error. Exim's code has been fixed.
1628 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1631 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1633 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1635 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1636 because it turns out to be too big or there is a timeout) there is no
1637 "Completed" line in the log. When some messages of this type were
1638 selected by exigrep, they were listed as "not completed". Others were
1639 picked up by some special patterns. I have improved the selection
1640 criteria to be more general.
1642 PH/38 The host_find_failed option in the manualroute router can now be set
1643 to "ignore", to completely ignore a host whose IP address cannot be
1644 found. If all hosts are ignored, the behaviour is controlled by the new
1645 host_all_ignored option.
1647 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1648 homing or because of multiple MX records with /mx) generated more than
1649 one IP address, and the following item turned out to be the local host,
1650 all the secondary addresses of the first item were incorrectly removed
1651 from the list, along with the local host and any following hosts (which
1652 is what is supposed to happen).
1654 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1655 whoever called Exim into the -H file. In the case of the daemon it was
1656 behaving confusingly. When first started, it used values for whoever
1657 started the daemon, but after a SIGHUP it used the Exim user (because it
1658 calls itself on a restart). I have changed the code so that it now always
1661 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1662 message are rejected with the same error (e.g. no authentication or bad
1663 sender address), and a DATA command is nevertheless sent (as can happen
1664 with PIPELINING or a stupid MUA), the error message that was given to the
1665 RCPT commands is included in the rejection of the DATA command. This is
1666 intended to be helpful for MUAs that show only the final error to their
1669 PH/42 Another patch from the Sieve maintainer.
1671 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1672 Eximstats - Fixed some broken HTML links and added missing column headers
1674 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1675 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1677 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1679 PH/43 Yet another patch from the Sieve maintainer.
1681 PH/44 I found a way to check for a TCP/IP connection going away before sending
1682 the response to the final '.' that terminates a message, but only in the
1683 case where the client has not sent further data following the '.'
1684 (unfortunately, this is allowed). However, in many cases there won't be
1685 any further data because there won't be any more messages to send. A call
1686 to select() can be used: if it shows that the input is "ready", there is
1687 either input waiting, or the socket has been closed. An attempt to read
1688 the next input character can distinguish the two cases. Previously, Exim
1689 would have sent an OK response which the client would never have see.
1690 This could lead to message repetition. This fix should cure that, at
1691 least in a lot of common cases.
1693 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1694 advertised in response to EHLO.
1700 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1701 fixed by 4.65/MH/01 (is this a record?) are fixed:
1703 (i) An empty string was always treated as zero by the numeric comparison
1704 operators. This behaviour has been restored.
1706 (ii) It is documented that the numeric comparison operators always treat
1707 their arguments as decimal numbers. This was broken in that numbers
1708 starting with 0 were being interpreted as octal.
1710 While fixing these problems I realized that there was another issue that
1711 hadn't been noticed. Values of message_size_limit (both the global option
1712 and the transport option) were treated as octal if they started with 0.
1713 The documentation was vague. These values are now always treated as
1714 decimal, and I will make that clear in the documentation.
1720 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1721 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1724 MH/01 Don't check that the operands of numeric comparison operators are
1725 integers when their expansion is in "skipping" mode (fixes bug
1726 introduced by 4.64-PH/07).
1728 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1729 child addresses, Exim now panics and dies. Previously, because the count
1730 is held in a short int, deliveries were likely to be lost. As such a
1731 large number of recipients for a single message is ridiculous
1732 (performance will be very, very poor), I have chosen to impose a limit
1733 rather than extend the field.
1739 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1740 leftover -K file (the existence of which was triggered by #402).
1741 While we were at it, introduced process PID as part of the -K
1742 filename. This should rule out race conditions when creating
1745 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1746 processing considerably. Previous code took too long for large mails,
1747 triggering a timeout which in turn triggers #401.
1749 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1750 in the DK code in transports.c. sendfile() is not really portable,
1751 hence the _LINUX specificness.
1753 TF/01 In the add_headers option to the mail command in an Exim filter,
1754 there was a bug that Exim would claim a syntax error in any
1755 header after the first one which had an odd number of characters
1758 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1759 callout verification, Exim cached a "reject" for the entire domain. This
1760 is correct for most verifications, but it is not correct for a recipient
1761 verification with use_sender or use_postmaster set, because in that case
1762 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1763 case of MAIL FROM:<> rejection from other early rejections (e.g.
1764 rejection of HELO). When verifying a recipient using a non-null MAIL
1765 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1766 Whatever the result of the callout, the value of the domain cache is
1767 left unchanged (for any other kind of callout, getting as far as trying
1768 RCPT means that the domain itself is ok).
1770 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1773 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1774 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1775 session, the daemon ignores ECONNECT errors and logs others; it now
1776 ignores EPIPE as well.
1778 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1779 (quoted-printable decoding).
1781 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1782 later the small subsequent patch to fix an introduced bug.
1784 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1786 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1788 PH/08 An error is now given if message_size_limit is specified negative.
1790 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1791 to be given (somewhat) arbitrary names.
1793 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1796 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1797 miscellaneous code fixes
1799 PH/10 Added the log_reject_target ACL modifier to specify where to log
1802 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1803 hostname. This is wrong, because it relates to the incoming message (and
1804 probably the interface on which it is arriving) and not to the outgoing
1805 callout (which could be using a different interface). This has been
1806 changed to use the value of the helo_data option from the smtp transport
1807 instead - this is what is used when a message is actually being sent. If
1808 there is no remote transport (possible with a router that sets up host
1809 addresses), $smtp_active_hostname is used.
1811 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1812 tweaks were necessary in order to get it to work (see also 21 below):
1813 (a) The code assumed that strncpy() returns a negative number on buffer
1814 overflow, which isn't the case. Replaced with Exim's string_format()
1816 (b) There were several signed/unsigned issues. I just did the minimum
1817 hacking in of casts. There is scope for a larger refactoring.
1818 (c) The code used strcasecmp() which is not a standard C function.
1819 Replaced with Exim's strcmpic() function.
1820 (d) The code set only $1; it now sets $auth1 as well.
1821 (e) A simple test gave the error "authentication client didn't specify
1822 service in request". It would seem that Dovecot has changed its
1823 interface. Fortunately there's a specification; I followed it and
1824 changed what the client sends and it appears to be working now.
1826 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1829 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1830 address A is aliased to B and C, where B exists and C does not. Without
1831 -v the output is "A verified" because verification stops after a
1832 successful redirection if more than one address is generated. However,
1833 with -v the child addresses are also verified. Exim was outputting "A
1834 failed to verify" and then showing the successful verification for C,
1835 with its parentage. It now outputs "B failed to verify", showing B's
1836 parentage before showing the successful verification of C.
1838 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1839 look up a TXT record in a specific list after matching in a combined
1842 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1843 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1844 they consult the DNS. I had assumed they would set it the way they
1845 wanted; and indeed my experiments on Linux seem to show that in some
1846 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1847 To be on the safe side, however, I have now made the interface to
1848 host_find_byname() similar to host_find_bydns(), with an argument
1849 containing the DNS resolver options. The host_find_byname() function now
1850 sets these options at its start, just as host_find_bydns() does. The smtp
1851 transport options dns_qualify_single and dns_search_parents are passed to
1852 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1853 of host_find_byname() use the default settings of RES_DEFNAMES
1854 (qualify_single) but not RES_DNSRCH (search_parents).
1856 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1857 spool_read_header() do less string testing, by means of a preliminary
1858 switch on the second character of optional "-foo" lines. (This is
1859 overdue, caused by the large number of possibilities that now exist.
1860 Originally there were few.) While I was there, I also converted the
1861 str(n)cmp tests so they don't re-test the leading "-" and the first
1862 character, in the hope this might squeeze out yet more improvement.
1864 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1865 flag allowing group syntax was set by the header_syntax check but not
1866 turned off, possible causing trouble later; (2) The flag was not being
1867 set at all for the header_verify test, causing "group"-style headers to
1868 be rejected. I have now set it in this case, and also caused header_
1869 verify to ignore an empty address taken from a group. While doing this, I
1870 came across some other cases where the code for allowing group syntax
1871 while scanning a header line wasn't quite right (mostly, not resetting
1872 the flag correctly in the right place). These bugs could have caused
1873 trouble for malformed header lines. I hope it is now all correct.
1875 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1876 with the "reply" argument non-NULL. The code, however (which originally
1877 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1878 but it didn't always do it. This confused somebody who was copying the
1879 code for some other use. I have removed all the tests.
1881 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1882 feature that was used to support insecure browsers during the U.S. crypto
1883 embargo. It requires special client support, and Exim is probably the
1884 only MTA that supported it -- and would never use it because real RSA is
1885 always available. This code has been removed, because it had the bad
1886 effect of slowing Exim down by computing (never used) parameters for the
1887 RSA_EXPORT functionality.
1889 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1890 authenticator to fail if there's a tab character in the incoming data
1891 (there should never be unless someone is messing about, as it's supposed
1892 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1893 if the connection is using TLS or if the remote IP is the same as the
1894 local IP, and the "valid-client-cert option" if a client certificate has
1897 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1898 authenticators. This can be used for authorization after authentication
1899 succeeds. (In the case of plaintext, it servers for both authentication
1902 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1903 if any retry times were supplied.
1905 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1906 connection, where there is no client IP address to check. In this
1907 situation, the verify now always succeeds.
1909 PH/25 Applied John Jetmore's -Mset patch.
1911 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1913 PH/27 In a string expansion for a processed (not raw) header when multiple
1914 headers of the same name were present, leading whitespace was being
1915 removed from all of them, but trailing whitespace was being removed only
1916 from the last one. Now trailing whitespace is removed from each header
1917 before concatenation. Completely empty headers in a concatenation (as
1918 before) are ignored.
1920 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1921 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1923 PH/29 [Removed. This was a change that I later backed out, and forgot to
1924 correct the ChangeLog entry (that I had efficiently created) before
1925 committing the later change.]
1927 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1928 address errors (4xx response to RCPT) over the same connection as other
1929 messages routed to the same hosts. Such deliveries are always "forced",
1930 so retry times are not inspected. This resulted in far too many retries
1931 for the affected addresses. The effect occurred only when there were more
1932 hosts than the hosts_max_try setting in the smtp transport when it had
1933 the 4xx errors. Those hosts that it had tried were not added to the list
1934 of hosts for which the message was waiting, so if all were tried, there
1935 was no problem. Two fixes have been applied:
1937 (i) If there are any address or message errors in an SMTP delivery, none
1938 of the hosts (tried or untried) are now added to the list of hosts
1939 for which the message is waiting, so the message should not be a
1940 candidate for sending over the same connection that was used for a
1941 successful delivery of some other message. This seems entirely
1942 reasonable: after all the message is NOT "waiting for some host".
1943 This is so "obvious" that I'm not sure why it wasn't done
1944 previously. Hope I haven't missed anything, but it can't do any
1945 harm, as the worst effect is to miss an optimization.
1947 (ii) If, despite (i), such a delivery is accidentally attempted, the
1948 routing retry time is respected, so at least it doesn't keep
1949 hammering the server.
1951 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1952 in ${readsocket because some servers need this prod.
1954 PH/32 Added some extra debug output when updating a wait-xxx database.
1956 PH/33 The hint "could be header name not terminated by colon", which has been
1957 given for certain expansion errors for a long time, was not being given
1958 for the ${if def:h_colon_omitted{... case.
1960 PH/34 The spec says: "With one important exception, whenever a domain list is
1961 being scanned, $domain contains the subject domain." There was at least
1962 one case where this was not true.
1964 PH/35 The error "getsockname() failed: connection reset by peer" was being
1965 written to the panic log as well as the main log, but it isn't really
1966 panic-worthy as it just means the connection died rather early on. I have
1967 removed the panic log writing for the ECONNRESET error when getsockname()
1970 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1971 runs only) independently of the message's sender address. This meant
1972 that, if the 4xx error was in fact related to the sender, a different
1973 message to the same recipient with a different sender could confuse
1974 things. In particualar, this can happen when sending to a greylisting
1975 server, but other circumstances could also provoke similar problems.
1976 I have changed the default so that the retry time for these errors is now
1977 based a combination of the sender and recipient addresses. This change
1978 can be overridden by setting address_retry_include_sender=false in the
1981 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1982 remote server are returned as part of bounce messages. This was not
1983 happening for LMTP over a pipe (the lmtp transport), but now it is the
1984 same for both kinds of LMTP.
1986 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1987 in header lines that were in fact CNAMEs. This is no longer the case.
1989 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1990 and queue runs started by the daemon processed all messages. This has
1991 been fixed so that -R and -S can now usefully be given with -q<time>.
1993 PH/40 Import PCRE release 6.7 (fixes some bugs).
1995 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1997 PH/42 Give an error if -q is specified more than once.
1999 PH/43 Renamed the variables $interface_address and $interface_port as
2000 $received_ip_address and $received_port, to make it clear that these
2001 values apply to message reception, and not to the outgoing interface when
2002 a message is delivered. (The old names remain recognized, of course.)
2004 PH/44 There was no timeout on the connect() call when using a Unix domain
2005 socket in the ${readsocket expansion. There now is.
2007 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2008 be meaningful with "accept".
2010 SC/01 Eximstats V1.43
2011 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2013 SC/02 Eximstats V1.44
2014 Use a glob alias rather than an array ref in the generated
2015 parser. This improves both readability and performance.
2017 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2018 Collect SpamAssassin and rejection statistics.
2019 Don't display local sender or destination tables unless
2020 there is data to show.
2021 Added average volumes into the top table text output.
2023 SC/04 Eximstats V1.46
2024 Collect data on the number of addresses (recipients)
2025 as well as the number of messages.
2027 SC/05 Eximstats V1.47
2028 Added 'Message too big' to the list of mail rejection
2029 reasons (thanks to Marco Gaiarin).
2031 SC/06 Eximstats V1.48
2032 Mainlog lines which have GMT offsets and are too short to
2033 have a flag are now skipped.
2035 SC/07 Eximstats V1.49 (Alain Williams)
2036 Added the -emptyok flag.
2038 SC/08 Eximstats V1.50
2039 Fixes for obtaining the IP address from reject messages.
2041 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2042 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2043 whitesspace changes from 4.64-PH/27
2045 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2048 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2049 are found, allow negative numbers in numeric criteria)
2051 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2053 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2054 to match changes made in 4.64-PH/43
2056 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2058 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2059 "domain" is actually an IP address, and force a failure. This locks out
2060 those revolvers/nameservers that support "A-for-A" lookups, in
2061 contravention of the specifications.
2063 PH/48 When a host name was looked up from an IP address, and the subsequent
2064 forward lookup of the name timed out, the host name was left in
2065 $sender_host_name, contrary to the specification.
2067 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2068 restricted to single-key lookups, Exim was not diagnosing an error if
2069 * or *@ was used with a query-style lookup.
2071 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2073 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2074 long ago, but noone interested enough thought of it. Let's just say that
2075 the "1.1" means that there are some new functions that weren't there at
2076 some point in the past.
2078 PH/51 Error processing for expansion failure of helo_data from an smtp
2079 transport during callout processing was broken.
2081 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2082 tested/used via the -bh/-bhc/-bs options.
2084 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2085 bug, fixed in subsequent PCRE releases).
2087 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2088 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2090 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2096 SC/01 Use a glob alias rather than an array ref in eximstats generated
2097 parser. This improves both readability and performance.
2099 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2100 Don't display local sender or destination tables in eximstats unless
2101 there is data to show.
2102 Added average volumes into the eximstats top table text output.
2104 SC/03 Collect data on the number of addresses (recipients) as well
2105 as the number of messages in eximstats.
2107 TF/01 Correct an error in the documentation for the redirect router. Exim
2108 does (usually) call initgroups() when daemonizing.
2110 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2111 with consistent privilege compared to when running as a daemon.
2113 TF/03 Note in the spec that $authenticated_id is not set for local
2114 submissions from trusted users.
2116 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2117 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2119 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2120 by adding some example configuration directives to the default
2121 configuration file. A little bit of work is required to uncomment the
2122 directives and define how usernames and passwords are checked, but
2123 there is now a framework to start from.
2125 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2126 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2127 without this. I don't know how relevant this is to other LDAP libraries.
2129 PH/02 Add the verb name to the "unknown ACL verb" error.
2131 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2133 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2135 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2136 directory not expanded when it should be if an expanded home directory
2137 was set for the address (which is overridden by the transport).
2139 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2142 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2143 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2144 because it is too late at that time, and has no effect.
2146 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2147 security issue with \' (bugzilla #107). I could not use the
2148 PQescapeStringConn() function, because it needs a PGconn value as one of
2151 PH/08 When testing addresses using -bt, indicate those final addresses that
2152 are duplicates that would not cause an additional delivery. At least one
2153 person was confused, thinking that -bt output corresponded to deliveries.
2154 (Suppressing duplicates isn't a good idea as you lose the information
2155 about possibly different redirections that led to the duplicates.)
2157 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2158 systems where poll() doesn't work, in particular OS X.
2160 PH/10 Added more information to debugging output for retry time not reached.
2162 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2163 operations in malware.c.
2165 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2168 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2169 syslog_duplication set false, log lines that would normally be written
2170 both the the main log and to the reject log were not written to syslog at
2173 PH/14 In the default configuration, change the use of "message" in ACL warn
2174 statements to "add_header".
2176 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2177 not followed by a command (e.g. "seen endif").
2179 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2180 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2183 PH/17 Added extra conditions to the default value of delay_warning_condition
2187 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2188 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2189 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2192 The Auto-Submitted: and various List- headers are standardised, whereas I
2193 don't think Precedence: ever was.
2195 PH/18 Refactored debugging code in route_finduser() to show more information,
2196 in particular, the error code if getpwnam() issues one.
2198 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2199 This is apparently needed in addition to the PH/07 change above to avoid
2200 any possible encoding problems.
2202 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2203 but not after initializing Perl.
2205 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2206 output them only if debugging. By default they are written stderr,
2207 apparently, which is not desirable.
2209 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2212 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2215 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2217 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2218 authenticated or an ident call has been made. Suppress the default
2219 values for $authenticated_id and $authenticated_sender (but permit -oMai
2220 and -oMas) when testing with -bh.
2222 PH/24 Re-jigged the order of the tests in the default configuration so that the
2223 tests for valid domains and recipients precede the DNS black list and CSA
2224 tests, on the grounds that those ones are more expensive.
2226 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2227 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2228 This bug exists in every version of Exim that I still have, right back to
2231 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2232 However, an attempt to turn on case-sensitivity in a regex key by
2233 including (?-i) didn't work because the subject string was already
2234 lowercased, and the effects were non-intuitive. It turns out that a
2235 one-line patch can be used to allow (?-i) to work as expected.
2241 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2242 other effects) broke the use of negated acl sub-conditions.
2244 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2247 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2248 "Deny" causes Exim to reject the incoming connection with a 554 error.
2249 Unfortunately, if there is a major crisis, such as a disk failure,
2250 tcp-wrappers gives "deny", whereas what one would like would be some
2251 kind of temporary error. A kludge has been added to help with this.
2252 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2253 554 error is used if errno is still zero or contains ENOENT (which occurs
2254 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2257 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2259 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2260 errors. Otherwise a message that provokes a temporary error (when other
2261 messages do not) can cause a whole host to time out.
2263 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2264 addresses were routed directly to files or pipes from a redirect router.
2265 File deliveries just didn't batch; pipe deliveries might have suffered
2268 PH/06 A failure to get a lock for a hints database would erroneously always say
2269 "Failed to get write lock", even when it was really a read lock.
2271 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2272 of 0600. This has been changed to use the value of the lockfile_mode
2273 option (which defaults to 0600).
2275 PH/08 Applied small patch from the Sieve maintainer.
2277 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2278 folder from quota calculations, a direct delivery into this folder messed
2279 up the contents of the maildirsize file. This was because the regex was
2280 used only to exclude .Trash (or whatever) when the size of the mailbox
2281 was calculated. There was no check that a delivery was happening into an
2282 excluded directory. This bug has been fixed by ignoring all quota
2283 processing for deliveries into excluded directories.
2285 PH/10 Added the maildirfolder_create_regex option to appendfile.
2291 PH/01 The code for finding all the local interface addresses on a FreeBSD
2292 system running IPv6 was broken. This may well have applied to all BSD
2293 systems, as well as to others that have similar system calls. The broken
2294 code found IPv4 interfaces correctly, but gave incorrect values for the
2295 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2296 that it would not match correctly against @[] and not recognize the IPv6
2299 PH/02 The ipliteral router was not recognizing addresses of the form user@
2300 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2302 PH/03 Added disable_ipv6.
2304 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2305 decoded form, because it is most often used to construct To: headers
2306 lines in autoreplies, and the decoded form may well be syntactically
2307 invalid. However, $reply_address has leading white space removed, and all
2308 newlines turned into spaces so that the autoreply transport does not
2311 PH/05 If group was specified without a user on a router, and no group or user
2312 was specified on a transport, the group from the router was ignored.
2314 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2315 for visible compile-time settings that can be used to change these
2316 numbers, for those that want even more. Backwards compatibility with old
2317 spool files has been maintained. However, going back to a previous Exim
2318 release will lost any variables that are in spool files.
2320 PH/07 Two small changes when running in the test harness: increase delay when
2321 passing a TCP/IP connection to a new process, in case the original
2322 process has to generate a bounce, and remove special handling of
2323 127.0.0.2 (sic), which is no longer necessary.
2325 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2326 be the same on different OS.
2328 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2331 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2332 whether --show-vars was specified or not
2334 JJ/02 exipick: Added support for new ACL variable spool format introduced
2337 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2338 syntactically invalid From: or Reply-to: line, and a filter used this to
2339 generate an autoreply, and therefore failed to obtain an address for the
2340 autoreply, Exim could try to deliver to a non-existent relative file
2341 name, causing unrelated and misleading errors. What now happens is that
2342 it logs this as a hard delivery error, but does not attempt to create a
2345 PH/11 The exinext utility has a -C option for testing purposes, but although
2346 the given file was scanned by exinext itself; it wasn't being passed on
2347 when Exim was called.
2349 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2350 an end-of-file indication when reading a command response.
2352 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2353 compiled. In many other places in Exim, IPv6 addresses are always
2354 recognized, so I have changed this. It also means that IPv4 domain
2355 literals of the form [IPV4:n.n.n.n] are now always recognized.
2357 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2358 used if the router is not running as root, for example, when verifying at
2359 ACL time, or when using -bh. The debugging output from this situation was
2360 non-existent - all you got was a failure to exec. I have made two
2363 (a) Failures to set uid/gid, the current directory, or a process leader
2364 in a subprocess such as that created by queryprogram now generate
2365 suitable debugging ouput when -d is set.
2367 (b) The queryprogram router detects when it is not running as root,
2368 outputs suitable debugging information if -d is set, and then runs
2369 the subprocess without attempting to change uid/gid.
2371 PH/15 Minor change to Makefile for building test_host (undocumented testing
2374 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2375 additional section of a DNS packet that returns MX or SRV records.
2376 Instead, it always explicitly searches for A/AAAA records. This avoids
2377 major problems that occur when a DNS server includes only records of one
2378 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2379 fixed another bug: if SRV records were looked up and the corresponding
2380 address records were *not* found in the additional section, the port
2381 values from the SRV records were lost.
2383 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2384 using the correct key (the original address) when searching the retry
2385 rules in order to find which one to use for generating the retry hint.
2387 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2388 adding the default one. Similarly, if it contains a Reply-To: header, the
2389 errors_reply_to option, if set, is not used.
2391 PH/19 When calculating a retry time, Exim used to measure the "time since
2392 failure" by looking at the "first failed" field in the retry record. Now
2393 it does not use this if it is later than than the arrival time of the
2394 message. Instead it uses the arrival time. This makes for better
2395 behaviour in cases where some deliveries succeed, thus re-setting the
2396 "first failed" field. An example is a quota failure for a huge message
2397 when small messages continue to be delivered. Without this change, the
2398 "time since failure" will always be short, possible causing more frequent
2399 delivery attempts for the huge message than are intended.
2400 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2402 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2403 $1, $2, $3) because the numerical variables can be reset during some
2404 expansion items (e.g. "match"), thereby losing the authentication data.
2406 PH/21 Make -bV show the size of off_t variables so that the test suite can
2407 decide whether to run tests for quotas > 2G.
2409 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2410 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2411 filecount value is greater than 2G or if a quota value is greater than 2G
2412 on a system where the size of off_t is not greater than 4, a panic error
2415 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2416 never match. The debug and -bh output now contains an explicit error
2417 message indicating a malformed IPv4 address or mask.
2419 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2420 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2421 PH/23 above applies.
2423 PH/25 Do not write to syslog when running in the test harness. The only
2424 occasion when this arises is a failure to open the main or panic logs
2425 (for which there is an explicit test).
2427 PH/26 Added the /no_tell option to "control=freeze".
2429 PH/27 If a host name lookup failed very early in a connection, for example, if
2430 the IP address matched host_lookup and the reverse lookup yielded a name
2431 that did not have a forward lookup, an error message of the form "no IP
2432 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2433 could be logged. Now it outputs the IP address instead of "NULL".
2435 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2436 allows the sender and the authenticated sender to be set when
2437 submitting a message from within Exim. Since child_open_exim() is
2438 documented for local_scan(), the new function should be too.
2440 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2441 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2442 results in an empty string is now treated as unset.
2444 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2446 PH/31 Added sender_verify_fail logging option.
2448 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2449 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2450 tidied the source and removed it altogether.
2452 PH/33 When a queue run was abandoned because the load average was too high, a
2453 log line was always written; now it is written only if the queue_run log
2454 selector is set. In addition, the log line for abandonment now contains
2455 information about the queue run such as the pid. This is always present
2456 in "start" and "stop" lines but was omitted from the "abandon" line.
2458 PH/34 Omit spaces between a header name and the colon in the error message that
2459 is given when verify = headers_syntax fails (if there are lots of them,
2460 the message gets confusing).
2462 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2463 names, as there are now some PTR records that contain slashes. This check
2464 is only to protect against broken name servers that fall over on strange
2465 characters, so the fact that it applies to all lookups doesn't matter.
2467 PH/36 Now that the new test suite is complete, we can remove some of the
2468 special code in Exim that was needed for the old test suite. For example,
2469 sorting DNS records because real resolvers return them in an arbitrary
2470 order. The new test suite's fake resolver always returns records in the
2473 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2474 bounces) except when queue_only is set, to avoid logging races between
2475 the different processes.
2477 PH/38 Panic-die if .include specifies a non-absolute path.
2479 PH/39 A tweak to the "H" retry rule from its user.
2481 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2482 a label. They prevented compilation on older perls.
2484 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2485 a warning to be raised on newish perls.
2487 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2488 on queue. Changes to match documented behaviour of showing count of
2489 messages matching specified criteria.
2491 PH/40 Changed the default ident timeout from 30s to 5s.
2493 PH/41 Added support for the use of login_cap features, on those BSD systems
2494 that have them, for controlling the resources used by pipe deliveries.
2496 PH/42 The content-scanning code uses fopen() to create files in which to put
2497 message data. Previously it was not paying any attention to the mode of
2498 the files. Exim runs with umask(0) because the rest of the code creates
2499 files with open(), and sets the required mode explicitly. Thus, these
2500 files were ending up world-writeable. This was not a big issue, because,
2501 being within the spool directory, they were not world-accessible. I have
2502 created a function called modefopen, which takes an additional mode
2503 argument. It sets umask(777), creates the file, chmods it to the required
2504 mode, then resets the umask. All the relevant calls to fopen() in the
2505 content scanning code have been changed to use this function.
2507 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2508 to 24 hours. This avoids potential overflow problems when processing G
2509 and H retry rules. I suspect nobody ever tinkers with this value.
2511 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2513 PH/45 When the plaintext authenticator is running as a client, the server's
2514 challenges are checked to ensure they are valid base64 strings. By
2515 default, the authentication attempt is cancelled if an invalid string is
2516 received. Setting client_ignore_invalid_base64 true ignores these errors.
2517 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2518 they are received. Thus, the responses can be made to depend on the
2519 challenges. If an invalid string is ignored, an empty string is placed in
2522 PH/46 Messages that are created by the autoreply transport now contains a
2523 References: header, in accordance with RFCs 2822 and 3834.
2525 PH/47 Added authenticated_sender_force to the smtp transport.
2527 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2529 PH/49 Installed latest patch from the Sieve maintainer.
2531 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2532 was also set, the appendfile transport was unnecessarily scanning a
2533 directory of message files (e.g. for maildir delivery) to find the count
2534 of files (along with the size), even though it did not need this
2535 information. It now does the scan only if it needs to find either the
2536 size of the count of files.
2538 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2540 PH/52 Two bugs concerned with error handling when the smtp transport is
2543 (i) Exim was not creating retry information for temporary errors given
2544 for individual recipients after the DATA command when the smtp transport
2545 was used in LMTP mode. This meant that they could be retried too
2546 frequently, and not timed out correctly.
2548 (ii) Exim was setting the flag that allows error details to be returned
2549 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2550 recipients that were returned after the DATA command.
2552 PH/53 This is related to PH/52, but is more general: for any failing address,
2553 when detailed error information was permitted to be returned to the
2554 sender, but the error was temporary, then after the final timeout, only
2555 "retry timeout exceeded" was returned. Now it returns the full error as
2556 well as "retry timeout exceeded".
2558 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2559 do this, and (what is worse) MTAs that accept it.
2561 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2562 will now be deprecated.
2564 PH/56 New os.c-cygwin from the Cygwin maintainer.
2566 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2567 formats (previously only available in exim formats via -bpr, -bpru,
2568 and -bpra. Now also available in native and exiqgrep formats)
2570 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2571 with very large, slow to parse queues
2573 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2575 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2577 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2578 responses to authentication challenges, though it was showing the
2579 challenges; (ii) I've removed the CR characters from the debug output for
2582 PH/58 Allow for the insertion of a newline as well as a space when a string
2583 is turned into more than one encoded-word during RFC 2047 encoding. The
2584 Sieve code now uses this.
2586 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2587 data_4xx, lost_connection, tls_required.
2589 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2590 message was being sent as an SMTP response.
2592 PH/61 Add -l and -k options to exicyclog.
2594 PH/62 When verifying, if an address was redirected to one new address, so that
2595 verification continued, and the new address failed or deferred after
2596 having set something in $address_data, the value of $address_data was not
2597 passed back to the ACL. This was different to the case when no
2598 redirection occurred. The value is now passed back in both cases.
2600 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2601 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2602 use login_cap.h, so on its own it isn't the distinguishing feature. The
2603 new name refers directly to the setclassresources() function.
2605 PH/65 Added configuration files for NetBSD3.
2607 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2609 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2610 is preferred over IPv4.
2612 PH/68 The bounce_return_message and bounce_return_body options were not being
2613 honoured for bounces generated during the reception of non-SMTP messages.
2614 In particular, this applied to messages rejected by the ACL. This bug has
2615 been fixed. However, if bounce_return_message is true and bounce_return_
2616 body is false, the headers that are returned for a non-SMTP message
2617 include only those that have been read before the error was detected.
2618 (In the case of an ACL rejection, they have all been read.)
2620 PH/69 The HTML version of the specification is now built in a directory called
2621 spec_html instead of spec.html, because the latter looks like a path with
2622 a MIME-type, and this confuses some software.
2624 PH/70 Catch two compiler warnings in sieve.c.
2626 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2627 function verify_get_ident() calls ip_connect() to connect a socket, but
2628 if the "connect()" function timed out, ip_connect() used to close the
2629 socket. However, verify_get_ident() also closes the socket later, and in
2630 between Exim writes to the log, which may get opened at this point. When
2631 the socket was closed in ip_connect(), the log could get the same file
2632 descriptor number as the socket. This naturally causes chaos. The fix is
2633 not to close the socket in ip_connect(); the socket should be closed by
2634 the function that creates it. There was only one place in the code where
2635 this was missing, in the iplookup router, which I don't think anybody now
2636 uses, but I've fixed it anyway.
2638 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2639 well as to direct DNS lookups. Otherwise the handling of names in host
2640 lists is inconsistent and therefore confusing.
2646 PH/01 Two changes to the default runtime configuration:
2648 (1) Move the checks for relay_from_hosts and authenticated clients from
2649 after to before the (commented out) DNS black list checks.
2651 (2) Add control=submission to the relay_from_hosts and authenticated
2652 clients checks, on the grounds that messages accepted by these
2653 statements are most likely to be submissions.
2655 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2657 (1) Generate an error if the third argument for the ${prvs expansion is
2660 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2663 (3) Reset the variables that are obtained from the first argument of
2664 ${prvscheck and used in the second argument before leaving the code,
2665 because their memory is reclaimed, so using them afterwards may do
2668 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2669 one (it's much easier than Tom thought :-).
2671 (5) Because of (4), we can now allow for the use of $prvscheck_result
2672 inside the third argument.
2674 PH/03 For some reason, the default setting of PATH when running a command from
2675 a pipe transport was just "/usr/bin". I have changed it to
2678 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2679 anything to be listed in the output from -bV.
2681 PH/05 When a filter generated an autoreply, the entire To: header line was
2682 quoted in the delivery log line, like this:
2684 => >A.N.Other <ano@some.domain> <original@ddress> ...
2686 This has been changed so that it extracts the operative address. There
2687 may be more than one such address. If so, they are comma-separated, like
2690 => >ano@some.domain,ona@other.domain <original@ddress> ...
2692 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2693 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2694 not being looked up in the rDNS to get a host name, Exim was showing the
2695 IP address twice in Received: lines, even though the IP addresses were
2696 identical. For example:
2698 Received: from [1.2.3.4] (helo=[1.2.3.4])
2700 However, if the real host name was known, it was omitting the HELO data
2701 if it matched the actual IP address. This has been tidied up so that it
2702 doesn't show the same IP address twice.
2704 PH/07 When both +timestamp and +memory debugging was on, the value given by
2705 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2706 called by the debug printing, thereby overwriting the timestamp buffer.
2707 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2709 PH/08 When the original message was included in an autoreply transport, it
2710 always said "this is a copy of the message, including all the headers",
2711 even if body_only or headers_only was set. It now gives an appropriate
2714 PH/09 Applied a patch from the Sieve maintainer which:
2716 o fixes some comments
2717 o adds the (disabled) notify extension core
2718 o adds some debug output for the result of if/elsif tests
2719 o points to the current vacation draft in the documentation
2720 and documents the missing references header update
2724 o fixes a bug in processing the envelope test (when testing
2725 multiple envelope elements, the last element determinted the
2728 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2729 Electronic Mail") by including:
2731 Auto-submitted: auto-generated
2733 in the messages that it generates (bounce messages and others, such as
2734 warnings). In the case of bounce messages for non-SMTP mesages, there was
2735 also a typo: it was using "Auto_submitted" (underscore instead of
2736 hyphen). Since every message generated by Exim is necessarily in response
2737 to another message, thes have all been changed to:
2739 Auto-Submitted: auto-replied
2741 in accordance with these statements in the RFC:
2743 The auto-replied keyword:
2745 - SHOULD be used on messages sent in direct response to another
2746 message by an automatic process,
2748 - MUST NOT be used on manually-generated messages,
2750 - MAY be used on Delivery Status Notifications (DSNs) and Message
2751 Disposition Notifications (MDNs),
2753 - MUST NOT be used on messages generated by automatic or periodic
2754 processes, except for messages which are automatic responses to
2757 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2758 to the default Received: header definition.
2760 PH/12 Added log selector acl_warn_skipped (default on).
2762 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2763 variables because (a) they are in the wrong storage pool and (b) even if
2764 they were copied, it wouldn't work properly because of the caching.
2766 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2767 checking when decoding. Apparently there are clients that generate
2768 overlong encoded strings. Why am I not surprised?
2770 PH/15 If the first argument of "${if match_address" was not empty, but did not
2771 contain an "@" character, Exim crashed. Now it writes a panic log message
2772 and treats the condition as false.
2774 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2776 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2777 extension "envelope-auth". The code is finished and in agreement with
2778 other implementations, but there is no documentation so far and in fact,
2779 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2780 not changing the active code.
2782 Print executed "if" and "elsif" statements when debugging is used. This
2783 helps a great deal to understand what a filter does.
2785 Document more things not specified clearly in RFC3028. I had all this
2786 sorted out, when out of a sudden new issues came to my mind. Oops."
2788 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2791 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2792 canonical form (as documented). However, after a host name lookup from
2793 the IP address, check_host() was doing a simple string comparison with
2794 addresses acquired from the DNS when checking that the found name did
2795 have the original IP as one of its addresses. Since any found IPv6
2796 addresses are likely to be in abbreviated form, the comparison could
2797 fail. Luckily, there already exists a function for doing the comparison
2798 by converting both addresses to binary, so now that is used instead of
2799 the text comparison.
2801 PH/20 There was another similar case to PH/19, when a complete host name was
2802 given in a host list; looking up its IP address could give an abbreviated
2803 form, whereas the current host's name might or might not be abbreviated.
2804 The same fix has been applied.
2810 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2811 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2814 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2815 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2817 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2819 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2820 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2821 Also, for "Auto-Submitted", treat anything other than "no" as
2822 non-personal, in accordance with RFC 3834. (Previously it treated
2823 anything starting "auto-" as non-personal.)
2825 TF/01 The control=submission/name=... option had a problem with syntax
2826 errors if the name included a slash character. The /name= option
2827 now slurps the rest of the string, so it can include any characters
2828 but it must come last in the list of options (after /sender_retain
2831 PH/05 Some modifications to the interface to the fake nameserver for the new
2839 TK/01 Added the "success_on_redirect" address verification option. See
2840 NewStuff for rationale and an example.
2842 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2844 PH/02 Patch to exigrep to allow it to work on syslog lines.
2846 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2847 fread() to skip over the body file's header line, because in Cygwin the
2848 header line is locked and is inaccessible.
2850 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2851 co-exist for some time) to make it clear that it is the Exim ID that is
2852 referenced, not the Message-ID: header line.
2854 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2855 string_format() function, because snprintf() does not exist on all
2858 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2859 ${stat: expansion item.
2861 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2862 protocol synchronization error", to keep the pedants happy.
2864 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2865 well as for IRIX systems, when gcc is being used. See the host.c source
2868 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2870 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2873 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2874 transport and to the smtp transport in LMTP mode.
2876 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2878 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2879 run a filter in a subprocess. This could lead to confusion in subsequent
2880 lookups in the parent process. There should also be a search_tidyup() at
2881 the end of the subprocess.
2883 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2884 only if the host matched helo_try_verify_hosts, which caused the
2885 verification to occur when the EHLO/HELO command was issued. The ACL just
2886 tested the remembered result. Now, if a previous verification attempt has
2887 not happened, "verify = helo" does it there and then.
2889 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2891 TK/03 Fix log output including CR from clamd.
2893 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2894 exist provoked a memory error which could cause a segfault.
2896 PH/15 Installed PCRE 6.2
2898 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2900 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2901 of the problem. Specifically, suggested +O2 rather than +O1 for the
2904 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2906 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2907 record, the port was not correctly logged when the outgoing_port log
2908 selector was set (it logged the transort's default port).
2910 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2911 fallback_hosts, and "hosts" in the smtp transport.
2913 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2914 host errors such as "Connection refused".
2916 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2917 authentication with radiusclient 0.4.9:
2919 - Error returned from rc_read_config was caught wrongly
2920 - Username/password not passed on to radius server due to wrong length.
2922 The presumption is that some radiusclient API changes for 4.51/PH/17
2923 were not taken care of correctly. The code is still untested by me (my
2924 Linux distribution still has 0.3.2 of radiusclient), but it was
2925 contributed by a Radius user.
2927 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2928 expanding the "port" option of the smtp transport.
2930 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2931 while reading a MIME header. Thanks to Tom Hughes for a patch.
2933 PH/24 Include config.h inside local_scan.h so that configuration settings are
2936 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2937 that in an ACL for RCPT (for example), you can examine exactly what was
2940 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2941 commands, but it was not correctly comparing the address with the actual
2942 client host address. Thus, it would show the EHLO address in Received:
2943 header lines when this was not necessary.
2945 PH/27 Added the % operator to ${eval:}.
2947 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2948 it should be ignoring failures (because with -C, for example, it has lost
2949 privilege). It wasn't ignoring creation failures other than "already
2952 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2955 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2956 because an input line was too long, either on its own, or by virtue of
2957 too many continuations, the temporary file was not being removed, and the
2958 return code was incorrect.
2960 PH/31 Missing "BOOL" in function definition in filtertest.c.
2962 PH/32 Applied Sieve patches from the maintainer.
2964 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2966 PH/33 Added "verify = not_blind".
2968 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2969 Local/Makefile (with some defaults set). These are used in built scripts
2970 such as exicyclog, but they have never been used in the exim_install
2971 script (though there are many overriding facilities there). I have
2972 arranged that the exim_install script now takes note of these two
2975 PH/35 Installed configuration files for Dragonfly.
2977 PH/36 When a locally submitted message by a trusted user did not contain a
2978 From: header, and the sender address was obtained from -f or from an SMTP
2979 MAIL command, and the trusted user did not use -F to supply a sender
2980 name, $originator_name was incorrectly used when constructing a From:
2981 header. Furthermore, $originator_name was used for submission mode
2982 messages from external hosts without From: headers in a similar way,
2983 which is clearly wrong.
2985 PH/37 Added control=suppress_local_fixups.
2987 PH/38 When log_selector = +received_sender was set, and the addition of the
2988 sender made the log line's construction buffer exactly full, or one byte
2989 less than full, an overflow happened when the terminating "\n" was
2992 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2993 when the result of a list match is failure because a DNS lookup failed.
2995 PH/40 RM_COMMAND is now used in the building process.
2997 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2998 the "build-* directories that it finds.
3000 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3001 address in a domain literal was a prefix of an interface address.
3003 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3004 when verifying a sender address, unless rewrite_headers is false.
3006 PH/44 Wrote a long comment about why errors_to addresses are verified as
3007 recipients, not senders.
3009 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3010 the ratelimit ACL was added.
3012 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3014 PH/46 Added extra information about PostgreSQL errors to the error string.
3016 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3017 avoiding the need to install special zones in a real server. This is
3018 backwards compatible; if it can't find the fake resolver, it drops back.
3019 Thus, both old and new test suites can be run.
3021 TF/02 Added util/ratelimit.pl
3023 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3024 clock is set back in time.
3026 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3027 Candler <B.Candler@pobox.com>.
3029 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3030 OK for addresses that are the result of redirections.
3032 PH/48 A number of further additions for the benefit of the new test suite,
3033 including a fake gethostbyname() that interfaces to the fake DNS resolver
3036 TF/06 The fix for widen_domains has also been applied to qualify_single and
3037 search_parents which are the other dnslookup options that can cause
3040 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3043 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3045 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3046 service" if it encountered an MX record with an empty target hostname.
3047 The message is now "an MX or SRV record indicated no SMTP service".
3049 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3050 if the DNS of the sending site is misconfigured. This is quite a
3051 common situation. This change restores the behaviour of treating a
3052 helo verification defer as a failure.
3054 PH/51 If self=fail was set on a router, the bounce message did not include the
3055 actual error message.
3061 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3063 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3064 command itself ended in error, the underlying message about the transport
3065 filter timeout was being overwritten with the pipe command error. Now the
3066 underlying error message should be appended to the second error message.
3068 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3070 PH/02 Reduce the amount of output that "make" produces by default. Full output
3071 can still be requested.
3073 PH/03 The warning log line about a condition test deferring for a "warn" verb
3074 was being output only once per connection, rather than after each
3075 occurrence (because it was using the same function as for successful
3076 "warn" verbs). This seems wrong, so I have changed it.
3078 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3079 it should not have, which might have caused a crash in the right
3080 circumstances, but probably never did.
3082 PH/04 Installed a modified version of Tony Finch's patch to make submission
3083 mode fix the return path as well as the Sender: header line, and to
3084 add a /name= option so that you can make the user's friendly name appear
3087 TF/03 Added the control = fakedefer ACL modifier.
3089 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3090 Mark Lowes for thorough testing.
3092 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3094 TK/03 Merged latest SRS patch from Miles Wilton.
3096 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3097 with the definition in sysexits.h (which is #included earlier).
3098 Fortunately, Exim does not actually use EX_OK. The code used to try to
3099 preserve the sysexits.h value, by assumimg that macro definitions were
3100 scanned for macro replacements. I have been disabused of this notion,
3101 so now the code just undefines EX_OK before #including unistd.h.
3103 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3104 in the smtp transport. When a block could not be written in a single
3105 write() function, the timeout was being re-applied to each part-write.
3106 This seems wrong - if the receiver was accepting one byte at a time it
3107 would take for ever. The timeout is now adjusted when this happens. It
3108 doesn't have to be particularly precise.
3110 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3111 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3113 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3114 without a domain if the check to <postmaster@domain> fails.
3116 SC/01 Eximstats: added -xls and the ability to specify output files
3117 (patch written by Frank Heydlauf).
3119 SC/02 Eximstats: use FileHandles for outputing results.
3121 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3123 SC/04 Eximstats: fixed display of large numbers with -nvr option
3125 SC/05 Eximstats: fixed merging of reports with empty tables.
3127 SC/06 Eximstats: added the -include_original_destination flag
3129 SC/07 Eximstats: removed tabs and trailing whitespace.
3131 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3133 TK/06 MBOX spool code: Add real "From " MBOX separator line
3134 so the .eml file is really in mbox format (even though
3135 most programs do not really care). Patch from Alex Miller.
3137 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3138 The latter is generated from $received_to and is only set if the
3139 message has one envelope recipient. SA can use these headers,
3140 obviously out-of-the-box. Patch from Alex Miller.
3142 PH/08 The ${def test on a variable was returning false if the variable's
3143 value was "0", contrary to what the specification has always said!
3144 The result should be true unless the variable is empty.
3146 PH/09 The syntax error of a character other than { following "${if
3147 def:variable_name" (after optional whitespace) was not being diagnosed.
3148 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3149 accidental colon was present, for example, could give incorrect results.
3151 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3152 result is used (not including appendfile, where other changes are about
3155 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3156 This involved changing a lot of size variables from int to off_t. It
3157 should work with maildirs and everything.
3159 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3160 spamd dying while we are connected to it.
3162 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3165 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3166 function for BDB 4.3.
3168 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3170 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3171 the sources for PCRE have been re-arranged and more files are now
3174 PH/15 The code I had for printing potentially long long variables in PH/11
3175 above was not the best (it lost precision). The length of off_t variables
3176 is now inspected at build time, and an appropriate printing format (%ld
3177 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3178 to be "long long int" or "long int". This is needed for the internal
3179 formatting function string_vformat().
3181 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3182 the configuration file to be ":syslog", then the script "guesses" where
3183 the logs files are, rather than using the compiled in default. In our
3184 case the guess is not the same as the compiled default, so the script
3185 suddenly stopped working when I started to use syslog. The patch checks
3186 to see if log_file_path is "". If so, it attempts to read it from exim
3187 with no configuration file to get the compiled in version, before it
3188 falls back to the previous guessing code."
3190 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3191 implementing BATV in an Exim configuration. See NewStuff for the gory
3194 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3195 Makefile that are specific to HP-UX.
3197 PH/18 If the "use_postmaster" option was set for a recipient callout together
3198 with the "random" option, the postmaster address was used as the MAIL
3199 FROM address for the random test, but not for the subsequent recipient
3200 test. It is now used for both.
3202 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3203 patch removes a few documentation additions to RFC 3028, because the
3204 latest draft now contains them. It adds the new en;ascii-case comparator
3205 and a new error check for 8bit text in MIME parts. Comparator and
3206 require names are now matched exactly. I enabled the subaddress
3207 extension, but it is not well tested yet (read: it works for me)."
3209 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3210 rework some of the code of TK/09 above to avoid the hardwired use of
3211 "%lld" and "long long". Replaced the call to snprintf() with a call to
3214 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3215 records point to non-existent hosts", "retry timeout exceeded", and
3216 "retry time not reached for any host after a long failure period".
3218 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3219 experimental DomainKeys support:
3221 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3222 (2) On an error such as an illegally used "control", the wrong name for
3223 the control was given.
3225 These problems did NOT occur unless DomainKeys support was compiled.
3227 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3229 PH/24 Added ${if match_ip condition.
3231 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3232 databases so that it will be absolutely obvious if a crash occurs in the
3233 DB library. This is a regular occurrence (often caused by mis-matched
3236 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3237 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3238 picked up on a user's system that detects such things. There doesn't seem
3239 to be a gcc warning option for this - only an attribute that has to be
3240 put on the function's prototype. It seems that in Fedora Core 4 they have
3241 set this on a number of new functions. No doubt there will be more in due
3244 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3245 specify a transport. However, if an address that was verified by such a
3246 router was the subject of a callout, Exim crashed because it tried to
3247 read the rcpt_include_affixes from the non-existent transport. Now it
3248 just assumes that the setting of that option is false. This bug was
3249 introduced by 4.51/PH/31.
3251 PH/28 Changed -d+all to exclude +memory, because that information is very
3252 rarely of interest, but it makes the output a lot bigger. People tend to
3253 do -d+all out of habit.
3255 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3256 code in os-type was giving problems when libc.so lives in lib64, like on
3259 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3260 aren't the modern standard, and it seems that some systems' include files
3261 don't always have them. Exim was already checking for some of the newer
3262 ones like T_AAAA, and defining it itself. I've added checks for all the
3263 record types that Exim uses.
3265 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3266 not automatically generating a new one, as it is supposed to. This
3267 prevented TLS from working. If the file did exist, but contained invalid
3268 data, a new version was generated, as expected. It was only the case of a
3269 non-existent file that was broken.
3271 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3272 with a change in libdomainkeys > 0.64.
3274 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3275 from DNS. If the selector record carries the flag, it now has
3276 precedence over the domain-wide flag.
3278 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3280 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3281 the use of a port name that isn't defined in /etc/services) occurred, the
3282 message was deferred as in a normal delivery, and thus remained on the
3283 spool, instead of being failed because of the mua_wrapper setting. This
3284 is now fixed, and I tidied up some of the mua_wrapper messages at the
3287 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3288 the messages in a hash of arrays rather than using individual hashes.
3289 This is a bit cleaner and results in dramatic memory savings, albeit
3290 at a slight CPU cost.
3292 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3293 as requested by Marc Sherman.
3295 SC/10 Eximstats: added histograms for user specified patterns as requested
3298 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3300 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3301 fopen() in the content-scanning modules that did not already have it.
3307 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3308 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3310 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3312 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3314 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3315 to test Sieve filters that use "vacation".
3317 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3318 that changes the way the GnuTLS parameters are stored in the cache file.
3319 The new format can be generated externally. For backward compatibility,
3320 if the data in the cache doesn't make sense, Exim assumes it has read an
3321 old-format file, and it generates new data and writes a new file. This
3322 means that you can't go back to an older release without removing the
3325 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3326 work if there are any delivery delays because "one_time" forces the
3327 parent to be marked "delivered", so its unseen clone is never tried
3328 again. For this reason, Exim now forbids the simultaneous setting of
3331 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3332 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3333 entry if you want to know the details. The fix, however, neglected to
3334 consider the case when local delivery batching is involved. The test for
3335 "previously delivered" was not happening when checking to see if an
3336 address could be batched with a previous (undelivered) one; under
3337 certain circumstances this could lead to multiple deliveries to the same
3340 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3341 in its include files, and this causes problems building Exim.
3343 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3344 header_syntax) but Exim was just ignoring anything given after a slash.
3345 In particular, this caused confusion with an attempt to use "verify =
3346 reverse_host_lookup/defer_ok". An error is now given when options are
3347 supplied for verify items that do not have them. (Maybe reverse_host_
3348 lookup should have a defer_ok option, but that's a different point.)
3350 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3351 defined by RFC 821) to 2048, because there were problems with some AUTH
3352 commands, and RFC 1869 says the size should be increased for extended
3353 SMTP commands that take arguments.
3355 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3358 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3359 "unknown" error; now it says that the functionality isn't in the binary.
3361 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3362 an address' error message when a string expansion fails (syntax or
3363 whatever). Otherwise the password may appear in the log. Following change
3364 PH/42 below, there is no longer a chance of it appearing in a bounce
3367 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3369 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3370 of its addresses was ever tried. (Bugzilla bug #2.)
3372 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3373 the result incorrectly in the debug output. (It correctly added a newline
3374 to what was transported.)
3376 TF/01 Added $received_time.
3378 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3379 commented out examples of how to interface to a virus scanner and to
3380 SpamAssassin. Also added commented examples of av_scanner and
3381 spamd_address settings.
3383 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3384 and controls are allowed in which ACLs. There were a couple of minor
3385 errors. Some of the entries in the conditions table (which is a table of
3386 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3387 negation of where the condition IS allowed.
3389 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3391 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3392 header file does not have a version number, so I've had to invent a new
3393 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3394 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3395 radiusclient), but it was contributed by a Radius user.
3397 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3398 files or directories, for OpenSSL.
3400 PH/19 When an Exim process that is running as root has to create an Exim log
3401 file, it does so in a subprocess that runs as exim:exim so as to get the
3402 ownership right at creation (otherwise, other Exim processes might see
3403 the file with the wrong ownership). There was no test for failure of this
3404 fork() call, which would lead to the process getting stuck as it waited
3405 for a non-existent subprocess. Forks do occasionally fail when resources
3406 run out. I reviewed all the other calls to fork(); they all seem to check
3409 PH/20 When checking for unexpected SMTP input at connect time (before writing
3410 the banner), Exim was not dealing correctly with a non-positive return
3411 from the read() function. If the client had disconnected by this time,
3412 the result was a log entry for a synchronization error with an empty
3413 string after "input=" when read() returned zero. If read() returned -1
3414 (an event I could not check), uninitialized data bytes were printed.
3415 There were reports of junk text (parts of files, etc) appearing after
3418 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3420 PH/22 Added support for macro redefinition, and (re)definition in between
3421 driver and ACL definitions.
3423 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3424 forgetting to use the resulting value; it was using the unexpanded value.
3426 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3427 hadn't been configured. The fix is from Juergen Kreileder, who
3428 understands it better than I do:
3430 "Here's what I see happening with three configured cyrus_sasl
3431 authenticators configured (plain, login, cram-md5):
3433 On startup auth_cyrus_sasl_init() gets called for each of these.
3434 This means three calls to sasl_listmech() without a specified mech_list.
3435 => SASL tests which mechs of all available mechs actually work
3436 => three warnings about OTP not working
3437 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3439 With the patch, sasl_listmech() also gets called three times. But now
3440 SASL's mech_list option is set to the server_mech specified in the the
3441 authenticator. Or in other words, the answer from sasl_listmech()
3442 gets limited to just the mech you're testing for (which is different
3444 => the return list contains just 'plain' or 'login', 'cram-md5' or
3445 nothing depending on the value of ob->server_mech.
3447 I've just tested the patch: Authentication still works fine,
3448 unavailable mechs specified in the exim configuration are still
3449 caught, and the auth.log warnings about OTP are gone."
3451 PH/25 When debugging is enabled, the contents of the command line are added
3452 to the debugging output, even when log_selector=+arguments is not
3455 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3456 answer is "GNU", and only if the return is "GNU/something" is the answer
3459 PH/27 $acl_verify_message is now set immediately after the failure of a
3460 verification in an ACL, and so is available in subsequent modifiers. In
3461 particular, the message can be preserved by coding like this:
3463 warn !verify = sender
3464 set acl_m0 = $acl_verify_message
3466 Previously, $acl_verify_message was set only while expanding "message"
3467 and "log_message" when a very denied access.
3469 PH/28 Modified OS/os.c-Linux with
3471 -#ifndef OS_LOAD_AVERAGE
3472 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3474 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3475 nomenclature these days.)
3477 PH/29 Installed patch from the Sieve maintainer that adds the options
3478 sieve_useraddress and sieve_subaddress to the redirect router.
3480 PH/30 In these circumstances:
3481 . Two addresses routed to the same list of hosts;
3482 . First host does not offer TLS;
3483 . First host accepts first address;
3484 . First host gives temporary error to second address;
3485 . Second host offers TLS and a TLS session is established;
3486 . Second host accepts second address.
3487 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3488 and peerdn, if requested) that were in fact used only for the second
3491 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3492 attention to any local part prefix or suffix that was matched by the
3493 router that accepted the address. It now behaves in the same way as it
3494 does for delivery: the affixes are removed from the local part unless
3495 rcpt_include_affixes is set on the transport.
3497 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3498 timeout during the DATA phase of an incoming message.
3500 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3501 applied a patch sanctioned by the Sieve maintainer.
3503 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3504 the uid or gid is negative. A case of a negative gid caused this to be
3505 noticed. The fix allows for either to be negative.
3507 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3508 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3511 PH/36 The change PH/12 above was broken. Fixed it.
3513 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3514 the grounds that routing the same address twice would always produce the
3515 same answer. This might have been true once, but it is certainly no
3516 longer true now. Routing a child address may depend on the previous
3517 routing that produced that child. Some complicated redirection strategies
3518 went wrong when messages had multiple recipients, and made Exim's
3519 behaviour dependent on the order in which the addresses were given.
3521 I have moved the duplicate checking until after the routing is complete.
3522 Exim scans the addresses that are assigned to local and remote
3523 transports, and removes any duplicates. This means that more work will be
3524 done, as duplicates will always all be routed, but duplicates are
3525 presumably rare, so I don't expect this is of any significance.
3527 For deliveries to pipes, files, and autoreplies, the duplicate checking
3528 still happens during the routing process, since they are not going to be
3531 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3532 It corrects a timeout issue with spamd. This is Ian's comment: "The
3533 background is that sometimes spamd either never reads data from a
3534 connection it has accepted, or it never writes response data. The exiscan
3535 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3536 blindly assumes that writes won't block so it may never time out."
3538 PH/39 Allow G after quota size as well as K and M.
3540 PH/40 The value set for $authenticated_id in an authenticator may not contain
3541 binary zeroes or newlines because the value is written to log lines and
3542 to spool files. There was no check on this. Now the value is run through
3543 the string_printing() function so that such characters are converted to
3544 printable escape sequences.
3546 PH/41 $message_linecount is a new variable that contains the total number of
3547 lines in the message. Compare $body_linecount, which is the count for the
3550 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3551 bounce and delay warning messages, except in certain special cases, which
3554 (a) An SMTP error message from a remote host;
3555 (b) A message specified in a :fail: redirection;
3556 (c) A message specified in a "fail" command in a system filter;
3557 (d) A message specified in a FAIL return from the queryprogram router;
3558 (e) A message specified by the cannot_route_message router option.
3560 In these cases only, Exim does include the error details in bounce and
3561 warning messages. There are also a few cases where bland messages such
3562 as "unrouteable address" or "local delivery error" are given.
3564 PH/43 $value is now also set for the "else" part of a ${run expansion.
3566 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3567 being worked on, but at least Exim now implements the latest version to
3570 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3571 process to complete was treated as a delivery failure, a timeout while
3572 writing the message to the pipe was logged, but erroneously treated as a
3573 successful delivery. Such timeouts include transport filter timeouts. For
3574 consistency with the overall process timeout, these timeouts are now
3575 treated as errors, giving rise to delivery failures by default. However,
3576 there is now a new Boolean option for the pipe transport called
3577 timeout_defer, which, if set TRUE, converts the failures into defers for
3578 both kinds of timeout. A transport filter timeout is now identified in
3581 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3582 systems where "make" and "gmake" are different, calling "gmake" at top
3583 level broke things. I've arranged for the value of $(MAKE) to be passed
3584 from the Makefile to this script so that it can call the same version of
3588 A note about Exim versions 4.44 and 4.50
3589 ----------------------------------------
3591 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3592 changes of various kinds. As a consequence, a big documentation update was
3593 needed. This delayed the release for rather longer than seemed good, especially
3594 in the light of a couple of (minor) security issues. Therefore, the changes
3595 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3596 release. So 4.44 and 4.50 are in effect two different branches that both start
3599 I have left the 4.50 change log unchanged; it contains all the changes since
3600 4.43. The change log for 4.44 is below; many of its items are identical to
3601 those for 4.50. This seems to be the most sensible way to preserve the
3602 historical information.
3608 1. Minor wording change to the doc/README.SIEVE file.
3610 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3611 computation of the current number of files was incorrect.
3613 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3614 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3617 4. Give more explanation in the error message when the command for a transport
3618 filter fails to execute.
3620 5. There are several places where Exim runs a non-Exim command in a
3621 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3622 was being done only for the command run by the queryprogram router. It is
3623 now done for all such subprocesses. The other cases are: ${run, transport
3624 filters, and the commands run by the lmtp and pipe transports.
3626 6. Added CONFIGURE_GROUP build-time option.
3628 7. Some older OS have a limit of 256 on the maximum number of file
3629 descriptors. Exim was using setrlimit() to set 1000 as a large value
3630 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3631 systems. I've change it so that if it can't get 1000, it tries for 256.
3633 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3634 was an oversight, and furthermore, ever since the addition of extra
3635 controls (e.g. 4.43/32), the checks on when to allow different forms of
3636 "control" were broken. There should now be diagnostics for all cases when a
3637 control that does not make sense is encountered.
3639 9. Added the /retain_sender option to "control=submission".
3641 10. $recipients is now available in the predata ACL (oversight).
3643 11. Tidy the search cache before the fork to do a delivery from a message
3644 received from the command line. Otherwise the child will trigger a lookup
3645 failure and thereby defer the delivery if it tries to use (for example) a
3646 cached ldap connection that the parent has called unbind on.
3648 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3649 of $address_data from the recipient verification was clobbered by the
3650 sender verification.
3652 13. The value of address_data from a sender verification is now available in
3653 $sender_address_data in subsequent conditions in the ACL statement.
3655 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3657 15. Added a new option "connect=<time>" to callout options, to set a different
3660 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3661 was its contents. (It was OK if the option was not defined at all.)
3663 17. A "Completed" log line is now written for messages that are removed from
3664 the spool by the -Mrm option.
3666 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3667 information about exactly what failed.
3669 19. Added -dd to debug only the daemon process.
3671 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3672 handles timeouts, both on the server side and network timeouts. Renamed the
3673 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3675 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3676 It is now set to "smtps".
3678 22. $host_address is now set to the target address during the checking of
3679 ignore_target_hosts.
3681 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3682 being passed; this would have caused $sender_host_name to have been used if
3683 matching the list had actually called for a host name (not very likely,
3684 since this list is usually IP addresses). A host name is now passed as
3687 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3688 code that specifies a non-restarting handler (typically sigaction() in
3689 modern systems) in an attempt to fix a rare and obscure crash bug.
3691 25. Narrowed the window for a race in the daemon that could cause it to ignore
3692 SIGCHLD signals. This is not a major problem, because they are used only to
3693 wake it up if nothing else does.
3695 26. A malformed maildirsize file could cause Exim to calculate negative values
3696 for the mailbox size or file count. Odd effects could occur as a result.
3697 The maildirsize information is now recalculated if the size or filecount
3700 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3701 support for a long time. Removed HAVE_SYS_VFS_H.
3703 28. Installed the latest version of exipick from John Jetmore.
3705 29. In an address list, if the pattern was not a regular expression, an empty
3706 subject address (from a bounce message) matched only if the pattern was an
3707 empty string. Non-empty patterns were not even tested. This was the wrong
3708 because it is perfectly reasonable to use an empty address as part of a
3709 database query. An empty address is now tested by patterns that are
3710 lookups. However, all the other forms of pattern expect the subject to
3711 contain a local part and a domain, and therefore, for them, an empty
3712 address still always fails if the pattern is not itself empty.
3714 30. Exim went into a mad DNS loop when attempting to do a callout where the
3715 host was specified on an smtp transport, and looking it up yielded more
3716 than one IP address.
3718 31. Re-factored the code for checking spool and log partition space into a
3719 function that finds that data and another that does the check. The former
3720 is then used to implement four new variables: $spool_space, $log_space,
3721 $spool_inodes, and $log_inodes.
3723 32. The RFC2047 encoding function was originally intended for short strings
3724 such as real names; it was not keeping to the 75-character limit for
3725 encoded words that the RFC imposes. It now respects the limit, and
3726 generates multiple encoded words if necessary. To be on the safe side, I
3727 have increased the buffer size for the ${rfc2047: expansion operator from
3730 33. It is now permitted to omit both strings after an "if" condition; if the
3731 condition is true, the result is "true". As before, when the second string
3732 is omitted, a false condition yields an empty string. This makes it less
3733 cumbersome to write custom ACL and router conditions.
3735 34. Failure to deliver a bounce message always caused it to be frozen, even if
3736 there was an errors_to setting on the router. The errors_to setting is now
3739 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3740 canonical form (fully expanded) before being placed in
3741 $sender_host_address.
3743 36. The table in the code that translates DNS record types into text (T_A to
3744 "A" for instance) was missing entries for NS and CNAME. It is just possible
3745 that this could have caused confusion if both these types were looked up
3746 for the same domain, because the text type is used as part of Exim's
3747 per-process caching. But the chance of anyone hitting this buglet seems
3750 37. The dnsdb lookup has been extended in a number of ways.
3752 (1) There is a new type, "zns", which walks up the domain tree until it
3753 finds some nameserver records. It should be used with care.
3755 (2) There is a new type, "mxh", which is like "mx" except that it returns
3756 just the host names, not the priorities.
3758 (3) It is now possible to give a list of domains (or IP addresses) to be
3759 looked up. The behaviour when one of the lookups defers can be
3760 controlled by a keyword.
3762 (4) It is now possible to specify the separator character for use when
3763 multiple records are returned.
3765 38. The dnslists ACL condition has been extended: it is now possible to supply
3766 a list of IP addresses and/or domains to be looked up in a particular DNS
3769 39. Added log_selector=+queue_time_overall.
3771 40. When running the queue in the test harness, wait just a tad after forking a
3772 delivery process, to get repeatability of debugging output.
3774 41. Include certificate and key file names in error message when GnuTLS fails
3775 to set them up, because the GnuTLS error message doesn't include the name
3776 of the failing file when there is a problem reading it.
3778 42. Allow both -bf and -bF in the same test run.
3780 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3782 44. The "Exiscan patch" is now merged into the mainline Exim source.
3784 45. Sometimes the final signoff response after QUIT could fail to get
3785 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3786 < 0 before doing a fflush(). This bug looks as though it goes back to the
3787 introduction of TLS in release 3.20, but "sometimes" must have been rare
3788 because the tests only now provoked it.
3790 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3791 (this can affect the format of dates).
3793 47. exim_tidydb, when checking for the continued existence of a message for
3794 which it has found a message-specific retry record, was not finding
3795 messages that were in split spool directories. Consequently, it was
3796 deleting retry records that should have stayed in existence.
3798 48. Steve fixed some bugs in eximstats.
3800 49. The SPA authentication driver was not abandoning authentication and moving
3801 on to the next authenticator when an expansion was forced to fail,
3802 contradicting the general specification for all authenticators. Instead it
3803 was generating a temporary error. It now behaves as specified.
3805 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3806 (the order specifies the preference for clients). The order is now AES256,
3807 AES128, 3DES, ARCFOUR128.
3809 51. Small patch to Sieve code - explicitly set From: when generating an
3812 52. Exim crashed if a remote delivery caused a very long error message to be
3813 recorded - for instance if somebody sent an entire SpamAssassin report back
3814 as a large number of 550 error lines. This bug was coincidentally fixed by
3815 increasing the size of one of Exim's internal buffers (big_buffer) that
3816 happened as part of the Exiscan merge. However, to be on the safe side, I
3817 have made the code more robust (and fixed the comments that describe what
3820 53. Now that there can be additional text after "Completed" in log lines (if
3821 the queue_time_overall log selector is set), a one-byte patch to exigrep
3822 was needed to allow it to recognize "Completed" as not the last thing in
3825 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3826 patch that reportedly fixes this has been added. I am not expert enough to
3827 create a test for it. This is what the patch creator wrote:
3829 "I found a little strange behaviour of ldap code when working with
3830 Windows 2003 AD Domain, where users was placed in more than one
3831 Organization Units. When I tried to give exim partial DN, the exit code
3832 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3833 But simultaneously result of request was absolutely normal ldap result,
3834 so I produce this patch..."
3836 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3837 REFERENCE, so I have modified the code to exclude the patch when that macro
3840 55. Some experimental protocols are using DNS PTR records for new purposes. The
3841 keys for these records are domain names, not reversed IP addresses. The
3842 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3843 leaves it alone. Component reversal etc. now happens only for IP addresses.
3846 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3848 57. Double the size of the debug message buffer (to 2048) so that more of very
3849 long debug lines gets shown.
3851 58. The exicyclog utility now does better if the number of log files to keep
3852 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3854 59. Two changes related to the smtp_active_hostname option:
3856 (1) $smtp_active_hostname is now available as a variable.
3857 (2) The default for smtp_banner uses $smtp_active_hostname instead
3858 of $primary_hostname.
3860 60. The host_aton() function is supposed to be passed a string that is known
3861 to be a valid IP address. However, in the case of IPv6 addresses, it was
3862 not checking this. This is a hostage to fortune. Exim now panics and dies
3863 if the condition is not met. A case was found where this could be provoked
3864 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3865 components; fortuitously, this particular loophole had already been fixed
3866 by change 4.50/55 above.
3868 If there are any other similar loopholes, the new check in host_aton()
3869 itself should stop them being exploited. The report I received stated that
3870 data on the command line could provoke the exploit when Exim was running as
3871 exim, but did not say which command line option was involved. All I could
3872 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3873 running as the user.
3876 61. There was a buffer overflow vulnerability in the SPA authentication code
3877 (which came originally from the Samba project). I have added a test to the
3878 spa_base64_to_bits() function which I hope fixes it.
3881 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3882 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3884 63. The daemon start-up calls getloadavg() while still root for those OS that
3885 need the first call to be done as root, but it missed one case: when
3886 deliver_queue_load_max is set with deliver_drop_privilege. This is
3887 necessary for the benefit of the queue runner, because there is no re-exec
3888 when deliver_drop_privilege is set.
3890 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3891 This has been fixed.
3893 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3894 was in use, was not putting the data itself into the right store pool;
3895 consequently, it could be overwritten for a subsequent message in the same
3896 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3899 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3901 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3902 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3903 the function were treating the return as a boolean value, which happened to
3904 work because 0=false and not-0=true, but is not correct code.
3906 68. The host_aton() function was not handling scoped IPv6 addresses (those
3907 with, for example, "%eth0" on the end) correctly.
3909 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3910 negated items (that is, ~something) in unsigned ints. Some compilers
3911 apparently mutter when there is no cast.
3913 70. If an address verification called from an ACL failed, and did not produce a
3914 user-specific message (i.e. there was only a "system" message), nothing was
3915 put in $acl_verify_message. In this situation, it now puts the system
3918 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3919 session; change 4.31/43 added the unwanted input to the log line - except
3920 that it did not do this in the start of session case. It now does.
3922 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3923 This is wrong and can cause the other end to generate a synchronization
3924 error if it is another Exim or anything else that does the synchronization
3925 check. A QUIT command is no longer sent after a timeout.
3927 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3928 during host lookups.
3930 74. The defer_ok option of callout verification was not working if it was used
3931 when verifying addresses in header lines, that is, for this case:
3933 verify = header_sender/callout=defer_ok
3935 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3936 those file descriptors could be used for SMTP connections. If anything
3937 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3938 could be sent to the SMTP client, causing chaos. The daemon now opens
3939 stdin, stdout, and stderr to /dev/null when it puts itself into the
3942 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3943 log by default. The user can override this with suitable Perl magic.
3945 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3946 the log message when discard triggers, was not working for the DATA ACL or
3947 for the non-SMTP ACL.
3949 78. Error message wording change in sieve.c.
3951 79. If smtp_accept_max_per_host was set, the number of connections could be
3952 restricted to fewer than expected, because the daemon was trying to set up
3953 a new connection before checking whether the processes handling previous
3954 connections had finished. The check for completed processes is now done
3955 earlier. On busy systems, this bug wouldn't be noticed because something
3956 else would have woken the daemon, and it would have reaped the completed
3959 80. If a message was submitted locally by a user whose login name contained one
3960 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3961 It caused a spool format error. I have fixed the spool reading code. A
3962 related problem was that the "from" clause in the Received: line became
3963 illegal because of the space(s). It is now covered by ${quote_local_part.
3965 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3968 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3970 83. Patch from Sieve maintainer to fix unterminated string problem in
3971 "vacation" handling.
3973 84. Some minor changes to the Linux configuration files to help with other
3974 OS variants using glibc.
3976 85. One more patch for Sieve to update vacation handling to latest spec.
3979 ----------------------------------------------------
3980 See the note above about the 4.44 and 4.50 releases.
3981 ----------------------------------------------------
3987 1. Change 4.43/35 introduced a bug that caused file counts to be
3988 incorrectly computed when quota_filecount was set in an appendfile
3991 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3992 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3995 3. Give more explanation in the error message when the command for a transport
3996 filter fails to execute.
3998 4. There are several places where Exim runs a non-Exim command in a
3999 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4000 was being done only for the command run by the queryprogram router. It is
4001 now done for all such subprocesses. The other cases are: ${run, transport
4002 filters, and the commands run by the lmtp and pipe transports.
4004 5. Some older OS have a limit of 256 on the maximum number of file
4005 descriptors. Exim was using setrlimit() to set 1000 as a large value
4006 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4007 systems. I've change it so that if it can't get 1000, it tries for 256.
4009 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4010 was an oversight, and furthermore, ever since the addition of extra
4011 controls (e.g. 4.43/32), the checks on when to allow different forms of
4012 "control" were broken. There should now be diagnostics for all cases when a
4013 control that does not make sense is encountered.
4015 7. $recipients is now available in the predata ACL (oversight).
4017 8. Tidy the search cache before the fork to do a delivery from a message
4018 received from the command line. Otherwise the child will trigger a lookup
4019 failure and thereby defer the delivery if it tries to use (for example) a
4020 cached ldap connection that the parent has called unbind on.
4022 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4023 of $address_data from the recipient verification was clobbered by the
4024 sender verification.
4026 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4027 was its contents. (It was OK if the option was not defined at all.)
4029 11. A "Completed" log line is now written for messages that are removed from
4030 the spool by the -Mrm option.
4032 12. $host_address is now set to the target address during the checking of
4033 ignore_target_hosts.
4035 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4036 being passed; this would have caused $sender_host_name to have been used if
4037 matching the list had actually called for a host name (not very likely,
4038 since this list is usually IP addresses). A host name is now passed as
4041 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4042 code that specifies a non-restarting handler (typically sigaction() in
4043 modern systems) in an attempt to fix a rare and obscure crash bug.
4045 15. Narrowed the window for a race in the daemon that could cause it to ignore
4046 SIGCHLD signals. This is not a major problem, because they are used only to
4047 wake it up if nothing else does.
4049 16. A malformed maildirsize file could cause Exim to calculate negative values
4050 for the mailbox size or file count. Odd effects could occur as a result.
4051 The maildirsize information is now recalculated if the size or filecount
4054 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4055 support for a long time. Removed HAVE_SYS_VFS_H.
4057 18. Updated exipick to current release from John Jetmore.
4059 19. Allow an empty sender to be matched against a lookup in an address list.
4060 Previously the only cases considered were a regular expression, or an
4063 20. Exim went into a mad DNS lookup loop when doing a callout where the
4064 host was specified on the transport, if the DNS lookup yielded more than
4067 21. The RFC2047 encoding function was originally intended for short strings
4068 such as real names; it was not keeping to the 75-character limit for
4069 encoded words that the RFC imposes. It now respects the limit, and
4070 generates multiple encoded words if necessary. To be on the safe side, I
4071 have increased the buffer size for the ${rfc2047: expansion operator from
4074 22. Failure to deliver a bounce message always caused it to be frozen, even if
4075 there was an errors_to setting on the router. The errors_to setting is now
4078 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4079 canonical form (fully expanded) before being placed in
4080 $sender_host_address.
4082 24. Updated eximstats to version 1.33
4084 25. Include certificate and key file names in error message when GnuTLS fails
4085 to set them up, because the GnuTLS error message doesn't include the name
4086 of the failing file when there is a problem reading it.
4088 26. Expand error message when OpenSSL has problems setting up cert/key files.
4091 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4092 (this can affect the format of dates).
4094 28. exim_tidydb, when checking for the continued existence of a message for
4095 which it has found a message-specific retry record, was not finding
4096 messages that were in split spool directories. Consequently, it was
4097 deleting retry records that should have stayed in existence.
4099 29. eximstats updated to version 1.35
4100 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4101 1.35 - bugfix such that pie charts by volume are generated correctly
4103 30. The SPA authentication driver was not abandoning authentication and moving
4104 on to the next authenticator when an expansion was forced to fail,
4105 contradicting the general specification for all authenticators. Instead it
4106 was generating a temporary error. It now behaves as specified.
4108 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4109 (the order specifies the preference for clients). The order is now AES256,
4110 AES128, 3DES, ARCFOUR128.
4112 31. Small patch to Sieve code - explicitly set From: when generating an
4115 32. Exim crashed if a remote delivery caused a very long error message to be
4116 recorded - for instance if somebody sent an entire SpamAssassin report back
4117 as a large number of 550 error lines. This bug was coincidentally fixed by
4118 increasing the size of one of Exim's internal buffers (big_buffer) that
4119 happened as part of the Exiscan merge. However, to be on the safe side, I
4120 have made the code more robust (and fixed the comments that describe what
4123 33. Some experimental protocols are using DNS PTR records for new purposes. The
4124 keys for these records are domain names, not reversed IP addresses. The
4125 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4126 leaves it alone. Component reversal etc. now happens only for IP addresses.
4129 34. The host_aton() function is supposed to be passed a string that is known
4130 to be a valid IP address. However, in the case of IPv6 addresses, it was
4131 not checking this. This is a hostage to fortune. Exim now panics and dies
4132 if the condition is not met. A case was found where this could be provoked
4133 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4134 components; fortuitously, this particular loophole had already been fixed
4135 by change 4.50/55 or 4.44/33 above.
4137 If there are any other similar loopholes, the new check in host_aton()
4138 itself should stop them being exploited. The report I received stated that
4139 data on the command line could provoke the exploit when Exim was running as
4140 exim, but did not say which command line option was involved. All I could
4141 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4142 running as the user.
4145 35. There was a buffer overflow vulnerability in the SPA authentication code
4146 (which came originally from the Samba project). I have added a test to the
4147 spa_base64_to_bits() function which I hope fixes it.
4150 36. The daemon start-up calls getloadavg() while still root for those OS that
4151 need the first call to be done as root, but it missed one case: when
4152 deliver_queue_load_max is set with deliver_drop_privilege. This is
4153 necessary for the benefit of the queue runner, because there is no re-exec
4154 when deliver_drop_privilege is set.
4156 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4157 was in use, was not putting the data itself into the right store pool;
4158 consequently, it could be overwritten for a subsequent message in the same
4159 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4162 38. Sometimes the final signoff response after QUIT could fail to get
4163 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4164 < 0 before doing a fflush(). This bug looks as though it goes back to the
4165 introduction of TLS in release 3.20, but "sometimes" must have been rare
4166 because the tests only now provoked it.
4172 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4173 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4174 Now it returns an integer. A number of calls were still expecting a T/F
4175 return. Fortuitously, in all cases, the tests worked in OK situations,
4176 which is the norm. However, things would have gone wrong on any write
4177 failures on the smtp file descriptor. This function is used when sending
4178 messages over SMTP and also when doing verify callouts.
4180 2. When Exim is called to do synchronous delivery of a locally submitted
4181 message (the -odf or -odi options), it no longer closes stderr before doing
4184 3. Implemented the mua_wrapper option.
4186 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4188 5. Implemented the functions header_remove(), header_testname(),
4189 header_add_at_position(), and receive_remove_recipient(), and exported them
4192 6. If an ACL "warn" statement specified the addition of headers, Exim already
4193 inserted X-ACL-Warn: at the start if there was no header name. However, it
4194 was not making this test for the second and subsequent header lines if
4195 there were newlines in the string. This meant that an invalid header could
4196 be inserted if Exim was badly configured.
4198 7. Allow an ACL "warn" statement to add header lines at the start or after all
4199 the Received: headers, as well as at the end.
4201 8. Added the rcpt_4xx retry error code.
4203 9. Added postmaster_mailfrom=xxx to callout verification option.
4205 10. Added mailfrom=xxxx to the callout verification option, for verify=
4208 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4209 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4211 12. Inserted some casts to stop certain compilers warning when using pointer
4212 differences as field lengths or precisions in printf-type calls (mostly
4213 affecting debugging statements).
4215 13. Added optional readline() support for -be (dynamically loaded).
4217 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4218 same clock tick as a message's arrival, so that its received time was the
4219 same as the "first fail" time on the retry record, and that message
4220 remained on the queue past the ultimate address timeout, every queue runner
4221 would try a delivery (because it was past the ultimate address timeout) but
4222 after another failure, the ultimate address timeout, which should have then
4223 bounced the address, did not kick in. This was a "< instead of <=" error;
4224 in most cases the first failure would have been in the next clock tick
4225 after the received time, and all would be well.
4227 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4228 being recognized when the domain list was tested by the match_domain
4229 condition in an expansion string.
4231 16. Added the ${str2b64: operator.
4233 17. Exim was always calling setrlimit() to set a large limit for the number of
4234 processes, without checking whether the existing limit was already
4235 adequate. (It did check for the limit on file descriptors.) Furthermore,
4236 errors from getrlimit() and setrlimit() were being ignored. Now they are
4237 logged to the main and panic logs, but Exim does carry on, to try to do its
4238 job under whatever limits there are.
4240 18. Imported PCRE 5.0.
4242 19. Trivial typo in log message " temporarily refused connection" (the leading
4245 20. If the log selector return_path_on_delivery was set and an address was
4246 redirected to /dev/null, the delivery process crashed because it assumed
4247 that a return path would always be set for a "successful" delivery. In this
4248 case, the whole delivery is bypassed as an optimization, and therefore no
4251 21. Internal re-arrangement: the function for sending a challenge and reading
4252 a response while authentication was assuming a zero-terminated challenge
4253 string. It's now changed to take a pointer and a length, to allow for
4254 binary data in such strings.
4256 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4258 23. Exim was not respecting finduser_retries when seeking the login of the
4259 uid under which it was called; it was always trying 10 times. (The default
4260 setting of finduser_retries is zero.) Also, it was sleeping after the final
4261 failure, which is pointless.
4263 24. Implemented tls_on_connect_ports.
4265 25. Implemented acl_smtp_predata.
4267 26. If the domain in control=submission is set empty, Exim assumes that the
4268 authenticated id is a complete email address when it generates From: or
4269 Sender: header lines.
4271 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4272 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4273 chown and chgrp in /bin and hostname in /usr/bin.
4275 28. Exim was keeping the "process log" file open after each use, just as it
4276 does for the main log. This opens the possibility of it remaining open for
4277 long periods when the USR1 signal hits a daemon. Occasional processlog
4278 errors were reported, that could have been caused by this. Anyway, it seems
4279 much more sensible not to leave this file open at all, so that is what now
4282 29. The long-running daemon process does not normally write to the log once it
4283 has entered its main loop, and it closes the log before doing so. This is
4284 so that log files can straightforwardly be renamed and moved. However,
4285 there are a couple of unusual error situations where the daemon does write
4286 log entries, and I had neglected to close the log afterwards.
4288 30. The text of an SMTP error response that was received during a remote
4289 delivery was being truncated at 512 bytes. This is too short for some of
4290 the long messages that one sometimes sees. I've increased the limit to
4293 31. It is now possible to make retry rules that apply only when a message has a
4294 specific sender, in particular, an empty sender.
4296 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4297 it possible to be selective about when SMTP synchronization is enforced.
4299 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4301 32. Implemented hosts_connection_nolog.
4303 33. Added an ACL for QUIT.
4305 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4308 35. Added mailbox_size and mailbox_filecount to appendfile.
4310 36. Added control = no_multiline_responses to ACLs.
4312 37. There was a bug in the logic of the code that waits for the clock to tick
4313 in the case where the clock went backwards by a substantial amount such
4314 that the microsecond fraction of "now" was more than the microsecond
4315 fraction of "then" (but the whole seconds number was less).
4317 38. Added support for the libradius Radius client library this is found on
4318 FreeBSD (previously only the radiusclient library was supported).
4324 1. When certain lookups returned multiple values in the form name=value, the
4325 quoting of the values was not always being done properly. Specifically:
4326 (a) If the value started with a double quote, but contained no whitespace,
4328 (b) If the value contained whitespace other than a space character (i.e.
4329 tabs or newlines or carriage returns) it was not quoted.
4330 This fix has been applied to the mysql and pgsql lookups by writing a
4331 separate quoting function and calling it from the lookup code. The fix
4332 should probably also be applied to nisplus, ibase and oracle lookups, but
4333 since I cannot test any of those, I have not disturbed their existing code.
4335 2. A hit in the callout cache for a specific address caused a log line with no
4336 reason for rejecting RCPT. Now it says "Previous (cached) callout
4337 verification failure".
4339 3. There was an off-by-one bug in the queryprogram router. An over-long
4340 return line was truncated at 256 instead of 255 characters, thereby
4341 overflowing its buffer with the terminating zero. As well as fixing this, I
4342 have increased the buffer size to 1024 (and made a note to document this).
4344 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4345 when Exim is waiting for an SMTP response from a remote server, Exim
4346 restarts its select() call on the socket, thereby resetting its timeout.
4347 This is not a problem when such interrupts are rare. Somebody set up a cron
4348 job to run exiwhat every 2 minutes, which is less than the normal select()
4349 timeout (5 or 10 minutes). This meant that the select() timeout never
4350 kicked in because it was always reset. I have fixed this by comparing the
4351 time when an interrupt arrives with the time at the start of the first call
4352 to select(). If more time than the timeout has elapsed, the interrupt is
4353 treated as a timeout.
4355 5. Some internal re-factoring in preparation for the addition of Sieve
4356 extensions (by MH). In particular, the "personal" test is moved to a
4357 separate function, and given an option for scanning Cc: and Bcc: (which is
4358 not set for Exim filters).
4360 6. When Exim created an email address using the login of the caller as the
4361 local part (e.g. when creating a From: or Sender: header line), it was not
4362 quoting the local part when it contained special characters such as @.
4364 7. Installed new OpenBSD configuration files.
4366 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4367 try to make them clearer.
4369 9. Callout options, other than the timeout value, were being ignored when
4370 verifying sender addresses in header lines. For example, when using
4372 verify = header_sender/callout=no_cache
4374 the cache was (incorrectly) being used.
4376 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4377 only the Cygwin environment.
4379 11. When return_path_on_delivery was set as a log selector, if different remote
4380 addresses in the same message used different return paths and parallel
4381 remote delivery occurred, the wrong values would sometimes be logged.
4382 (Whenever a remote delivery process finished, the return path value from
4383 the most recently started remote delivery process was logged.)
4385 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4386 lines when AUTH and/or TLS are in use. This is the "received protocol"
4387 field. Exim used to use "asmtp" for authenticated SMTP, without any
4388 indication (in the protocol name) for TLS use. Now it follows the RFC and
4389 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4390 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4391 names appear in log lines as well as in Received: header lines.
4393 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4394 extensions, and comparison tests, and to fix some bugs.
4396 14. Changes to the "personal" filter test:
4398 (1) The test was buggy in that it was just doing the equivalent of
4399 "contains" tests on header lines. For example, if a user's address was
4400 anne@some.where, the "personal" test would incorrectly be true for
4402 To: susanne@some.where
4404 This test is now done by extracting each address from the header in turn,
4405 and checking the entire address. Other tests that are part of "personal"
4406 are now done using regular expressions (for example, to check local parts
4407 of addresses in From: header lines).
4409 (2) The list of non-personal local parts in From: addresses has been
4410 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4411 taken from the Sieve specification recommendations.
4413 (3) If the message contains any header line starting with "List-" it is
4414 treated as non-personal.
4416 (4) The test for "circular" in the Subject: header line has been removed
4417 because it now seems ill-conceived.
4419 15. Minor typos in src/EDITME comments corrected.
4421 16. Installed latest exipick from John Jetmore.
4423 17. If headers_add on a router specified a text string that was too long for
4424 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4425 of string_sprintf() is now avoided.
4427 18. $message_body_size was not set (it was always zero) when running the DATA
4428 ACL and the local_scan() function.
4430 19. For the "mail" command in an Exim filter, no default was being set for
4431 the once_repeat time, causing a random time value to be used if "once" was
4432 specified. (If the value happened to be <= 0, no repeat happened.) The
4433 default is now 0s, meaning "never repeat". The "vacation" command was OK
4434 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4435 (I found it when inspecting the code).
4437 20. There is now an overall timeout for performing a callout verification. It
4438 defaults to 4 times the callout timeout, which applies to individual SMTP
4439 commands during the callout. The overall timeout applies when there is more
4440 than one host that can be tried. The timeout is checked before trying the
4441 next host. This prevents very long delays if there are a large number of
4442 hosts and all are timing out (e.g. when the network connections are timing
4443 out). The value of the overall timeout can be changed by specifying an
4444 additional sub-option for "callout", called "maxwait". For example:
4446 verify = sender/callout=5s,maxwait=20s
4448 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4449 to the end before writing, but this should make it even safer).
4451 22. Exim was forgetting that it had advertised PIPELINING for the second and
4452 subsequent messages on an SMTP connection. It was also not resetting its
4453 memory on STARTTLS and an internal HELO.
4455 23. When Exim logs an SMTP synchronization error within a session, it now
4456 records whether PIPELINING has been advertised or not.
4458 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4459 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4460 rather than long int.
4462 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4464 26. Added the never_mail option to autoreply.
4470 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4471 crash if the getsockname() call failed; this can happen if a connection is
4472 closed very soon after it is established. The problem was simply in the
4473 order in which certain operations were done, causing Exim to try to write
4474 to the SMTP stream before it had set up the file descriptor. The bug has
4475 been fixed by making things happen in the correct order.
4481 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4482 before the connection was closed, thus losing the rejection response.
4484 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4485 some early Solaris releases, but causes trouble in current releases where
4486 socklen_t is defined.
4488 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4491 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4494 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4495 that releasing the top of it at the end releases what was used for sub-
4496 expansions (unless the block got too big). However, discard this block if
4497 the first thing is a variable or header, so that we can use its block when
4498 it is dynamic (useful for very large $message_headers, for example).
4500 6. Lookups now cache *every* query, not just the most recent. A new, separate
4501 store pool is used for this. It can be recovered when all lookup caches are
4502 flushed. Lookups now release memory at the end of their result strings.
4503 This has involved some general refactoring of the lookup sources.
4505 7. Some code has been added to the store_xxx() functions to reduce the amount
4506 of flapping under certain conditions.
4508 8. log_incoming_interface used to affect only the <= reception log lines. Now
4509 it causes the local interface and port to be added to several more SMTP log
4510 lines, for example "SMTP connection from", and rejection lines.
4512 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4514 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4516 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4517 could be overwritten at the end of the current message (or the start of a
4518 new message if it was set in a HELO ACL). The value is now preserved for
4519 the duration of the SMTP connection.
4521 12. If a transport had a headers_rewrite setting, and a matching header line
4522 contained an unqualified address, that address was qualified, even if it
4523 did not match any rewriting rules. The underlying bug was that the values
4524 of the flags that permit the existence of unqualified sender and recipient
4525 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4526 non-local messages, and by -bnq for local messages) were not being
4527 preserved with the message after it was received.
4529 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4530 "next input=" as part of the text comprising the host identity instead of
4531 the correct text. The code was using the same buffer for two different
4532 strings. However, depending on which order the printing function evaluated
4533 its arguments, the bug did not always show up. Under Linux, for example, my
4534 test suite worked just fine.
4536 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4537 doesn't work with some older versions of Perl. It has been changed to "my",
4538 which in any case is probably the better facility to use.
4540 15. A really picky compiler found some instances of statements for creating
4541 error messages that either had too many or two few arguments for the format
4544 16. The size of the buffer for calls to the DNS resolver has been increased
4545 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4546 for addresses that have a lot of PTR records. This alleviates a problem; it
4547 does not fully solve it.
4549 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4550 buffer now truncates the list and logs the incident, which is the same
4551 action as happens when Exim is looking up a host name and its aliases.
4552 Previously in this situation something unpredictable would happen;
4553 sometimes it was "internal error: store_reset failed".
4555 18. If a server dropped the connection unexpectedly when an Exim client was
4556 using GnuTLS and trying to read a response, the client delivery process
4557 crashed while trying to generate an error log message.
4559 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4560 string, for example:
4562 warn message = H1: something\nH2: something
4564 the text was added as a single header line from Exim's point of view
4565 though it ended up OK in the delivered message. However, searching for the
4566 second and subsequent header lines using $h_h2: did not work. This has been
4567 fixed. Similarly, if a system filter added multiple headers in this way,
4568 the routers could not see them.
4570 20. Expanded the error message when iplsearch is called with an invalid key to
4571 suggest using net-iplsearch in a host list.
4573 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4574 ACLs are no longer actually imposed (and a message to that effect is
4577 22. If a "gecos" field in a passwd entry contained escaped characters, in
4578 particular, if it contained a \" sequence, Exim got it wrong when building
4579 a From: or a Sender: header from that name. A second bug also caused
4580 incorrect handling when an unquoted " was present following a character
4581 that needed quoting.
4583 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4584 was not being matched caselessly.
4586 24. Arranged for all hyphens in the exim.8 source to be escaped with
4589 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4590 properly. Recipient callout cache records were still being keyed to include
4591 the sender, even when use_sender was set false. This led to far more
4592 callouts that were necessary. The sender is no longer included in the key
4593 when use_sender is false.
4595 26. Added "control = submission" modifier to ACLs.
4597 27. Added the ${base62d: operator to decode base 62 numbers.
4599 28. dnsdb lookups can now access SRV records.
4601 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4602 the configuration file.
4604 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4605 (-v) mode. This makes the output for a verbose queue run more intelligible.
4607 31. Added a use_postmaster feature to recipient callouts.
4609 32. Added the $body_zerocount variable, containing the number of binary zero
4610 bytes in the message body.
4612 33. The time of last modification of the "new" subdirectory is now used as the
4613 "mailbox time last read" when there is a quota error for a maildir
4616 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4618 35. Added +ignore_unknown as a special item in host lists.
4620 36. Code for decoding IPv6 addresses in host lists is now included, even if
4621 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4622 address was recognized as an IP address, but was then not correctly decoded
4623 into binary, causing unexpected and incorrect effects when compared with
4630 1. Very minor rewording of debugging text in manualroute to say "list of
4631 hosts" instead of "hostlist".
4633 2. If verify=header_syntax was set, and a header line with an unqualified
4634 address (no domain) and a large number of spaces between the end of the
4635 name and the colon was received, the reception process suffered a buffer
4636 overflow, and (when I tested it) crashed. This was caused by some obsolete
4637 code that should have been removed. The fix is to remove it!
4639 3. When running in the test harness, delay a bit after writing a bounce
4640 message to get a bit more predictability in the log output.
4642 4. Added a call to search_tidyup() just before forking a reception process. In
4643 theory, someone could use a lookup in the expansion of smtp_accept_max_
4644 per_host which, without the tidyup, could leave open a database connection.
4646 5. Added the variables $recipient_data and $sender_data which get set from a
4647 lookup success in an ACL "recipients" or "senders" condition, or a router
4648 "senders" option, similar to $domain_data and $local_part_data.
4650 6. Moved the writing of debug_print from before to after the "senders" test
4653 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4654 problems for message scanning, either using a data ACL, or using
4655 local_scan() because the Received: header was not generated till after they
4656 were called (in order to set the time as the time of reception completion).
4657 I have revised the way this works. The header is now generated after the
4658 body is received, but before the ACL or local_scan() are called. After they
4659 are run, the timestamp in the header is updated.
4665 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4666 before starting a queue runner without re-exec. This happened only when
4667 deliver_drop_privilege was set or when the Exim user was set to root. The
4668 effect of the bug was that timeouts during subsequent deliveries caused
4669 crashes instead of being properly handled. The handler is now left at its
4670 default (and expected) setting.
4672 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4673 message, again when deliver_drop_privilege is set or Exim is run as root.
4674 The bug described in (1) was not present in this case, but the tidying up
4675 of the other signals was missing. I have made the two cases consistent.
4677 3. The ignore_target_hosts setting on a manualroute router was being ignored
4678 for hosts that were looked up using the /MX notation.
4680 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4683 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4684 operated on the sender address. After changing the $sender_address to <>
4685 for the sender address verify, Exim was re-instated it as the original
4686 (before rewriting) address, but remembering that it had rewritten it, so it
4687 wasn't rewriting it again. This bug also had the effect of breaking the
4688 sender address verification caching when the sender address was rewritten.
4690 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4691 This has been changed so that if the ip literal address matches
4692 ignore_target_hosts, the router declines.
4694 7. Added expansion conditions match_domain, match_address, and match_local_
4695 part (NOT match_host).
4697 8. The placeholder for the Received: header didn't have a length field set.
4699 9. Added code to Exim itself and to exim_lock to test for a specific race
4700 condition that could lead to file corruption when using MBX delivery. The
4701 issue is with the lockfile that is created in /tmp. If this file is removed
4702 after a process has opened it but before that process has acquired a lock,
4703 there is the potential for a second process to recreate the file and also
4704 acquire a lock. This could lead to two Exim processes writing to the file
4705 at the same time. The added code performs the same test as UW imapd; it
4706 checks after acquiring the lock that its file descriptor still refers to
4707 the same named file.
4709 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4710 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4713 11. Added the smtp_active_hostname option. If used, this will typically be made
4714 to depend on the incoming interface address. Because $interface_address is
4715 not set up until the daemon has forked a reception process, error responses
4716 that can happen earlier (such as "too many connections") no longer contain
4719 12. If an expansion in a condition on a "warn" statement fails because a lookup
4720 defers, the "warn" statement is abandoned, and the next ACL statement is
4721 processed. Previously this caused the whole ACL to be aborted.
4723 13. Added the iplsearch lookup type.
4725 14. Added ident_timeout as a log selector.
4727 15. Added tls_certificate_verified as a log selector.
4729 16. Added a global option tls_require_ciphers (compare the smtp transport
4730 option of the same name). This controls incoming TLS connections.
4732 17. I finally figured out how to make tls_require_ciphers do a similar thing
4733 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4734 before starting the TLS session.
4736 18. Tabs are now shown as \t in -bP output.
4738 19. If the log selector return_path_on_delivery was set, Exim crashed when
4739 bouncing a message because it had too many Received: header lines.
4741 20. If two routers both had headers_remove settings, and the first one included
4742 a superfluous trailing colon, the final name in the first list and the
4743 first name in the second list were incorrectly joined into one item (with a
4744 colon in the middle).
4750 1. Added -C and -D options to the exinext utility, mainly to make it easier
4751 to include in the automated testing, but these could be helpful when
4752 multiple configurations are in use.
4754 2. The exinext utility was not formatting the output nicely when there was
4755 an alternate port involved in the retry record key, nor when there was a
4756 message id as well (for retries that were specific to a specific message
4757 and a specific host). It was also confused by IPv6 addresses, because of
4758 the additional colons they contain. I have fixed the IPv4 problem, and
4759 patched it up to do a reasonable job for IPv6.
4761 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4762 delivery, the log line now contains "pipelined" if PIPELINING was used.
4764 4. An SMTP transport process used to panic and die if the bind() call to set
4765 an explicit outgoing interface failed. This has been changed; it is now
4766 treated in the same way as a connect() failure.
4768 5. A reference to $sender_host_name in the part of a conditional expansion
4769 that was being skipped was still causing a DNS lookup. This no longer
4772 6. The def: expansion condition was not recognizing references to header lines
4773 that used bh_ and bheader_.
4775 7. Added the _cache feature to named lists.
4777 8. The code for checking quota_filecount in the appendfile transport was
4778 allowing one more file than it should have been.
4780 9. For compatibility with Sendmail, the command line option
4788 and sets the incoming protocol and host name (for trusted callers). The
4789 host name and its colon can be omitted when only the protocol is to be set.
4790 Note the Exim already has two private options, -pd and -ps, that refer to
4791 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4792 "s", but I don't think that's a major issue.
4794 10. A number of refactoring changes to the code, none of which should affect
4797 (a) The number of logging options was getting close to filling up the
4798 32-bit word that was used as a bit map. I have split them into two classes:
4799 those that are passed in the argument to log_write(), and those that are
4800 only ever tested independently outside of that function. These are now in
4801 separate 32-bit words, so there is plenty of room for expansion again.
4802 There is no change in the user interface or the logging behaviour.
4804 (b) When building, for example, log lines, the code previously used a
4805 macro that called string_cat() twice, in order to add two strings. This is
4806 not really sufficiently general. Furthermore, there was one instance where
4807 it was actually wrong because one of the argument was used twice, and in
4808 one call a function was used. (As it happened, calling the function twice
4809 did not affect the overall behaviour.) The macro has been replaced by a
4810 function that can join an arbitrary number of extra strings onto a growing
4813 (c) The code for expansion conditions now uses a table and a binary chop
4814 instead of a serial search (which was left over from when there were very
4815 few conditions). Also, it now recognizes conditions like "pam" even when
4816 the relevant support is not compiled in: a suitably worded error message is
4817 given if an attempt is made to use such a condition.
4819 11. Added ${time_interval:xxxxx}.
4821 12. A bug was causing one of the ddress fields not to be passed back correctly
4822 from remote delivery subprocesses. The field in question was not being
4823 subsequently used, so this caused to problems in practice.
4825 13. Added new log selectors queue_time and deliver_time.
4827 14. Might have fixed a bug in maildirsizefile handling that threw up
4828 "unexpected character" debug warnings, and recalculated the data
4829 unnecessarily. In any case, I expanded the warning message to give more
4832 15. Added the message "Restricted characters in address" to the statements in
4833 the default ACL that block characters like @ and % in local parts.
4835 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4836 Three changes have been made:
4838 (a) There was a serious bug; a negative response to MAIL caused the whole
4839 recipient domain to be cached as invalid, thereby blocking all messages
4840 to all local parts at the same domain, from all senders. This bug has
4841 been fixed. The domain is no longer cached after a negative response to
4842 MAIL if the sender used is not empty.
4844 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4847 (c) A new callout option, "use_sender" has been added for people who want
4848 the modified behaviour.
4854 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4857 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4858 indeed breaks things for older releases.
4860 3. Added additional logging to the case where there is a problem reading data
4861 from a filter that is running in a subprocess using a pipe, in order to
4862 try to track down a specific problem.
4864 4. Testing facility fudge: when running in the test harness and attempting
4865 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4866 getting "No route to host". Convert this to a timeout.
4868 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4871 6. Some OS don't have socklen_t but use size_t instead. This affects the
4872 fifth argument of getsockopt() amongst other things. This is now
4873 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4874 can be set for individual OS. I have set it for SunOS5, OSF1, and
4875 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4876 some earlier ones do not.
4878 7. Change 4.30/15 was not doing the test caselessly.
4880 8. The standard form for an IPv6 address literal was being rejected by address
4881 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4882 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4883 this, as well as the form without the "IPv6" on the front (but only when
4884 address literals are enabled, of course).
4886 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4888 10. Exim crashed if a message with an empty sender address specified by -f
4889 encountered a router with an errors_to setting. This could be provoked only
4890 by a command such as
4894 where an empty string was supplied; "<>" did not hit this bug.
4896 11. Installed PCRE release 4.5.
4898 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4899 remained set. It is now erased.
4901 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4902 times from message ids (which are base 36 rather than the normal 62).
4904 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4905 were being counted as actual protocol errors, and logged if the log
4906 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4907 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4908 and DATA following a set of rejected RCPTs do not count as protocol errors.
4909 In other words, Exim assumes they were pipelined, though this may not
4910 actually be the case. Of course, in all cases the client gets an
4911 appropriate error code.
4913 15. If a lookup fails in an ACL condition, a message about the failure may
4914 be available; it is used if testing the ACL cannot continue, because most
4915 such messages specify what the cause of the deferral is. However, some
4916 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4917 that caused an old message to be retained and used if a later statement
4918 caused a defer, replacing the real cause of the deferral.
4920 16. If an IP address had so many PTR records that the DNS lookup buffer
4921 was not large enough to hold them, Exim could crash while trying to process
4922 the truncated data. It now detects and logs this case.
4924 17. Further to 4.21/58, another change has been made: if (and only if) the
4925 first line of a message (the first header line) ends with CRLF, a bare LF
4926 in a subsequent header line has a space inserted after it, so as not to
4927 terminate the header.
4929 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4930 unnecessarily, used atoi() instead of strtol(), and didn't check the
4931 termination when getting file sizes from file names by regex.
4933 19. Completely re-implemented the support for maildirsize files, in the light
4934 of a number of problems with the previous contributed implementation
4935 (4.30/29). In particular:
4937 . If the quota is zero, the maildirsize file is maintained, but no quota is
4940 . If the maildir directory does not exist, it is created before any attempt
4941 to write a maildirsize file.
4943 . The quota value in the file is just a cache; if the quota is changed in
4944 the transport, the new value overrides.
4946 . A regular expression is available for excluding directories from the
4949 20. The autoreply transport checks the characters in options that define the
4950 message's headers; it allows continued headers, but it was checking with
4951 isspace() after an embedded newline instead of explicitly looking for a
4954 21. If all the "regular" hosts to which an address was routed had passed their
4955 expiry times, and had not reached their retry times, the address was
4956 bounced, even if fallback hosts were defined. Now Exim should go on to try
4959 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4960 equivalent code in the SMTP transport. Some hosts send humungous responses
4961 to HELO/EHLO, more than 1024 it seems.
4963 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4964 gives compiler warnings in some environments. I've now done it "properly",
4967 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4968 (because of problems with the built-in one) was declared to return uschar *
4969 instead of char *, causing compiler failure.
4971 25. Fixed a file descriptor leak when processing alias/forward files.
4973 26. Fixed a minor format string issue in dbfn.c.
4975 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4977 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4978 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4981 29. When a sender address is verified, it is cached, to save repeating the test
4982 when there is more than one recipient in a message. However, when the
4983 verification involves a callout, it is possible for different callout
4984 options to be set for different recipients. It is too complicated to keep
4985 track of this in the cache, so now Exim always runs a verification when a
4986 callout is required, relying on the callout cache for the optimization.
4987 The overhead is duplication of the address routing, but this should not be
4990 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4991 to be verified with callout=postmaster, and the main callout worked but the
4992 postmaster check failed, the verification correctly failed. However, if a
4993 subsequent RCPT command asked for sender verification *without* the
4994 postmaster check, incorrect caching caused this verification also to fail,
4997 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4998 it was not caching the DNS options (qualify_single, search_parents) that
4999 were used when the lookup failed. A subsequent lookup with different
5000 options therefore always gave the same answer, though there were cases
5001 where it should not have. (Example: a "domains = !$mx_any" option on a
5002 dnslookup router: the "domains" option is always processed without any
5003 widening, but the router might have qualify_single set.) Now Exim uses the
5004 cached value only when the same options are set.
5006 32. Added John Jetmore's "exipick" utility to the distribution.
5008 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5009 than a timeout (e.g. a certificate is required, and is not provided), an
5010 Exim server now closes the connection immediately. Previously it waited for
5011 the client to close - but if the client is SSL, it seems that they each
5012 wait for each other, leading to a delay before one of them times out.
5014 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5015 maintained 0.8.x compatibility because I don't think many are using it, and
5016 it is clearly obsolete.
5018 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5021 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5022 client certificate was expired. A simple patch fixes this, though I don't
5023 understand the full logic of why the verify callback is called multiple
5026 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5027 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5028 which causes problems with some clients (such as the Certicom SSL Plus
5029 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5030 disables the coutermeasure allowing Eudora to connect."
5032 38. Exim was not checking that a write() to a log file succeeded. This could
5033 lead to Bad Things if a log got too big, in particular if it hit a file
5034 size limit. Exim now panics and dies if it cannot write to a log file, just
5035 as it does if it cannot open a log file.
5037 39. Modified OS/Makefile-Linux so that it now contains
5039 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5041 The two -D definitions ensure that Exim is compiled with large file
5042 support, which makes it possible to handle log files that are bigger than
5045 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5046 instance) a domain was checked against a named list that involved a lookup,
5047 causing $domain_data to be set, then another domain was checked against the
5048 same list, then the first domain was re-checked, the value of $domain_data
5049 after the final check could be wrong. In particular, if the second check
5050 failed, it could be set empty. This bug probably also applied to
5053 41. The strip_trailing_dot option was not being applied to the address given
5054 with the -f command-line option.
5056 42. The code for reading a message's header from the spool was incrementing
5057 $received_count, but never initializing it. This meant that the value was
5058 incorrect (doubled) while delivering a message in the same process in which
5059 it was received. In the most common configuration of Exim, this never
5060 happens - a fresh exec is done - but it can happen when
5061 deliver_drop_privilege is set.
5063 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5064 it now includes up to 150 characters of the unexpected data in the log
5067 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5068 and building data strings. The size of both of these buffers was 10 000
5069 bytes - far larger than anybody would *ever* want, thought I. Needless to
5070 say, somebody hit the limit. I have increased the maximum line length to
5071 20 000 and the maximum data length of concatenated lines to 100 000. I have
5072 also fixed two bugs, because there was no checking on these buffers. Tsk,
5073 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5074 buffer is too small.
5076 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5077 lsearch lookups. Now it does.
5079 46. When parsing a route_list item in a manualroute router, a fixed-length
5080 buffer was used for the list of hosts. I made this 1024 bytes long,
5081 thinking that nobody would ever have a list of hosts that long. Wrong.
5082 Somebody had a whole pile of complicated expansion conditions, and the
5083 string was silently truncated, leading to an expansion error. It turns out
5084 that it is easier to change to an unlimited length (owing to other changes
5085 that have happened since this code was originally written) than to build
5086 structure for giving a limitation error. The length of the item that
5087 expands into the list of hosts is now unlimited.
5089 47. The lsearch lookup could not handle data where the length of text line was
5090 more than 4095 characters. Such lines were truncated, leading to shortened
5091 data being returned. It should now handle lines of any length.
5093 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5094 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5097 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5098 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5099 no longer happens in comment lines. A list of replacements is now placed
5100 at the head of all of the source files, except those whose only change is
5101 to replace PERL_COMMAND in the very first #! line.
5103 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5104 messages on the queue, with a bottom-up merge sort, using code contributed
5105 by Michael Haardt. This should make operations like -bp somewhat faster on
5106 large queues. It won't affect queue runners, except when queue_run_in_order
5109 51. Installed eximstats 1.31 in the distribution.
5111 52. Added support for SRV lookups to the dnslookup router.
5113 53. If an ACL referred to $message_body or $message_body_end, the value was not
5114 reset for any messages that followed in the same SMTP session.
5116 54. The store-handling optimization for building very long strings was not
5117 differentiating between the different store pools. I don't think this
5118 actually made any difference in practice, but I've tidied it.
5120 55. While running the routers to verify a sender address, $sender_address
5121 was still set to the sender address. This is wrong, because when routing to
5122 send a bounce to the sender, it would be empty. Therefore, I have changed
5123 it so that, while verifying a sender address, $sender_address is set to <>.
5124 (There is no change to what happens when verifying a recipient address.)
5126 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5127 target A or AAAA records (if not already returned) without resetting the
5128 qualify_single or search_parents options of the DNS resolver. These are
5129 inappropriate in this case because the targets of MX and SRV records must
5130 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5131 target that, when qualified, matched something in the local domain. These
5132 two options are now turned off when doing these lookups.
5134 57. It seems that at least some releases of Reiserfs (which does not have the
5135 concept of a fixed number of inodes) returns zero and not -1 for the
5136 number of available inodes. This interacted badly with check_spool_inodes,
5137 which assumed that -1 was the "no such thing" setting. What I have done is
5138 to check that the total number of inodes is greater than zero before doing
5139 the test of how many are available.
5141 58. When a "warn" ACL statement has a log_message modifier, the message is
5142 remembered, and not repeated. This is to avoid a lot of repetition when a
5143 message has many recipients that cause the same warning to be written.
5144 Howewer, Exim was preserving the list of already written lines for an
5145 entire SMTP session, which doesn't seem right. The memory is now reset if a
5146 new message is started.
5148 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5149 debugging output unless log_rewrite was also set.
5151 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5152 of (int)(handle) when we know that handle contains (void *)(-1).
5154 61. The Exim daemon panic-logs an error return when it closes the incoming
5155 connection. However "connection reset by peer" seems to be common, and
5156 isn't really an error worthy of noting specially, so that particular error
5159 62. When Exim is trying to find all the local interfaces, it used to panic and
5160 die if the ioctl to get the interface flags failed. However, it seems that
5161 on at least one OS (Solaris 9) it is possible to have an interface that is
5162 included in the list of interfaces, but for which you get a failure error
5163 for this call. This happens when the interface is not "plumbed" into a
5164 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5165 failure of the "get flags" call assumes that the interface is down.
5167 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5168 makes life easier for people who are doing arithmetic on fields extracted
5169 from dates, where you often get leading zeros that should not be
5170 interpreted as octal.
5172 64. Added qualify_domain to the redirect router, to override the global
5175 65. If a pathologically long header line contained very many addresses (the
5176 report of this problem mentioned 10 000) and each of them was rewritten,
5177 Exim could use up a very large amount of memory. (It kept on making new
5178 copies of the header line as it rewrote, and never released the old ones.)
5179 At the expense of a bit more processing, the header rewriting function has
5180 been changed so that it no longer eats memory in this way.
5182 66. The generation of the Received: header has been moved from the time that a
5183 message starts to be received, to the time that it finishes. The timestamp
5184 in the Received: header should now be very close to that of the <= log
5185 line. There are two side-effects of this change:
5187 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5188 logged header lines no longer include the local Received: line, because
5189 it has not yet been created. The same applies to a copy of the message
5190 that is returned to a non-SMTP sender when a message is rejected.
5192 (b) When a filter file is tested using -bf, no additional Received: header
5193 is added to the test message. After some thought, I decided that this
5196 This change does not affect the value of $received_for. It is still set
5197 after address rewriting, but before local_scan() is called.
5199 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5201 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5202 gave an unhelpful panic error message, and a defer error. I have managed to
5203 change this behaviour so that it now rejects any supplied certificate,
5204 which seems right, as the list of acceptable certificates is empty.
5206 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5207 gave an unhelpful defer error. I have not managed to make this reject any
5208 supplied certificates, but the error message it gives is "no certificate
5209 supplied", which is not helpful.
5211 70. exigrep's output now also includes lines that are not associated with any
5212 message, but which match the given pattern. Implemented by a patch from
5213 Martin Sluka, which also tidied up the Perl a bit.
5215 71. Recipient callout verification, like sender verification, was using <> in
5216 the MAIL FROM command. This isn't really the right thing, since the actual
5217 sender may affect whether the remote host accepts the recipient or not. I
5218 have changed it to use the actual sender in the callout; this means that
5219 the cache record is now keyed on a recipient/sender pair, not just the
5220 recipient address. There doesn't seem to be a real danger of callout loops,
5221 since a callout by the remote host to check the sender would use <>.
5222 [SEE ABOVE: changed after hitting problems.]
5224 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5225 temporary errors. However, in the case of such a code being given after
5226 the end of a data transmission (i.e. after ".") Exim was failing to write
5227 a retry record for the message. (Yes, there was some broken host that was
5228 actually sending 8xx at this point.)
5230 73. An unknown lookup type in a host list could cause Exim to panic-die when
5231 the list was checked. (An example that provoked this was putting <; in the
5232 middle of a list instead of at the start.) If this happened during a DATA
5233 ACL check, a -D file could be left lying around. This kind of configuration
5234 error no longer causes Exim to die; instead it causes a defer errror. The
5235 incident is still logged to the main and panic logs.
5237 74. Buglet left over from Exim 3 conversion. The message "too many messages
5238 in one connection" was written to the rejectlog but not the mainlog, except
5239 when address rewriting (yes!) was being logged.
5241 75. Added write_rejectlog option.
5243 76. When a system filter was run not as root (that is, when system_filter_user
5244 was set), the values of the $n variables were not being returned to the
5245 main process; thus, they were not subsequently available in the $sn
5248 77. Added +return_path_on_delivery log selector.
5250 78. A connection timeout was being treated differently from recipients deferred
5251 when testing hosts_max_try with a message that was older than the host's
5252 retry timeout. (The host should not be counted, thus allowing all hosts to
5253 be tried at least once before bouncing.) This may have been the cause of an
5254 occasionally reported bug whereby a message would remain on the queue
5255 longer than the retry timeout, but would be bounced if a delivery was
5256 forced. I say "may" because I never totally pinned down the problem;
5257 setting up timeout/retry tests is difficult. See also the next item.
5259 79. The ultimate address timeout was not being applied to errors that involved
5260 a combination of host plus message (for example, a timeout on a MAIL
5261 command). When an address resolved to a number of possible hosts, and they
5262 were not all tried for each delivery (e.g. because of hosts_max_try), a
5263 message could remain on the queue longer than the retry timeout.
5265 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5268 81. Fixed an obscure SMTP outgoing bug which required at least the following
5269 conditions: (a) there was another message waiting for the same server;
5270 (b) the server returned 5xx to all RCPT commands in the first message so
5271 that the message was not completed; (c) the server dropped the connection
5272 or gave a negative response to the RSET that Exim sends to abort the
5273 transaction. The observed case was a dropped connection after DATA that had
5274 been sent in pipelining mode. That is, the server had advertised PIPELINING
5275 but was not implementing it correctly. The effect of the bug was incorrect
5276 behaviour, such as trying another host, and this could lead to a crash.
5282 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5283 and daemon.c were passed as pointers to ints; they should have been
5284 pointers to socklen_t variables (which are typically unsigned ints).
5286 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5289 3. Fixed a really odd bug that affected only the testing scheme; patching a
5290 certain fixed string in the binary changed the value of another string that
5291 happened to be identical to the end of the original first string.
5293 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5294 name", it returns that address as the IP address. On some operating
5295 systems (e.g. Solaris), it also passes back the IP address string as the
5296 "host name". However, on others (e.g. Linux), it passes back an empty
5297 string. Exim wasn't checking for this, and was changing the host name to an
5298 empty string, assuming it had been canonicized.
5300 5. Although rare, it is permitted to have more than one PTR record for a given
5301 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5302 all the names associated with an address, because they do in Solaris.
5303 However, it seems that they do not in Linux for data that comes from the
5304 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5305 I found this out when I moved to a new Linux workstation and tried to run
5306 the Exim test suite.
5308 To get round this problem I have changed the code so that it now does its
5309 own call to the DNS to look up PTR records when searching for a host name.
5310 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5311 addresses that are only in /etc/hosts are still found.
5313 This behaviour is, however, controlled by an option called host_lookup_
5314 order, which defaults to "bydns:byaddr". If people want to use the other
5315 order, or indeed, just use one or the other means of lookup, they can
5316 specify it in this variable.
5318 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5319 some operating systems, this comes back from gethostbyaddr() as an empty
5320 string, and this is what Exim used to test for. However, it seems that in
5321 other systems, "." is yielded. Exim now tests for this case too.
5323 7. The values of check_spool_space and check_log_space are now held internally
5324 as a number of kilobytes instead of an absolute number of bytes. If a
5325 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5326 kilobyte. This means that much larger values can be stored.
5328 8. Exim monitor: an attempt to get the action menu when not actually pointing
5329 at a message produces an empty menu entitled "No message selected". This
5330 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5331 no entries in it ("Shell widget menu has zero width and/or height"). So I
5332 have added a single, blank menu entry in this case.
5334 9. Added ${quote_local_part.
5336 10. MIME decoding is now applied to the contents of Subject: header lines when
5339 11. Now that a reference to $sender_host_address automatically causes a reverse
5340 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5341 host lookup before query-style lookups in lists that might use this
5342 variable. This has therefore been abolished, and the "net-" prefix is no
5343 longer necessary for query-style lookups.
5345 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5346 have been a typo for LFLAGS, so it has been changed.
5348 13. The install script calls Exim with "-C /dev/null" in order to find the
5349 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5350 to be output. Howeve, since Exim outputs its version number before the
5351 error, it didn't break the script. It just looked ugly. I fixed this by
5352 always allowing "-C /dev/null" if the caller is root.
5354 14. Ignore overlarge ACL variable number when reading spool file - insurance
5355 against a later release with more variables having written the file.
5357 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5358 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5359 this, as well as the form without the "IPv6" on the front.
5361 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5362 OS/Makefile-Darwin file.
5364 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5365 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5367 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5368 variable. However, some OS use ints for inodes. I've added cast to long int
5369 to get rid of the compiler warning.
5371 19. I had forgotten to lock out "/../" in configuration file names when
5372 ALT_CONFIG_PREFIX was set.
5374 20. Routers used for verification do not need to specify transports. However,
5375 if such a router generated a host list, and callout was configured, Exim
5376 crashed, because it could not find a port number from the (non-existent)
5377 transport. It now assumes port 25 in this circumstance.
5379 21. Added the -t option to exigrep.
5381 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5382 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5383 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5384 them is set, LOOKUP_LSEARCH is forced.
5386 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5388 24. Added sender and host information to the "rejected by local_scan()" log
5389 line; previously there was no indication of these.
5391 25. Added .include_if_exists.
5393 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5394 Linux. It turns out that not all file systems support this. Apparently some
5395 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5396 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5397 not supported on the file descriptor, is now ignored when Exim is trying to
5398 sync a directory. This applies only to Linux.
5400 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5402 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5403 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5406 29. Added support for maildirsize files from supplied patch (modified a bit).
5408 30. The use of :fail: followed by an empty string could lead Exim to respond to
5409 sender verification failures with (e.g.):
5411 550 Verification failed for <xxx>
5412 550 Sender verify failed
5414 where the first response line was missing the '-' that indicates it is not
5415 the final line of the response.
5417 31. The loop for finding the name of the user that called Exim had a hardwired
5418 limit of 10; it now uses the value of finduser_retries, which is used for
5419 all other user lookups.
5421 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5424 33. Exim was neglecting to zero errno before one call of strtol() when
5425 expanding a string and expecting an integer value. On some systems this
5426 resulted in spurious "integer overflow" errors. Also, it was casting the
5427 result into an int without checking.
5429 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5430 did not work. The code looks as if it has *never* worked, though it appears
5431 to have been documented since at least releast 1.62. I have made it work.
5433 35. The "timeout_DNS" error in retry rules, also documented since at least
5434 1.62, also never worked. As it isn't clear exactly what this means, and
5435 clearly it isn't a major issue, I have abolished the feature by treating it
5436 as "timeout", and writing a warning to the main and panic logs.
5438 36. The display of retry rules for -brt wasn't always showing the error code
5441 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5442 timeout_connect_A, timeout_connect_MX.
5444 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5445 to the empty sender.
5447 39. The daemon was not analysing the content of -oX till after it had closed
5448 stderr and disconnected from the controlling terminal. This meant that any
5449 syntax errors were only noted on the panic log, and the return code from
5450 the command was 0. By re-arranging the code a little, I've made the
5451 decoding happen first, so such errors now appear on stderr, and the return
5452 code is 1. However, the actual setting up of the sockets still happens in
5453 the disconnected process, so errors there are still only recorded on the
5456 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5457 connections (as happens on some IP stacks) was logged at start up time as
5458 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5459 it from "IPv6 and IPv4", which means that two separate sockets are being
5462 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5463 says whether AF_INET or AF_INET6 was passed as an argument.
5465 42. Exiwhat output was messed up when time zones were included in log
5468 43. Exiwhat now gives more information about the daemon's listening ports,
5469 and whether -tls-on-connect was used.
5471 44. The "port" option of the smtp transport is now expanded.
5473 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5474 silently ignored. Now an error message is written to the main and panic
5477 46. There's a new ACL modifier called "logwrite" which writes to a log file
5478 as soon as it is encountered.
5480 47. Added $local_user_uid and $local_user_gid at routing time.
5482 48. Exim crashed when trying to verify a sender address that was being
5485 49. Exim was recognizing only a space character after ".include". It now also
5486 recognizes a tab character.
5488 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5489 extracting the relevant information from the specification. The man page no
5490 longer contains scrambled data for the -d option, and I've added a section
5491 at the front about calling Exim under different names.
5493 51. Added "extra_headers" argument to the "mail" command in filter files.
5495 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5498 53. Installed eximstats 1.29.
5500 54. Added transport_filter_timeout as a generic transport option.
5502 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5503 Cc: header lines. This was required by RFC 822, but it not required by RFC
5506 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5507 incoming messages that did not have them. Now it does so only if the
5508 message originates locally, that is, if there is no associated remote host
5509 address. When Resent- header lines are present, this applies to the Resent-
5510 lines rather than the non-Resent- lines.
5512 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5513 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5515 58. Messages for configuration errors now include the name of the main
5516 configuration file - useful now that there may be more than one file in a
5517 list (.included file names were always shown).
5519 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5520 for those rare installations that do not start the daemon as root or run it
5521 setuid root. I've cut out the call to initgroups() if the daemon is not
5524 60. The Exim user and group can now be bound into the binary as text strings
5525 that are looked up at the start of Exim's processing.
5527 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5529 62. Added $mailstore_basename variable.
5531 63. Installed patch to sieve.c from Michael Haardt.
5533 64. When Exim failed to open the panic log after failing to open the main log,
5534 the original message it was trying to log was written to stderr and debug
5535 output, but if they were not available (the usual case in production), it
5536 was lost. Now it is written to syslog before the two lines that record the
5537 failures to open the logs.
5539 65. Users' Exim filters run in subprocesses under the user's uid. It is
5540 possible for a "deliver" command or an alias in a "personal" command to
5541 provoke an address rewrite. If logging of address rewriting is configured,
5542 this fails because the process is not running as root or exim. There may be
5543 a better way of dealing with this, but for the moment (because 4.30 needs
5544 to be released), I have disabled address rewrite logging when running a
5545 filter in a non-root, non-exim process.
5551 1. The buildconfig auxiliary program wasn't quoting the value set for
5552 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5553 not defined. This bug was masked in 4.22 by the effect that was fixed in
5556 2. Some messages that were rejected after a message id was allocated were
5557 shown as "incomplete" by exigrep. It no longer does this for messages that
5558 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5560 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5561 have allow_domain_literals set, the ID did not get logged in the <= line.
5562 Domain literals are now always recognized in Message-ID: header lines.
5564 4. The first argument for a ${extract expansion item is the key name or field
5565 number. Leading and trailing spaces in this item were not being ignored,
5566 causing some misleading effects.
5568 5. When deliver_drop_privilege was set, single queue runner processes started
5569 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5570 same command in the process it spins off) were not dropping privilege.
5572 6. When the daemon running as "exim" started a queue runner, it always
5573 re-executed Exim in the spun-off process. This is a waste of effort when
5574 deliver_drop_privilege is set. The new process now just calls the
5575 queue-runner function directly.
5581 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5584 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5585 ignored. Though the use of -oP was forcing the writing of a pid file, it
5586 was always written to the default place.
5588 3. If the message "no IP address found for host xxxx" is generated during
5589 incoming verification, it is now followed by identification of the incoming
5590 connection (so you can more easily find what provoked it).
5592 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5594 5. Added some features to "harden" Exim a bit more against certain attacks:
5596 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5597 be put in Local/Makefile. This is like the never_users runtime option,
5598 but it cannot be overridden. The default setting is "root".
5600 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5601 prefix string with which any file named in a -C command line option
5604 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5605 is retained for -C and -D only if the caller of Exim is root. Without
5606 it, the exim user may also use -C and -D and retain privilege.
5608 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5609 command line option is disabled.
5611 6. Macro names set by the -D option must start with an upper case letter, just
5612 like macro names defined in the configuration file.
5614 7. Added "dereference=" facility to LDAP.
5616 8. Two instances of the typo "uknown" in the source files are fixed.
5618 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5619 the Configure-Makefile script screwed up while processing it.
5621 10. Incorporated PCRE 4.4.
5623 11. The SMTP synchronization check was not operating right at the start of an
5624 SMTP session. For example, it could not catch a HELO sent before the client
5625 waited for the greeting. There is now a check for outstanding input at the
5626 point when the greeting is written. Because of the duplex, asynchronous
5627 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5628 way, but not yet received, when the check is performed.
5630 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5631 on TCP/IP sockets, because this apparently causes some broken clients to
5634 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5635 unchanged) from the Cygwin maintainer.
5637 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5638 was supported instead of testing for mbx. Effectively a typo.
5640 15. The spa authenticator server code was not checking that the input it
5641 received was valid base64.
5643 16. The debug output line for the "set" modifier in ACLs was not showing the
5644 name of the variable that was being set.
5646 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5648 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5649 lookup on its own. Something else was needed to trigger the lookup. For
5650 example, a match in host_lookup or the need for a host name in a host list.
5651 Now, if $sender_host_name is referenced and the host name has not yet been
5652 looked up, a lookup is performed. If the lookup fails, the variable remains
5653 empty, and $host_lookup_failed is set to "1".
5655 19. Added "eqi" as a case-independent comparison operator.
5657 20. The saslauthd authentication condition could segfault if neither service
5658 nor realm was specified.
5660 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5661 error message that was logged was misleading, and incoming SMTP
5662 connections were dropped. The message is now more accurate, and temporary
5663 errors are given to SMTP connections.
5665 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5666 (except RSET) with a 421 error, until QUIT is received. However, it was
5667 failing to send a response to QUIT.
5669 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5670 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5671 ACL is now run after the helo_try_verify_hosts code.
5673 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5674 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5675 case-independent, so other case variants are also recognized). Apparently
5676 some systems use these upper case variants.
5678 25. If more than two messages were waiting for the same host, and a transport
5679 filter was specified for the transport, Exim sent two messages over the
5680 same TCP/IP connection, and then failed with "socket operation on non-
5681 socket" when it tried to send the third.
5683 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5685 27. The extern definition of crypt16() in expand.c was not being excluded when
5686 the OS had its own crypt16() function.
5688 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5689 as a preferred synonym for return_size_limit, both as an option and as an
5692 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5694 30. Changed the default configuration ACL to relax the local part checking rule
5695 for addresses that are not in any local domains. For these addresses,
5696 slashes and pipe symbols are allowed within local parts, but the sequence
5697 /../ is explicitly forbidden.
5699 31. SPA server authentication was not clearing the challenge buffer before
5702 32. log_message in a "warn" ACL statement was writing to the reject log as
5703 well as to the main log, which contradicts the documentation and doesn't
5704 seem right (because no rejection is happening). So I have stopped it.
5706 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5707 However, I am unable to do any testing of this.
5709 34. Fixed an infelicity in the appendfile transport. When checking directories
5710 for a mailbox, to see if any needed to be created, it was accidentally
5711 using path names with one or more superfluous leading slashes; tracing
5712 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5714 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5715 contents are added to the log line that is written for every discarded
5716 recipient. (Previously a log_message setting was ignored.)
5718 36. The ${quote: operator now quotes the string if it is empty.
5720 37. The install script runs exim in order to find its version number. If for
5721 some reason other than non-existence or emptiness, which it checks, it
5722 could not run './exim', it was installing it with an empty version number,
5723 i.e. as "exim-". This error state is now caught, and the installation is
5726 38. An argument was missing from the function that creates an error message
5727 when Exim fails to connect to the socket for saslauthd authentication.
5728 This could cause Exim to crash, or give a corrupted message.
5730 39. Added isip, isip4, and isip6 to ${if conditions.
5732 40. The ACL variables $acl_xx are now saved with the message, and can be
5733 accessed later in routers, transports, and filters.
5735 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5736 strings in the file are not string-expanded.
5738 42. If a MAIL command specified a SIZE value that was too large to fit into an
5739 int variable, the check against message_size_limit failed. Such values are
5740 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5741 day this will have to be increased, but I don't think I want to be around
5742 when emails are that large.
5749 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5750 iconv() is not standard in FreeBSD.
5752 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5753 IPv6 enabled. The observed symptom was a segmentation fault on return from
5754 the function os_common_find_running_interfaces() in src/os.c.
5756 3. In the check_special_case() function in daemon.c I had used "errno" as an
5757 argument name, which causes warnings on some systems. This was basically a
5758 typo, since it was named "eno" in the comments!
5760 4. The code that waits for the clock to tick (at a resolution of some fraction
5761 of a second) so as to ensure message-id uniqueness was always waiting for
5762 at least one whole tick, when it could have waited for less. [This is
5763 almost certainly not relevant at current processor speeds, where it is
5764 unlikely to ever wait at all. But we try to future-proof.]
5766 5. The function that sleeps for a time interval that includes fractions of a
5767 second contained a race. It did not block SIGALRM between setting the
5768 timer, and suspending (a couple of lines later). If the interval was short
5769 and the sigsuspend() was delayed until after it had expired, the suspension
5770 never ended. On busy systems this could lead to processes getting stuck for
5773 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5774 process, before it forks any delivery processes. The open lookup caching
5775 mechanism meant that the open file or database connection was passed into
5776 the delivery process. The problem was that delivery processes always tidy
5777 up cached lookup data. This could cause a problem for the next delivery
5778 process started by the queue runner, because the external queue runner
5779 process does not know about the closure. So the next delivery process
5780 still has data in the lookup cache. In the case of a file lookup, there was
5781 no problem because closing a file descriptor in a subprocess doesn't affect
5782 the parent. However, if the lookup was caching a connection to a database,
5783 the connection was closed, and the second delivery process was likely to
5784 see errors such as "PGSQL: query failed: server closed the connection
5785 unexpectedly". The problem has been fixed by closing all cached lookups
5786 in a queue runner before running a delivery process.
5788 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5789 seem to have the "const" qualifier which it has on other OS. I've
5792 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5793 *of the same type* (client or server) with the same public name that an
5794 error should be diagnosed.
5796 9. When Exim looked up a host name for an IP address, but failed to find the
5797 original IP address when looking up the host name (a safety check), it
5798 output the message "<ip address> does not match any IP for NULL", which was
5799 confusing, to say the least. The bug was that the host name should have
5800 appeared instead of "NULL".
5802 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5803 user that is built into the binary, and the -C or -D options is used, root
5804 privilege is dropped before the configuration file is read. In addition,
5805 logging is switched to stderr instead of the normal log files. If the
5806 configuration then re-defines the Exim user, the unprivileged environment
5807 is probably not what is expected, so Exim logs a panic warning message (but
5810 However, if deliver_drop_privilege is set, the unprivileged state may well
5811 be exactly what is intended, so the warning has been cut out in that case,
5812 and Exim is allowed to try to write to its normal log files.
5818 1. smtp_return_error_details was not giving details for temporary sender
5819 or receiver verification errors.
5821 2. Diagnose a configuration error if two authenticators have the same public
5824 3. Exim used not to create the message log file for a message until the first
5825 delivery attempt. This could be confusing when incoming messages were held
5826 for policy or load reasons. The message log file is now created at the time
5827 the message is received, and an initial "Received" line is written to it.
5829 4. The automatically generated man page for command line options had a minor
5830 bug that caused no ill effects; however, a more serious problem was that
5831 the procedure for building the man page automatically didn't always
5832 operate. Consequently, release 4.20 contains an out-of-date version. This
5833 shouldn't happen again.
5835 5. When building Exim with embedded Perl support, the script that builds the
5836 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5837 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5839 6. The freeze_tell option was not being used for messages that were frozen on
5840 arrival, either by an ACL or by local_scan().
5842 7. Added the smtp_incomplete_transaction log selector.
5844 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5845 was accepting AUTH without a new EHLO.
5847 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5848 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5851 10. Logging of TCP/IP connections (when configured) now happens in the main
5852 daemon process instead of the child process, so that the TCP/IP connection
5853 count is more accurate (but it can never be perfect).
5855 11. The use of "drop" in a nested ACL was not being handled correctly in the
5856 outer ACL. Now, if condition failure induced by the nested "drop" causes
5857 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5858 or "require"), the connection is dropped.
5860 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5861 that yield "discard" can now be used with an "accept" or a "discard" verb,
5862 but an error is generated for any others (because I can't see a useful way
5863 to define what should happen).
5865 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5866 are now processed in the same way as lines in the Exim configuration file.
5867 In particular, continuation lines are supported.
5869 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5871 15. Added -ti meaning -t -i.
5873 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5874 domains, and warn by logging if others are found.
5876 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5877 returned from GIFCONF when Exim is trying to find the list of interfaces on
5878 a host. The code in os.c has been modified to copy each ifreq to an aligned
5879 structure in all cases.
5881 Also, in some cases, the returned ifreq's were being copied to a 'struct
5882 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5883 means the last couple of bytes of an IPv6 address could be chopped if the
5884 ifreq contained only a normal sockaddr (14 bytes storage).
5886 18. Named domain lists were not supported in the hosts_treat_as_local option.
5887 An entry such as +xxxx was not recognized, and was treated as a literal
5890 19. Ensure that header lines added by a DATA ACL are included in the reject log
5891 if the ACL subsequently rejects the message.
5893 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5894 MD5 (which is deprecated).
5896 21. When testing a filter file using -bf, Exim was writing a message when it
5897 took the sender from a "From " line in the message, but it was not doing so
5898 when it took $return_path from a Return-Path: header line. It now does.
5900 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5901 with a valid header line field name (a series of printing characters
5902 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5904 23. Changed "disc" in the source to "disk" to conform to the documentation and
5905 the book and for uniformity.
5907 24. Ignore Sendmail's -Ooption=value command line item.
5909 25. When execve() failed while trying to run a command in a pipe transport,
5910 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5911 could be confused with a return value of 69 from the command itself. This
5912 has been changed to 127, the value the shell returns if it is asked to run
5913 a non-existent command. The wording for the related log line suggests a
5914 non-existent command as the problem.
5916 26. If received_header_text expands to an empty string, do not add a Received:
5917 header line to the message. (Well, it adds a token one on the spool, but
5918 marks it "old" so that it doesn't get used or transmitted.)
5920 27. Installed eximstats 1.28 (addition of -nt option).
5922 28. There was no check for failure on the call to getsockname() in the daemon
5923 code. This can fail if there is a shortage of resources on the system, with
5924 ENOMEM, for example. A temporary error is now given on failure.
5926 29. Contrary to the C standard, it seems that in some environments, the
5927 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5928 program. Exim now does this explicitly; it affects the formatting of
5929 timestamps using strftime().
5931 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5932 complaints. I've now initialized all the variables, to avoid this.
5934 32. Header lines added by a system filter were not being "seen" during
5935 transport-time rewrites.
5937 33. The info_callback() function passed to OpenSSL is set up with type void
5938 (*)(SSL *, int, int), as described somewhere. However, when calling the
5939 function (actually a macro) that sets it up, the type void(*)() is
5940 expected. I've put in a cast to prevent warnings from picky compilers.
5942 34. If a DNS black list lookup found a CNAME record, but there were no A
5943 records associated with the domain it pointed at, Exim crashed.
5945 35. If a DNS black list lookup returned more than one A record, Exim ignored
5946 all but the first. It now scans all returned addresses if a particular IP
5947 value is being sought. In this situation, the contents of the
5948 $dnslist_value variable are a list of all the addresses, separated by a
5951 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5952 to accept a host name and all its aliases if the forward lookup for any of
5953 them yielded the IP address of the incoming connection. Now it accepts only
5954 those names whose forward lookup yields the correct IP address. Any other
5955 names are discarded. This closes a loophole whereby a rogue DNS
5956 administrator could create reverse DNS records to break through a
5957 wildcarded host restriction in an ACL.
5959 37. If a user filter or a system filter that ran in a subprocess used any of
5960 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5961 the wrong values were passed to the pipe command ($thisaddress had the
5962 value of $0, $0 had the value of $1, etc). This bug was introduced by
5963 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5965 38. Improved the line breaking for long SMTP error messages from ACLs.
5966 Previously, if there was no break point between 40 and 75 characters, Exim
5967 left the rest of the message alone. Two changes have been made: (a) I've
5968 reduced the minimum length to 35 characters; (b) if it can't find a break
5969 point between 35 and 75 characters, it looks ahead and uses the first one
5970 that it finds. This may give the occasional overlong line, but at least the
5971 remaining text gets split now.
5973 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5974 file descriptors might be low, and that setting 1000 would always raise it.
5975 It turns out that in some environments, the limit is already over 1000 and
5976 that lowering it causes trouble. So now Exim takes care not to decrease it.
5978 40. When delivering a message, the value of $return_path is set to $sender_
5979 address at the start of routing (routers may change the value). By an
5980 oversight, this default was not being set up when an address was tested by
5981 -bt or -bv, which affected the outcome if any router or filter referred to
5984 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5985 log only when "message" or "log_message" are set. However, if one of the
5986 conditions was an address verification, or a call to a nested ACL, the
5987 messages generated by the underlying test were being passed through. This
5988 no longer happens. The underlying message is available in $acl_verify_
5989 message for both "message" and "log_message" expansions, so it can be
5990 passed through if needed.
5992 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5993 new expansion $bh_ to give the encoded byte string without charset
5994 translation. Translation happens only if iconv() is available; HAVE_ICONV
5995 indicates this at build time. HEADERS_CHARSET gives the charset to
5996 translate to; headers_charset can change it in the configuration, and
5997 "headers charset" can change it in an individual filter file.
5999 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6000 that creates RFC 2047 encoded "words" labels them as that charset instead
6001 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6002 expansion operator; (ii) when Exim creates a From: line for a local
6003 message; (iii) when a header line is rewritten to include a "phrase" part.
6005 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6006 buggy, causing it to skip the first lines of messages whose message ID
6007 ended in 'D'. This would not have bitten before Exim release 4.14, because
6008 message IDs were unlikely to end in 'D' before then. The effect was to have
6009 incorrect size information for certain domains.
6011 45. #include "config.h" was missing at the start of the crypt16.c module. This
6012 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6015 46. If there was a timeout during a "random" callout check, Exim treated it as
6016 a failure of the random address, and carried on sending RSET and the real
6017 address. If the delay was just some slowness somewhere, the response to the
6018 original RCPT would be taken as a response to RSET and so on, causing
6019 mayhem of various kinds.
6021 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6022 when I implemented it. It didn't allow for the fact that some option values
6023 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6024 the right test for positive values.
6026 48. Domain names in DNS records are case-independent. Exim always looks them up
6027 in lower case. Some resolvers return domain names in exactly the case they
6028 appear in the zone file, that is, they may contain uppercase letters. Not
6029 all resolvers do this - some return always lower case. Exim was treating a
6030 change of case by a resolver as a change of domain, similar to a widening
6031 of a domain abbreviation. This triggered its re-routing code and so it was
6032 trying to route what was effectively the same domain again. This normally
6033 caused routing to fail (because the router wouldn't handle the domain
6034 twice). Now Exim checks for this case specially, and just changes the
6035 casing of the domain that it ultimately uses when it transmits the message
6038 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6041 50. If a filter generated a file delivery with a non-absolute name (possible if
6042 no home directory exists for the router), the forbid_file option was not
6045 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6046 the existing equality matching.
6048 52. Exim was using ints instead of ino_t variables in some places where it was
6049 dealing with inode numbers.
6051 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6052 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6053 called TMPDIR, and if it finds it is different, it changes its value.
6055 54. The smtp_printf() function is now made available to local_scan() so
6056 additional output lines can be written before returning. There is also an
6057 smtp_fflush() function to enable the detection of a dropped connection.
6058 The variables smtp_input and smtp_batched_input are exported to
6061 55. Changed the default runtime configuration: the message "Unknown user"
6062 has been removed from the ACL, and instead placed on the localuser router,
6063 using the cannot_route_message feature. This means that any verification
6064 failures that generate their own messages won't get overridden. Similarly,
6065 the "Unrouteable address" message that was in the ACL for unverifiable
6066 relay addresses has also been removed.
6068 56. Added hosts_avoid_esmtp to the smtp transport.
6070 57. The exicyclog script was not checking for the esoteric option
6071 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6072 will work only if exicyclog is run under the appropriate euid.
6074 58. Following a discussion on the list, the rules by which Exim recognises line
6075 endings on incoming messages have been changed. The -dropcr and drop_cr
6076 options are now no-ops, retained only for backwards compatibility. The
6077 following line terminators are recognized: LF CRLF CR. However, special
6078 processing applies to CR:
6080 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6081 nor a local message in the state where . is a terminator.
6083 (ii) If a bare CR is encountered in a header line, an extra space is added
6084 after the line terminator so as not to end the header. The reasoning
6085 behind this is that bare CRs in header lines are most likely either
6086 to be mistakes, or people trying to play silly games.
6088 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6089 was being incorrectly given as 18 bytes larger than it should have been.
6090 This is a VOB (very old bug).
6092 60. This may never have affected anything current, but just in case it has:
6093 When the local host is found other than at the start of a list of hosts,
6094 the local host, those with the same MX, and any that follow, are discarded.
6095 When the list in question was part of a longer list of hosts, the following
6096 hosts (not currently being processed) were also being discarded. This no
6097 longer happens. I'm not sure if this situation could ever has previously
6100 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6103 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6104 header. This is something that is recommended in a new Internet Draft, and
6105 is something that is documented as being done by Sendmail. There are two
6106 possible values. For messages generated by the autoreply transport, Exim
6109 Auto-Submitted: auto-replied
6111 whereas for all other generated messages (e.g. bounces) it adds
6113 Auto-Submitted: auto-generated
6115 63. The "personal" condition in filters now includes a test for the
6116 Auto-Submitted: header. If it contains the string "auto-" the message it
6117 not considered personal.
6119 64. Added rcpt_include_affixes as a generic transport option.
6121 65. Added queue_only_override (default true).
6123 66. Added the syslog_duplication option.
6125 67. If what should have been the first header line of a message consisted of
6126 a space followed by a colon, Exim was mis-interpreting it as a header line.
6127 It isn't of course - it is syntactically invalid and should therefore be
6128 treated as the start of the message body. The misbehaviour could have
6129 caused a number of strange effects, including loss of data in subsequent
6130 header lines, and spool format errors.
6132 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6133 client host had authenticated. This control can now be exercised by an ACL
6134 for more flexibility.
6136 69. By default, callouts do not happen when testing with -bh. There is now a
6137 variant, -bhc, which does actually run the callout code, including
6138 consulting and updating the callout cache.
6140 70. Added support for saslauthd authentication, courtesy of Alexander
6143 71. If statvfs() failed on the spool or log directories while checking their
6144 size for availability, Exim confusingly gave the error "space shortage".
6145 Furthermore, in debugging mode it crashed with a floating point exception.
6146 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6147 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6148 serious problem, Exim now writes to the main and panic logs when this
6149 happens, with details of the failure. It then refuses to accept the
6150 incoming message, giving the message "spool directory problem" or "log
6151 directory problem" with a 421 code for SMTP messages.
6153 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6154 0, 1, and 2 exist, because some OS complain for execs without them (see
6155 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6156 descriptors. However, the code omitted to check that the open succeeded,
6157 causing mysterious errors if for some reason the permissions on /dev/null
6158 got screwed. Now Exim writes a message to the main and panic logs, and
6159 bombs out if it can't open /dev/null.
6161 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6162 interact so that it is all more flexible. It is supposed to remain
6163 backwards compatible. Also added extra_local_interfaces.
6165 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6166 to bomb out with an assertion failure - to the client this appears as a
6167 connection drop. This problem occurs in the part of the code that was taken
6168 from the Samba project. Fortunately, the assertion is in a very simple
6169 function, so I have fixed this by reproducing the function inline in the
6170 one place where it is called, and arranging for authentication to fail
6171 instead of killing the process with assert().
6173 75. The SPA client code was not working when the server requested OEM rather
6174 than Unicode encoding.
6176 76. Added code to make require_files with a specific uid setting more usable in
6177 the case where statting the file as root fails - usually a non-root-mounted
6178 NFS file system. When this happens and the failure is EACCES, Exim now
6179 forks a subprocess and does the per-uid checking as the relevant uid.
6181 77. Added process_log_path.
6183 78. If log_file_path was not explicitly set, a setting of check_log_space or
6184 check_log_inodes was ignored.
6186 79. If a space check for the spool or log partitions fails, the incident is now
6187 logged. Of course, in the latter case the data may get lost...
6189 80. Added the %p formatting code to string_format() so that it can be used to
6190 print addresses in debug_print(). Adjusted all the address printing in the
6191 debugging in store.c to use %p rather than %d.
6193 81. There was a concern that a line of code in smtp_in.c could overflow a
6194 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6195 initially expressed, the concern was not well-founded, because trailing
6196 spaces are removed early. However, if the trailing spaces were followed by
6197 a NULL, they did not get removed, so the overflow was possible. Two fixes
6200 (a) I re-wrote the offending code in a cleaner fashion.
6201 (b) If an incoming SMTP command contains a NULL character, it is rejected
6204 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6205 runs initgroups(), so that if the Exim user is in any additional groups,
6206 they will be used during message reception.
6212 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob