1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and backup-only
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
77 TL/01 When running the test suite, the README says that variables such as
78 no_msglog_check are global and can be placed anywhere in a specific
79 test's script, however it was observed that placement needed to be near
80 the beginning for it to behave that way. Changed the runtest perl
81 script to read through the entire script once to detect and set these
82 variables, reset to the beginning of the script, and then run through
83 the script parsing/test process like normal.
85 TL/02 The BSD's have an arc4random API. One of the functions to induce
86 adding randomness was arc4random_stir(), but it has been removed in
87 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
88 function when detected.
90 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
91 cause callback expansion.
93 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
94 syntax errors in an expansion can be treated as a string instead of
95 logging or causing an error, due to the internal use of bool_lax
96 instead of bool when processing it.
98 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
99 server certificates when making smtp deliveries.
101 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
103 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
105 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
107 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
108 Merged patch from Sebastian Wiedenroth.
110 JH/05 Fix results-pipe from transport process. Several recipients, combined
111 with certificate use, exposed issues where response data items split
112 over buffer boundaries were not parsed properly. This eventually
113 resulted in duplicates being sent. This issue only became common enough
114 to notice due to the introduction of conection certificate information,
115 the item size being so much larger. Found and fixed by Wolfgang Breyha.
117 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
118 size buffer was used, resulting in syntax errors when an expansion
121 JH/07 Add support for directories of certificates when compiled with a GnuTLS
122 version 3.3.6 or later.
124 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
125 is EXPERIMENTAL_EVENT, the main-configuration and transport options
126 both become "event_action", the variables become $event_name, $event_data
127 and $event_defer_errno. There is a new variable $verify_mode, usable in
128 routers, transports and related events. The tls:cert event is now also
129 raised for inbound connections, if the main configuration event_action
132 TL/06 In test suite, disable OCSP for old versions of openssl which contained
133 early OCSP support, but no stapling (appears to be less than 1.0.0).
135 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
136 server certificate names available under the smtp transport option
137 "tls_verify_cert_hostname" now do not permit multi-component wildcard
140 JH/10 Time-related extraction expansions from certificates now use the main
141 option "timezone" setting for output formatting, and are consistent
142 between OpenSSL and GnuTLS compilations. Bug 1541.
144 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
145 encoded parameter in the incoming message. Bug 1558.
147 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
148 include certificate info, eximon was claiming there were spoolfile
151 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
153 JH/14 Log delivery-related information more consistently, using the sequence
154 "H=<name> [<ip>]" wherever possible.
156 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
157 are problematic for Debian distribution, omit them from the release
160 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
162 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
165 JH/17 Fix a null-indirection in certextract expansions when a nondefault
166 output list separator was used.
171 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
172 checkers that were complaining about end of non-void function with no
175 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
176 This was a regression intruduced in 4.83 by another bugfix.
178 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
180 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
181 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
187 TF/01 Correctly close the server side of TLS when forking for delivery.
189 When a message was received over SMTP with TLS, Exim failed to clear up
190 the incoming connection properly after forking off the child process to
191 deliver the message. In some situations the subsequent outgoing
192 delivery connection happened to have the same fd number as the incoming
193 connection previously had. Exim would try to use TLS and fail, logging
194 a "Bad file descriptor" error.
196 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
197 utilities have not been installed.
199 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
200 temporary space as the ACL may create new global variables.
202 TL/01 LDAP support uses per connection or global context settings, depending
203 upon the detected version of the libraries at build time.
205 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
206 to extract and use the src ip:port in logging and expansions as if it
207 were a direct connection from the outside internet. PPv2 support was
208 updated based on HAProxy spec change in May 2014.
210 JH/02 Add ${listextract {number}{list}{success}{fail}}.
212 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
213 Properly escape header and check for NULL return.
215 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
218 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
220 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
221 characters in header names, implemented as a verify condition.
222 Contributed by Michael Fischer v. Mollard.
224 TL/05 Rename SPF condition results err_perm and err_temp to standardized
225 results permerror and temperror. Previous values are deprecated but
226 still accepted. In a future release, err_perm and err_temp will be
227 completely removed, which will be a backward incompatibility if the
228 ACL tests for either of these two old results. Patch contributed by
229 user bes-internal on the mailing list.
231 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
233 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
234 selectors, in both main and reject logs.
236 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
239 JH/07 Add malware type "sock" for talking to simple daemon.
241 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
243 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
244 routers/transports under cutthrough routing.
246 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
247 numbers. Touch up "bool" conditional to keep the same definition.
249 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
251 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
253 JH/12 Expand items in router/transport headers_add or headers_remove lists
254 individually rather than the list as a whole. Bug 1452.
256 Required for reasonable handling of multiple headers_ options when
257 they may be empty; requires that headers_remove items with embedded
258 colons must have them doubled (or the list-separator changed).
260 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
261 view the policy declared in the DMARC record. Currently, $dmarc_status
262 is a combined value of both the record presence and the result of the
265 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
267 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
268 dnslookup router and the smtp transport (applying to the forward
271 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
272 of ldap servers used for a specific lookup. Patch provided by Heiko
275 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
276 New variable $lookup_dnssec_authenticated for observability.
278 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
279 Patch submitted by Lars Timman.
281 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
283 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
284 Requires trusted mode and valid format message id, aborts otherwise.
285 Patch contributed by Heiko Schlichting.
287 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
288 certextract with support for various fields. Bug 1358.
290 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
291 is requested by default, modifiable by smtp transport option
294 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
295 operate on certificate variables to give certificate fingerprints
296 Also new ${sha256:cert_variable}.
298 JH/23 The PRDR feature is moved from being Experimental into the mainline.
300 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
301 Christian Aistleitner.
303 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
305 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
306 file. Patch from Wolfgang Breyha.
308 JH/25 Expand the coverage of the delivery $host and $host_address to
309 client authenticators run in verify callout. Bug 1476.
311 JH/26 Port service names are now accepted for tls_on_connect_ports, to
312 align with daemon_smtp_ports. Bug 72.
314 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
315 support and error reporting did not work properly.
317 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
318 and is readable. Patch from Andrew Colin Kissa.
320 TL/14 Enhance documentation of ${run expansion and how it parses the
321 commandline after expansion, particularly in the case when an
322 unquoted variable expansion results in an empty value.
324 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
326 PP/02 Fix internal collision of T_APL on systems which support RFC3123
327 by renaming away from it. Addresses GH issue 15, reported by
330 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
332 TL/15 SECURITY: prevent double expansion in math comparison functions
333 (can expand unsanitized data). Not remotely exploitable.
340 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
342 PP/02 Make -n do something, by making it not do something.
343 When combined with -bP, the name of an option is not output.
345 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
348 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
349 $sender_host_name and config options to manage this, and basic check
352 PP/05 DSCP support for outbound connections and control modifier for inbound.
354 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
355 (Only plugin which currently uses this is kerberos4, which nobody should
356 be using, but we should make it available and other future plugins might
357 conceivably use it, even though it would break NAT; stuff *should* be
358 using channel bindings instead).
360 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
361 name; added for Sendmail compatibility; requires admin caller.
362 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
363 just ignore it); requires trusted caller.
364 Also parse but ignore: -Ac -Am -X<logfile>
367 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
369 TL/02 Add +smtp_confirmation as a default logging option.
371 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
372 Patch by Magnus Holmgren from 2007-02-20.
374 TL/04 Bugzilla 1281 - Spec typo.
375 Bugzilla 1283 - Spec typo.
376 Bugzilla 1290 - Spec grammar fixes.
378 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
380 TL/06 Add Experimental DMARC support using libopendmarc libraries.
382 TL/07 Fix an out of order global option causing a segfault. Reported to dev
383 mailing list by by Dmitry Isaikin.
385 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
387 JH/02 Support "G" suffix to numbers in ${if comparisons.
389 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
391 NM/01 Bugzilla 1197 - Spec typo
392 Bugzilla 1196 - Spec examples corrections
394 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
396 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
397 gnutls_enable_pkcs11, but renamed to more accurately indicate its
400 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
401 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
403 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
404 "acl {{name}{arg}...}", and optional args on acl condition
407 JH/05 Permit multiple router/transport headers_add/remove lines.
409 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
411 JH/07 Avoid using a waiting database for a single-message-only transport.
412 Performance patch from Paul Fisher. Bugzilla 1262.
414 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
417 JH/09 Add $headers_added variable, with content from use of ACL modifier
418 add_header (but not yet added to the message). Bugzilla 199.
420 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
421 Pulled from Bugzilla 817 by Wolfgang Breyha.
423 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
425 (nb: this is the same fix as in Exim 4.80.1)
427 JH/11 Add A= logging on delivery lines, and a client_set_id option on
430 JH/12 Add optional authenticated_sender logging to A= and a log_selector
433 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
435 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
436 advertise SMTP AUTH mechanism to us, instead of a generic
437 protocol violation error. Also, make Exim more robust to bad
438 data from the Dovecot auth socket.
440 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
442 When a queue runner is handling a message, Exim first routes the
443 recipient addresses, during which it prunes them based on the retry
444 hints database. After that it attempts to deliver the message to
445 any remaining recipients. It then updates the hints database using
448 So if a recipient address works intermittently, it can get repeatedly
449 deferred at routing time. The retry hints record remains fresh so the
450 address never reaches the final cutoff time.
452 This is a fairly common occurrence when a user is bumping up against
453 their storage quota. Exim had some logic in its local delivery code
454 to deal with this. However it did not apply to per-recipient defers
455 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
457 This change adds a proper retry rule check during routing so that the
458 final cutoff time is checked against the message's age. We only do
459 this check if there is an address retry record and there is not a
460 domain retry record; this implies that previous attempts to handle
461 the address had the retry_use_local_parts option turned on. We use
462 this as an approximation for the destination being like a local
463 delivery, as in LMTP.
465 I suspect this new check makes the old local delivery cutoff check
466 redundant, but I have not verified this so I left the code in place.
468 TF/02 Correct gecos expansion when From: is a prefix of the username.
470 Test 0254 submits a message to Exim with the header
474 When I ran the test suite under the user fanf2, Exim expanded
475 the header to contain my full name, whereas it should have added
476 a Resent-Sender: header. It erroneously treats any prefix of the
477 username as equal to the username.
479 This change corrects that bug.
481 GF/01 DCC debug and logging tidyup
482 Error conditions log to paniclog rather than rejectlog.
483 Debug lines prefixed by "DCC: " to remove any ambiguity.
485 TF/03 Avoid unnecessary rebuilds of lookup-related code.
487 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
488 Bug spotted by Jeremy Harris; was flawed since initial commit.
489 Would have resulted in OCSP responses post-SNI triggering an Exim
490 NULL dereference and crash.
492 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
494 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
495 Bug detection, analysis and fix by Samuel Thibault.
496 Bugzilla 1331, Debian bug #698092.
498 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
500 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
501 Server implementation by Todd Lyons, client by JH.
502 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
503 config variable "prdr_enable" controls whether the server
504 advertises the facility. If the client requests PRDR a new
505 acl_data_smtp_prdr ACL is called once for each recipient, after
506 the body content is received and before the acl_smtp_data ACL.
507 The client is controlled by bolth of: a hosts_try_prdr option
508 on the smtp transport, and the server advertisement.
509 Default client logging of deliveries and rejections involving
510 PRDR are flagged with the string "PRDR".
512 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
513 fclose(). Diagnosis by Todd Lyons.
515 PP/17 Update configure.default to handle IPv6 localhost better.
516 Patch by Alain Williams (plus minor tweaks).
519 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
520 This is now consistent with GnuTLS, and is now documented: the
521 previous undocumented portable approach to treating the option as
522 unset was to force an expansion failure. That still works, and
523 an empty string is now equivalent.
525 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
526 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
527 not performing validation itself.
529 PP/20 Added force_command boolean option to pipe transport.
530 Patch from Nick Koston, of cPanel Inc.
532 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
535 TF/04 Added udpsend ACL modifer and hexquote expansion operator
537 PP/21 Fix eximon continuous updating with timestamped log-files.
538 Broken in a format-string cleanup in 4.80, missed when I repaired the
539 other false fix of the same issue.
540 Report and fix from Heiko Schlichting.
543 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
544 Report from Prashanth Katuri.
546 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
547 It's SecureTransport, so affects any MacOS clients which use the
548 system-integrated TLS libraries, including email clients.
550 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
551 using a MIME ACL for non-SMTP local injection.
552 Report and assistance in diagnosis by Warren Baker.
554 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
556 JH/16 Fix comparisons for 64b. Bugzilla 1385.
558 TL/09 Add expansion variable $authenticated_fail_id to keep track of
559 last id that failed so it may be referenced in subsequent ACL's.
561 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
564 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
565 ldap library initialization, allowing self-signed CA's to be
566 used. Also properly sets require_cert option later in code by
567 using NULL (global ldap config) instead of ldap handle (per
568 session). Bug diagnosis and testing by alxgomz.
570 TL/12 Enhanced documentation in the ratelimit.pl script provided in
571 the src/util/ subdirectory.
573 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
574 renamed to Transport Post Delivery Action by Jeremy Harris, as
577 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
578 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
579 redis_servers = needs to be configured which will be used by the redis
580 lookup. Patch from Warren Baker, of The Packet Hub.
582 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
584 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
585 hostname or reverse DNS when processing a host list. Used suggestions
586 from multiple comments on this bug.
588 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
590 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
591 Missed a few lines, added it to make the runtest require no keyboard
594 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
595 contains upper case chars. Make router use caseful_local_part.
597 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
598 support when GnuTLS has been built with p11-kit.
604 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
606 This, or similar/improved, will also be change PP/11 of 4.82.
612 PP/01 Handle short writes when writing local log-files.
613 In practice, only affects FreeBSD (8 onwards).
614 Bugzilla 1053, with thanks to Dmitry Isaikin.
616 NM/01 Bugzilla 949 - Documentation tweak
618 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
621 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
623 PP/02 Implemented gsasl authenticator.
625 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
627 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
628 `pkg-config foo` for cflags/libs.
630 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
631 with rest of GSASL and with heimdal_gssapi.
633 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
634 `pkg-config foo` for cflags/libs for the TLS implementation.
636 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
637 properties get this fed in as external SSF. A number of robustness
638 and debugging improvements to the cyrus_sasl authenticator.
640 PP/08 cyrus_sasl server now expands the server_realm option.
642 PP/09 Bugzilla 1214 - Log authentication information in reject log.
643 Patch by Jeremy Harris.
645 PP/10 Added dbmjz lookup type.
647 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
649 PP/12 MAIL args handles TAB as well as SP, for better interop with
650 non-compliant senders.
651 Analysis and variant patch by Todd Lyons.
653 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
654 Bug report from Lars Müller <lars@samba.org> (via SUSE),
655 Patch from Dirk Mueller <dmueller@suse.com>
657 PP/13 tls_peerdn now print-escaped for spool files.
658 Observed some $tls_peerdn in wild which contained \n, which resulted
659 in spool file corruption.
661 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
662 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
663 or write after TLS renegotiation, which otherwise led to messages
666 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
667 as a tracking header (ie: a signed header comes before the signature).
668 Patch from Wolfgang Breyha.
670 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
671 comma-sep list; embedded commas doubled.
673 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
675 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
677 Report and patch from Dmitry Banschikov.
679 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
680 Removed SSL_clear() after SSL_new() which led to protocol negotiation
681 failures. We appear to now support TLS1.1+ with Exim.
683 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
684 lets Exim select keys and certificates based upon TLS SNI from client.
685 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
686 before an outbound SMTP session. New log_selector, +tls_sni.
688 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
689 NULL dereference. Report and patch from Alun Jones.
691 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
692 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
693 Not seeing resolver debug output on NetBSD, but suspect this is a
694 resolver implementation change.
696 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
697 Left warnings. Added "eximon gdb" invocation mode.
699 PP/21 Defaulting "accept_8bitmime" to true, not false.
701 PP/22 Added -bw for inetd wait mode support.
703 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
704 locate the relevant includes and libraries. Made this the default.
706 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
707 Bugzilla 1246, report and most of solution from Tomasz Kusy.
709 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
710 This may cause build issues on older platforms.
712 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
713 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
714 gnutls_require_mac & gnutls_require_protocols (no longer supported).
715 Added SNI support via GnuTLS too.
716 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
718 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
720 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
721 Applied second patch from Janne, implementing suggestion to default
722 multiple-strings-in-record handling to match SPF spec.
724 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
726 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
727 read-only, out of scope).
728 Patch from Wolfgang Breyha, report from Stuart Northfield.
730 PP/29 Fix three issues highlighted by clang analyser static analysis.
731 Only crash-plausible issue would require the Cambridge-specific
732 iplookup router and a misconfiguration.
733 Report from Marcin Mirosław.
735 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
737 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
738 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
739 As part of this, removing so much warning spew let me fix some minor
740 real issues in debug logging.
742 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
743 assignment on my part. Fixed.
745 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
746 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
747 Janne Snabb (who went above and beyond: thank you).
749 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
750 string otherwise requires a connection and a bunch more work and it's
751 relatively easy to get wrong. Should also expose TLS library linkage
754 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
755 64-bit ${eval} (JH/03).
757 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
758 GNU libc to support some of the 64-bit stuff, should not lead to
759 conflicts. Defined before os.h is pulled in, so if a given platform
760 needs to override this, it can.
762 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
763 protection layer was required, which is not implemented.
764 Bugzilla 1254, patch from Wolfgang Breyha.
766 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
767 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
768 tls_dhparam take prime identifiers. Also unbreak combination of
769 OpenSSL+DH_params+TLSSNI.
771 PP/39 Disable SSLv2 by default in OpenSSL support.
777 PP/01 Solaris build fix for Oracle's LDAP libraries.
778 Bugzilla 1109, patch from Stephen Usher.
780 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
782 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
785 TF/02 Fix a couple more cases where we did not log the error message
786 when unlink() failed. See also change 4.74-TF/03.
788 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
789 lock up or crash if it happened to be inside a call to libc when it
790 got a SIGUSR1 from exiwhat.
792 The SIGUSR1 handler appends the current process status to the process
793 log which is later printed by exiwhat. It used to use the general
794 purpose logging code to do this, but several functions it calls are
795 not safe for signals.
797 The new output code in the SIGUSR1 handler is specific to the process
798 log, and simple enough that it's easy to inspect for signal safety.
799 Removing some special cases also simplifies the general logging code.
800 Removing the spurious timestamps from the process log simplifies
803 TF/04 Improved ratelimit ACL condition.
805 The /noupdate option has been deprecated in favour of /readonly which
806 has clearer semantics. The /leaky, /strict, and /readonly update modes
807 are mutually exclusive. The update mode is no longer included in the
808 database key; it just determines when the database is updated. (This
809 means that when you upgrde Exim will forget old rate measurements.)
811 Exim now checks that the per_* options are used with an update mode that
812 makes sense for the current ACL. For example, when Exim is processing a
813 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
814 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
815 must specify per_mail/readonly. If you omit the update mode it defaults to
816 /leaky where that makes sense (as before) or /readonly where required.
818 The /noupdate option is now undocumented but still supported for
819 backwards compatibility. It is equivalent to /readonly except that in
820 ACLs where /readonly is required you may specify /leaky/noupdate or
821 /strict/noupdate which are treated the same as /readonly.
823 A useful new feature is the /count= option. This is a generalization
824 of the per_byte option, so that you can measure the throughput of other
825 aggregate values. For example, the per_byte option is now equivalent
826 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
828 The per_rcpt option has been generalized using the /count= mechanism
829 (though it's more complicated than the per_byte equivalence). When it is
830 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
831 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
832 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
833 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
834 Note that using per_rcpt with a non-readonly update mode in more than
835 one ACL will cause the recipients to be double-counted. (The per_mail
836 and per_byte options don't have this problem.)
838 The handling of very low rates has changed slightly. If the computed rate
839 is less than the event's count (usually one) then this event is the first
840 after a long gap. In this case the rate is set to the same as this event's
841 count, so that the first message of a spam run is counted properly.
843 The major new feature is a mechanism for counting the rate of unique
844 events. The new per_addr option counts the number of different
845 recipients that someone has sent messages to in the last time period. It
846 behaves like per_rcpt if all the recipient addresses are different, but
847 duplicate recipient addresses do not increase the measured rate. Like
848 the /count= option this is a general mechanism, so the per_addr option
849 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
850 example, measure the rate that a client uses different sender addresses
851 with the options per_mail/unique=$sender_address. There are further
852 details in the main documentation.
854 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
856 TF/06 Removed a few PCRE remnants.
858 TF/07 Automatically extract Exim's version number from tags in the git
859 repository when doing development or release builds.
861 PP/02 Raise smtp_cmd_buffer_size to 16kB.
862 Bugzilla 879. Patch from Paul Fisher.
864 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
865 Heavily based on revision 40f9a89a from Simon Arlott's tree.
868 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
870 PP/05 Variable $av_failed, true if the AV scanner deferred.
871 Bugzilla 1078. Patch from John Horne.
873 PP/06 Stop make process more reliably on build failure.
874 Bugzilla 1087. Patch from Heiko Schlittermann.
876 PP/07 Make maildir_use_size_file an _expandable_ boolean.
877 Bugzilla 1089. Patch from Heiko Schlittermann.
879 PP/08 Handle ${run} returning more data than OS pipe buffer size.
880 Bugzilla 1131. Patch from Holger Weiß.
882 PP/09 Handle IPv6 addresses with SPF.
883 Bugzilla 860. Patch from Wolfgang Breyha.
885 PP/10 GnuTLS: support TLS 1.2 & 1.1.
887 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
890 PP/11 match_* no longer expand right-hand-side by default.
891 New compile-time build option, EXPAND_LISTMATCH_RHS.
892 New expansion conditions, "inlist", "inlisti".
894 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
896 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
898 PP/14 fix log_write() format string regression from TF/03.
899 Bugzilla 1152. Patch from Dmitry Isaikin.
905 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
907 PP/02 Harmonised TLS library version reporting; only show if debugging.
908 Layout now matches that introduced for other libraries in 4.74 PP/03.
910 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
912 PP/04 New "dns_use_edns0" global option.
914 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
917 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
918 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
920 TK/01 Updated PolarSSL code to 0.14.2.
921 Bugzilla 1097. Patch from Andreas Metzler.
923 PP/07 Catch divide-by-zero in ${eval:...}.
926 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
929 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
930 format-string attack -- SECURITY: remote arbitrary code execution.
932 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
933 time unintentionally subject to list matching rules, letting the header
934 cause arbitrary Exim lookups (of items which can occur in lists, *not*
935 arbitrary string expansion). This allowed for information disclosure.
937 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
938 INT_MIN/-1 -- value coerced to INT_MAX.
944 NM/01 Workround for PCRE version dependency in version reporting
947 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
948 This fixes portability to compilers other than gcc, notably
949 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
951 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
952 makefiles for portability to HP-UX and POSIX correctness.
954 PP/01 Permit LOOKUP_foo enabling on the make command-line.
955 Also via indented variable definition in the Makefile.
956 (Debugging by Oliver Heesakkers).
958 PP/02 Restore caching of spamd results with expanded spamd_address.
959 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
961 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
962 Improves build reliability. Fix from: Frank Elsner
964 NM/02 Fix wide character breakage in the rfc2047 coding
965 Fixes bug 1064. Patch from Andrey N. Oktyabrski
967 NM/03 Allow underscore in dnslist lookups
968 Fixes bug 1026. Patch from Graeme Fowler
970 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
971 Code patches from Adam Ciarcinski of NetBSD.
973 NM/04 Fixed exiqgrep to cope with mailq missing size issue
976 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
977 is logged, to avoid truncation. Patch from John Horne.
979 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
980 Patch from Jakob Hirsch.
982 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
983 SQL string expansion failure details.
984 Patch from Andrey Oktyabrski.
986 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
987 Patch from Simon Arlott.
989 PP/09 New lookups functionality failed to compile on old gcc which rejects
990 extern declarations in function scope.
991 Patch from Oliver Fleischmann
993 PP/10 Use sig_atomic_t for flags set from signal handlers.
994 Check getgroups() return and improve debugging.
995 Fixed developed for diagnosis in bug 927 (which turned out to be
998 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
999 Patch from Mark Zealey.
1001 PP/12 Bugzilla 1056: Improved spamd server selection.
1002 Patch from Mark Zealey.
1004 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1005 Based on patch from Heiko Schlittermann.
1007 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1008 Patch from Uwe Doering, sign-off by Michael Haardt.
1010 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1011 variable declaration deep within a block. Bug and patch from
1014 PP/15 lookups-Makefile IRIX compatibilty coercion.
1016 PP/16 Make DISABLE_DKIM build knob functional.
1018 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1019 Patch by Simon Arlott
1021 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1022 variable argument macros. Our copy now differs from upstream.
1028 TF/01 Failure to get a lock on a hints database can have serious
1029 consequences so log it to the panic log.
1031 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1032 controlled using the smtp_confirmation log selector.
1034 TF/03 Include the error message when we fail to unlink a spool file.
1036 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1037 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1038 for maintaining out-of-tree patches for some time.
1040 PP/01 Bugzilla 139: Documentation and portability issues.
1041 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1042 Handle per-OS dynamic-module compilation flags.
1044 PP/02 Let /dev/null have normal permissions.
1045 The 4.73 fixes were a little too stringent and complained about the
1046 permissions on /dev/null. Exempt it from some checks.
1047 Reported by Andreas M. Kirchwitz.
1049 PP/03 Report version information for many libraries, including
1050 Exim version information for dynamically loaded libraries. Created
1051 version.h, now support a version extension string for distributors
1052 who patch heavily. Dynamic module ABI change.
1054 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1055 privilege escalation vulnerability whereby the Exim run-time user
1056 can cause root to append content of the attacker's choosing to
1059 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1062 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1063 If dropping privileges for untrusted macros, we disabled normal logging
1064 on the basis that it would fail; for the Exim run-time user, this is not
1065 the case, and it resulted in successful deliveries going unlogged.
1066 Fixed. Reported by Andreas Metzler.
1072 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1073 only prepend for the Resent-* case. Fixes regression introduced in
1074 Exim 4.70 by NM/22 for Bugzilla 607.
1076 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1077 increasing numbers of administrators be bitten by this.
1079 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1081 PP/03 Bugzilla 994: added openssl_options main configuration option.
1083 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1085 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1087 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1089 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1090 without a peer certificate, leading to a segfault because of an
1091 assumption that peers always have certificates. Be a little more
1092 paranoid. Problem reported by Martin Tscholak.
1094 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1095 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1096 NB: ClamAV planning to remove STREAM in "middle of 2010".
1097 CL also introduces -bmalware, various -d+acl logging additions and
1098 more caution in buffer sizes.
1100 PP/09 Implemented reverse_ip expansion operator.
1102 PP/10 Bugzilla 937: provide a "debug" ACL control.
1104 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1106 PP/12 Bugzilla 973: Implement --version.
1108 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1110 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1112 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1114 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1115 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1116 ignore trailing whitespace.
1118 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1120 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1123 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1124 Notification from Dr Andrew Aitchison.
1126 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1127 ExtendedDetectionInfo response format.
1128 Notification from John Horne.
1130 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1133 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1134 XSL and documented dependency on system catalogs, with examples of how
1137 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1140 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1141 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1142 configuration file which is writeable by the Exim user or group.
1144 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1145 of configuration files to cover files specified with the -C option if
1146 they are going to be used with root privileges, not just the default
1149 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1150 option (effectively making it always true).
1152 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1153 files to be used while preserving root privileges.
1155 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1156 that rogue child processes cannot use them.
1158 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1159 run-time user, instead of root.
1161 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1162 Exim run-time user without dropping privileges.
1164 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1165 result string, instead of calling string_vformat() twice with the same
1168 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1169 for other users. Others should always drop root privileges if they use
1170 -C on the command line, even for a whitelisted configure file.
1172 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1174 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1180 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1181 $data_path, and $header_path variables; fixed documentation bugs and
1184 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1185 exipick to access non-standard spools, including the "frozen" queue
1188 NM/01 Bugzilla 965: Support mysql stored procedures.
1189 Patch from Alain Williams
1191 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1193 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1194 Patch from Andreas Metzler
1196 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1197 Patch from Kirill Miazine
1199 NM/05 Bugzilla 671: Added umask to procmail example.
1201 JJ/03 installed exipick 20100323.0, fixing doc bug
1203 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1204 directory. Notification and patch from Dan Rosenberg.
1206 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1208 TK/02 Improve log output when DKIM signing operation fails.
1210 MH/01 Treat the transport option dkim_domain as a colon separated
1211 list, not as a single string, and sign the message with each element,
1212 omitting multiple occurences of the same signer.
1214 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1215 Bugzilla 985, 986. Patch by Simon Arlott
1217 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1218 Patch by Simon Arlott
1220 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1221 MBX locking. Notification from Dan Rosenberg.
1227 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1229 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1231 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1233 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1235 NM/04 Bugzilla 915: Flag broken perl installation during build.
1241 TK/01 Added patch by Johannes Berg that expands the main option
1242 "spamd_address" if it starts with a dollar sign.
1244 TK/02 Write list of recipients to X-Envelope-Sender header when building
1245 the mbox-format spool file for content scanning (suggested by Jakob
1248 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1249 (http://www.dcc-servers.net/) support via dccifd. Activated by
1250 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1252 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1253 by Mark Daniel Reidel <mr@df.eu>.
1255 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1256 When building exim an external PCRE library is now needed -
1257 PCRE is a system library on the majority of modern systems.
1258 See entry on PCRE_LIBS in EDITME file.
1260 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1261 conversation. Added nologin parameter to request.
1262 Patch contributed by Kirill Miazine.
1264 TF/01 Do not log submission mode rewrites if they do not change the address.
1266 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1268 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1269 log files in place. Contributed by Roberto Lima.
1271 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1273 TF/03 Bugzilla 615: When checking the local_parts router precondition
1274 after a local_part_suffix or local_part_prefix option, Exim now
1275 does not use the address's named list lookup cache, since this
1276 contains cached lookups for the whole local part.
1278 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1279 Robert Millan. Documentation is in experimental-spec.txt.
1281 TF/04 Bugzilla 668: Fix parallel build (make -j).
1283 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1285 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1286 Patch provided by Jan Srzednicki.
1288 TF/05 Leading white space used to be stripped from $spam_report which
1289 wrecked the formatting. Now it is preserved.
1291 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1292 that they are available at delivery time.
1294 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1296 TF/08 TLS error reporting now respects the incoming_interface and
1297 incoming_port log selectors.
1299 TF/09 Produce a more useful error message if an SMTP transport's hosts
1300 setting expands to an empty string.
1302 NM/06 Bugzilla 744: EXPN did not work under TLS.
1303 Patch provided by Phil Pennock.
1305 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1306 Patch provided by Richard Godbee.
1308 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1309 acl_smtp_notquit, added index entry.
1311 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1312 Patch provided by Eugene Bujak.
1314 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1315 accept(). Patch provided by Maxim Dounin.
1317 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1318 Patch provided by Phil Pennock.
1320 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1322 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1323 Patch provided by Brad "anomie" Jorsch.
1325 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1326 Patch provided by Dean Brooks.
1328 TK/05 Add native DKIM support (does not depend on external libraries).
1330 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1331 Patch provided by Graeme Fowler.
1333 NM/16 Bugzilla 851: Documentation example syntax fix.
1335 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1337 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1340 NM/19 Bugzilla 745: TLS version reporting.
1341 Patch provided by Phil Pennock.
1343 NM/20 Bugzilla 167: bool: condition support.
1344 Patch provided by Phil Pennock.
1346 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1347 clients. Patch provided by Phil Pennock.
1349 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1350 Patch provided by Brad "anomie" Jorsch.
1352 NM/23 Bugzilla 687: Fix misparses in eximstats.
1353 Patch provided by Heiko Schlittermann.
1355 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1356 Patch provided by Heiko Schlittermann.
1358 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1359 plus update to original patch.
1361 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1363 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1364 Patch provided by David Brownlee.
1366 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1368 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1370 NM/30 Bugzilla 888: TLS documentation bugfixes.
1372 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1374 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1375 Unlike the original bugzilla I have changed all shell scripts in src tree.
1377 NM/33 Bugzilla 898: Transport filter timeout fix.
1378 Patch by Todd Rinaldo.
1380 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1381 Patch by Serge Demonchaux.
1383 NM/35 Bugzilla 39: Base64 decode bug fixes.
1384 Patch by Jakob Hirsch.
1386 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1388 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1390 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1392 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1398 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1399 ALT-N's libdkim that I have put here:
1400 http://duncanthrax.net/exim-experimental/
1402 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1403 were called 'true' and it seems that C99 defines that as a reserved
1404 keyword to be used with 'bool' variable types. That means you could
1405 not include C99-style headers which use bools without triggering
1406 build errors in sieve.c.
1408 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1409 as mailq or other aliases. Changed the --help handling significantly
1410 to do whats expected. exim_usage() emits usage/help information.
1412 SC/01 Added the -bylocaldomain option to eximstats.
1414 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1416 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1418 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1424 PH/01 Another patch from the Sieve maintainer.
1426 PH/02 When an IPv6 address is converted to a string for single-key lookup
1427 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1428 dots are used instead of colons so that keys in lsearch files need not
1429 contain colons. This was done some time before quoting was made available
1430 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1431 (notated using the quote facility) so as to distinguish them from IPv4
1432 keys. This meant that lookups for IP addresses in host lists did not work
1433 for iplsearch lookups.
1435 This has been fixed by arranging for IPv6 addresses to be expressed with
1436 colons if the lookup type is iplsearch. This is not incompatible, because
1437 previously such lookups could never work.
1439 The situation is now rather anomolous, since one *can* have colons in
1440 ordinary lsearch keys. However, making the change in all cases is
1441 incompatible and would probably break a number of configurations.
1443 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1446 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1447 conversion specification without a maximum field width, thereby enabling
1448 a rogue spamd server to cause a buffer overflow. While nobody in their
1449 right mind would setup Exim to query an untrusted spamd server, an
1450 attacker that gains access to a server running spamd could potentially
1451 exploit this vulnerability to run arbitrary code as the Exim user.
1453 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1454 $primary_hostname instead of what libspf2 thinks the hosts name is.
1456 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1457 a directory entry by the name of the lookup key. Previously, if a
1458 symlink pointed to a non-existing file or a file in a directory that
1459 Exim lacked permissions to read, a lookup for a key matching that
1460 symlink would fail. Now it is enough that a matching directory entry
1461 exists, symlink or not. (Bugzilla 503.)
1463 PH/03 The body_linecount and body_zerocount variables are now exported in the
1466 PH/04 Added the $dnslist_matched variable.
1468 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1469 This means they are set thereafter only if the connection becomes
1472 PH/06 Added the client_condition to authenticators so that some can be skipped
1473 by clients under certain conditions.
1475 PH/07 The error message for a badly-placed control=no_multiline_responses left
1476 "_responses" off the end of the name.
1478 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1480 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1481 (without spaces) instead of just copying the configuration text.
1483 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1485 PH/11 Added $max_received_linelength.
1487 PH/12 Added +ignore_defer and +include_defer to host lists.
1489 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1490 way in which PCRE > 7.0 is built.
1492 PH/14 Implemented queue_only_load_latch.
1494 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1495 MAIL command. The effect was to mangle the value on 64-bit systems.
1497 PH/16 Another patch from the Sieve maintainer.
1499 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1501 PH/18 If a system quota error occurred while trying to create the file for
1502 a maildir delivery, the message "Mailbox is full" was not appended to the
1503 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1504 only to a quota excession during the actual writing of the file.
1506 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1507 characters?) which causes problems in log lines. The DN values are now
1508 passed through string_printing() before being added to log lines.
1510 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1511 and InterBase are left for another time.)
1513 PH/21 Added message_body_newlines option.
1515 PH/22 Guard against possible overflow in moan_check_errorcopy().
1517 PH/23 POSIX allows open() to be a macro; guard against that.
1519 PH/24 If the recipient of an error message contained an @ in the local part
1520 (suitably quoted, of course), incorrect values were put in $domain and
1521 $local_part during the evaluation of errors_copy.
1527 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1528 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1531 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1532 issue a MAIL command.
1534 PH/02 In an ACL statement such as
1536 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1538 if a client was not listed at all, or was listed with a value other than
1539 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1540 the condition was not true (as it should be), so access was not denied.
1541 The bug was that the ! inversion was incorrectly passed on to the second
1542 item. This has been fixed.
1544 PH/03 Added additional dnslists conditions == and =& which are different from
1545 = and & when the dns lookup returns more than one IP address.
1547 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1548 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1550 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1551 FSYNC, which compiles an option called disable_fsync that allows for
1552 bypassing fsync(). The documentation is heavily laced with warnings.
1554 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1556 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1557 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1558 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1559 including adding "make clean"; (3) Added -fPIC when compiling the test
1560 dynamically loaded module, to get rid of a warning.
1562 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1563 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1564 = 0s. The bug is otherwise harmless.
1566 PH/07 There was a bug in the dovecot authenticator such that the value of
1567 $auth1 could be overwritten, and so not correctly preserved, after a
1568 successful authentication. This usually meant that the value preserved by
1569 the server_setid option was incorrect.
1571 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1573 PH/09 Installed PCRE release 7.0.
1575 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1576 run for batched SMTP input. It is now run at the start of every message
1577 in the batch. While fixing this I discovered that the process information
1578 (output by running exiwhat) was not always getting set for -bs and -bS
1579 input. This is fixed, and it now also says "batched" for BSMTP.
1581 PH/11 Added control=no_pipelining.
1583 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1584 patch, slightly modified), and move the expansion of helo_data till after
1585 the connection is made in the smtp transport (so it can use these
1588 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1590 PH/14 Added log_selector = +pid.
1592 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1594 PH/16 Add ${if forany and ${if forall.
1596 PH/17 Added dsn_from option to vary the From: line in DSNs.
1598 PH/18 Flush SMTP output before performing a callout, unless control =
1599 no_callout_flush is set.
1601 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1602 was true (the default) a successful delivery failed to delete the retry
1603 item, thus causing premature timeout of the address. The bug is now
1606 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1608 PH/21 Long custom messages for fakedefer and fakereject are now split up
1609 into multiline reponses in the same way that messages for "deny" and
1610 other ACL rejections are.
1612 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1613 with slight modification.
1615 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1616 draft, changing the syntax and factoring some duplicate code".
1618 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1619 for deliveries of the second and subsequent messages over the same SMTP
1622 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1623 ${reduce, with only minor "tidies".
1625 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1627 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1628 expansion side effects.
1630 PH/27 When a message times out after an over-quota error from an Exim-imposed
1631 quota, the bounce message says "mailbox is full". This message was not
1632 being given when it was a system quota that was exceeded. It now should
1635 MH/03 Made $recipients available in local_scan(). local_scan() already has
1636 better access to the recipient list through recipients_list[], but
1637 $recipients can be useful in postmaster-provided expansion strings.
1639 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1640 in the case of a MAIL command with additional options following the
1641 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1642 were accidentally chopped off.
1644 PH/29 SMTP synchronization checks are implemented when a command is read -
1645 there is a check that no more input is waiting when there shouldn't be
1646 any. However, for some commands, a delay in an ACL can mean that it is
1647 some time before the response is written. In this time, more input might
1648 arrive, invalidly. So now there are extra checks after an ACL has run for
1649 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1650 pipelining has not been advertised.
1652 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1654 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1655 returned in the SMTP response when smtp_return_error_details was false.
1656 This has been fixed.
1658 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1659 instead of the C I/O that was originally supplied, because problems were
1660 reported on Solaris.
1662 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1663 Exim which did not show up earlier: it was assuming that a call to
1664 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1665 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1666 was a macro that became an assignment, so it seemed to work. This has
1667 changed to a proper function call with a void return, hence the compile
1668 error. Exim's code has been fixed.
1670 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1673 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1675 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1677 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1678 because it turns out to be too big or there is a timeout) there is no
1679 "Completed" line in the log. When some messages of this type were
1680 selected by exigrep, they were listed as "not completed". Others were
1681 picked up by some special patterns. I have improved the selection
1682 criteria to be more general.
1684 PH/38 The host_find_failed option in the manualroute router can now be set
1685 to "ignore", to completely ignore a host whose IP address cannot be
1686 found. If all hosts are ignored, the behaviour is controlled by the new
1687 host_all_ignored option.
1689 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1690 homing or because of multiple MX records with /mx) generated more than
1691 one IP address, and the following item turned out to be the local host,
1692 all the secondary addresses of the first item were incorrectly removed
1693 from the list, along with the local host and any following hosts (which
1694 is what is supposed to happen).
1696 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1697 whoever called Exim into the -H file. In the case of the daemon it was
1698 behaving confusingly. When first started, it used values for whoever
1699 started the daemon, but after a SIGHUP it used the Exim user (because it
1700 calls itself on a restart). I have changed the code so that it now always
1703 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1704 message are rejected with the same error (e.g. no authentication or bad
1705 sender address), and a DATA command is nevertheless sent (as can happen
1706 with PIPELINING or a stupid MUA), the error message that was given to the
1707 RCPT commands is included in the rejection of the DATA command. This is
1708 intended to be helpful for MUAs that show only the final error to their
1711 PH/42 Another patch from the Sieve maintainer.
1713 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1714 Eximstats - Fixed some broken HTML links and added missing column headers
1716 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1717 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1719 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1721 PH/43 Yet another patch from the Sieve maintainer.
1723 PH/44 I found a way to check for a TCP/IP connection going away before sending
1724 the response to the final '.' that terminates a message, but only in the
1725 case where the client has not sent further data following the '.'
1726 (unfortunately, this is allowed). However, in many cases there won't be
1727 any further data because there won't be any more messages to send. A call
1728 to select() can be used: if it shows that the input is "ready", there is
1729 either input waiting, or the socket has been closed. An attempt to read
1730 the next input character can distinguish the two cases. Previously, Exim
1731 would have sent an OK response which the client would never have see.
1732 This could lead to message repetition. This fix should cure that, at
1733 least in a lot of common cases.
1735 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1736 advertised in response to EHLO.
1742 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1743 fixed by 4.65/MH/01 (is this a record?) are fixed:
1745 (i) An empty string was always treated as zero by the numeric comparison
1746 operators. This behaviour has been restored.
1748 (ii) It is documented that the numeric comparison operators always treat
1749 their arguments as decimal numbers. This was broken in that numbers
1750 starting with 0 were being interpreted as octal.
1752 While fixing these problems I realized that there was another issue that
1753 hadn't been noticed. Values of message_size_limit (both the global option
1754 and the transport option) were treated as octal if they started with 0.
1755 The documentation was vague. These values are now always treated as
1756 decimal, and I will make that clear in the documentation.
1762 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1763 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1766 MH/01 Don't check that the operands of numeric comparison operators are
1767 integers when their expansion is in "skipping" mode (fixes bug
1768 introduced by 4.64-PH/07).
1770 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1771 child addresses, Exim now panics and dies. Previously, because the count
1772 is held in a short int, deliveries were likely to be lost. As such a
1773 large number of recipients for a single message is ridiculous
1774 (performance will be very, very poor), I have chosen to impose a limit
1775 rather than extend the field.
1781 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1782 leftover -K file (the existence of which was triggered by #402).
1783 While we were at it, introduced process PID as part of the -K
1784 filename. This should rule out race conditions when creating
1787 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1788 processing considerably. Previous code took too long for large mails,
1789 triggering a timeout which in turn triggers #401.
1791 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1792 in the DK code in transports.c. sendfile() is not really portable,
1793 hence the _LINUX specificness.
1795 TF/01 In the add_headers option to the mail command in an Exim filter,
1796 there was a bug that Exim would claim a syntax error in any
1797 header after the first one which had an odd number of characters
1800 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1801 callout verification, Exim cached a "reject" for the entire domain. This
1802 is correct for most verifications, but it is not correct for a recipient
1803 verification with use_sender or use_postmaster set, because in that case
1804 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1805 case of MAIL FROM:<> rejection from other early rejections (e.g.
1806 rejection of HELO). When verifying a recipient using a non-null MAIL
1807 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1808 Whatever the result of the callout, the value of the domain cache is
1809 left unchanged (for any other kind of callout, getting as far as trying
1810 RCPT means that the domain itself is ok).
1812 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1815 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1816 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1817 session, the daemon ignores ECONNECT errors and logs others; it now
1818 ignores EPIPE as well.
1820 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1821 (quoted-printable decoding).
1823 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1824 later the small subsequent patch to fix an introduced bug.
1826 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1828 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1830 PH/08 An error is now given if message_size_limit is specified negative.
1832 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1833 to be given (somewhat) arbitrary names.
1835 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1838 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1839 miscellaneous code fixes
1841 PH/10 Added the log_reject_target ACL modifier to specify where to log
1844 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1845 hostname. This is wrong, because it relates to the incoming message (and
1846 probably the interface on which it is arriving) and not to the outgoing
1847 callout (which could be using a different interface). This has been
1848 changed to use the value of the helo_data option from the smtp transport
1849 instead - this is what is used when a message is actually being sent. If
1850 there is no remote transport (possible with a router that sets up host
1851 addresses), $smtp_active_hostname is used.
1853 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1854 tweaks were necessary in order to get it to work (see also 21 below):
1855 (a) The code assumed that strncpy() returns a negative number on buffer
1856 overflow, which isn't the case. Replaced with Exim's string_format()
1858 (b) There were several signed/unsigned issues. I just did the minimum
1859 hacking in of casts. There is scope for a larger refactoring.
1860 (c) The code used strcasecmp() which is not a standard C function.
1861 Replaced with Exim's strcmpic() function.
1862 (d) The code set only $1; it now sets $auth1 as well.
1863 (e) A simple test gave the error "authentication client didn't specify
1864 service in request". It would seem that Dovecot has changed its
1865 interface. Fortunately there's a specification; I followed it and
1866 changed what the client sends and it appears to be working now.
1868 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1871 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1872 address A is aliased to B and C, where B exists and C does not. Without
1873 -v the output is "A verified" because verification stops after a
1874 successful redirection if more than one address is generated. However,
1875 with -v the child addresses are also verified. Exim was outputting "A
1876 failed to verify" and then showing the successful verification for C,
1877 with its parentage. It now outputs "B failed to verify", showing B's
1878 parentage before showing the successful verification of C.
1880 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1881 look up a TXT record in a specific list after matching in a combined
1884 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1885 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1886 they consult the DNS. I had assumed they would set it the way they
1887 wanted; and indeed my experiments on Linux seem to show that in some
1888 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1889 To be on the safe side, however, I have now made the interface to
1890 host_find_byname() similar to host_find_bydns(), with an argument
1891 containing the DNS resolver options. The host_find_byname() function now
1892 sets these options at its start, just as host_find_bydns() does. The smtp
1893 transport options dns_qualify_single and dns_search_parents are passed to
1894 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1895 of host_find_byname() use the default settings of RES_DEFNAMES
1896 (qualify_single) but not RES_DNSRCH (search_parents).
1898 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1899 spool_read_header() do less string testing, by means of a preliminary
1900 switch on the second character of optional "-foo" lines. (This is
1901 overdue, caused by the large number of possibilities that now exist.
1902 Originally there were few.) While I was there, I also converted the
1903 str(n)cmp tests so they don't re-test the leading "-" and the first
1904 character, in the hope this might squeeze out yet more improvement.
1906 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1907 flag allowing group syntax was set by the header_syntax check but not
1908 turned off, possible causing trouble later; (2) The flag was not being
1909 set at all for the header_verify test, causing "group"-style headers to
1910 be rejected. I have now set it in this case, and also caused header_
1911 verify to ignore an empty address taken from a group. While doing this, I
1912 came across some other cases where the code for allowing group syntax
1913 while scanning a header line wasn't quite right (mostly, not resetting
1914 the flag correctly in the right place). These bugs could have caused
1915 trouble for malformed header lines. I hope it is now all correct.
1917 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1918 with the "reply" argument non-NULL. The code, however (which originally
1919 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1920 but it didn't always do it. This confused somebody who was copying the
1921 code for some other use. I have removed all the tests.
1923 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1924 feature that was used to support insecure browsers during the U.S. crypto
1925 embargo. It requires special client support, and Exim is probably the
1926 only MTA that supported it -- and would never use it because real RSA is
1927 always available. This code has been removed, because it had the bad
1928 effect of slowing Exim down by computing (never used) parameters for the
1929 RSA_EXPORT functionality.
1931 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1932 authenticator to fail if there's a tab character in the incoming data
1933 (there should never be unless someone is messing about, as it's supposed
1934 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1935 if the connection is using TLS or if the remote IP is the same as the
1936 local IP, and the "valid-client-cert option" if a client certificate has
1939 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1940 authenticators. This can be used for authorization after authentication
1941 succeeds. (In the case of plaintext, it servers for both authentication
1944 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1945 if any retry times were supplied.
1947 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1948 connection, where there is no client IP address to check. In this
1949 situation, the verify now always succeeds.
1951 PH/25 Applied John Jetmore's -Mset patch.
1953 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1955 PH/27 In a string expansion for a processed (not raw) header when multiple
1956 headers of the same name were present, leading whitespace was being
1957 removed from all of them, but trailing whitespace was being removed only
1958 from the last one. Now trailing whitespace is removed from each header
1959 before concatenation. Completely empty headers in a concatenation (as
1960 before) are ignored.
1962 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1963 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1965 PH/29 [Removed. This was a change that I later backed out, and forgot to
1966 correct the ChangeLog entry (that I had efficiently created) before
1967 committing the later change.]
1969 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1970 address errors (4xx response to RCPT) over the same connection as other
1971 messages routed to the same hosts. Such deliveries are always "forced",
1972 so retry times are not inspected. This resulted in far too many retries
1973 for the affected addresses. The effect occurred only when there were more
1974 hosts than the hosts_max_try setting in the smtp transport when it had
1975 the 4xx errors. Those hosts that it had tried were not added to the list
1976 of hosts for which the message was waiting, so if all were tried, there
1977 was no problem. Two fixes have been applied:
1979 (i) If there are any address or message errors in an SMTP delivery, none
1980 of the hosts (tried or untried) are now added to the list of hosts
1981 for which the message is waiting, so the message should not be a
1982 candidate for sending over the same connection that was used for a
1983 successful delivery of some other message. This seems entirely
1984 reasonable: after all the message is NOT "waiting for some host".
1985 This is so "obvious" that I'm not sure why it wasn't done
1986 previously. Hope I haven't missed anything, but it can't do any
1987 harm, as the worst effect is to miss an optimization.
1989 (ii) If, despite (i), such a delivery is accidentally attempted, the
1990 routing retry time is respected, so at least it doesn't keep
1991 hammering the server.
1993 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1994 in ${readsocket because some servers need this prod.
1996 PH/32 Added some extra debug output when updating a wait-xxx database.
1998 PH/33 The hint "could be header name not terminated by colon", which has been
1999 given for certain expansion errors for a long time, was not being given
2000 for the ${if def:h_colon_omitted{... case.
2002 PH/34 The spec says: "With one important exception, whenever a domain list is
2003 being scanned, $domain contains the subject domain." There was at least
2004 one case where this was not true.
2006 PH/35 The error "getsockname() failed: connection reset by peer" was being
2007 written to the panic log as well as the main log, but it isn't really
2008 panic-worthy as it just means the connection died rather early on. I have
2009 removed the panic log writing for the ECONNRESET error when getsockname()
2012 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2013 runs only) independently of the message's sender address. This meant
2014 that, if the 4xx error was in fact related to the sender, a different
2015 message to the same recipient with a different sender could confuse
2016 things. In particualar, this can happen when sending to a greylisting
2017 server, but other circumstances could also provoke similar problems.
2018 I have changed the default so that the retry time for these errors is now
2019 based a combination of the sender and recipient addresses. This change
2020 can be overridden by setting address_retry_include_sender=false in the
2023 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2024 remote server are returned as part of bounce messages. This was not
2025 happening for LMTP over a pipe (the lmtp transport), but now it is the
2026 same for both kinds of LMTP.
2028 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2029 in header lines that were in fact CNAMEs. This is no longer the case.
2031 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2032 and queue runs started by the daemon processed all messages. This has
2033 been fixed so that -R and -S can now usefully be given with -q<time>.
2035 PH/40 Import PCRE release 6.7 (fixes some bugs).
2037 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2039 PH/42 Give an error if -q is specified more than once.
2041 PH/43 Renamed the variables $interface_address and $interface_port as
2042 $received_ip_address and $received_port, to make it clear that these
2043 values apply to message reception, and not to the outgoing interface when
2044 a message is delivered. (The old names remain recognized, of course.)
2046 PH/44 There was no timeout on the connect() call when using a Unix domain
2047 socket in the ${readsocket expansion. There now is.
2049 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2050 be meaningful with "accept".
2052 SC/01 Eximstats V1.43
2053 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2055 SC/02 Eximstats V1.44
2056 Use a glob alias rather than an array ref in the generated
2057 parser. This improves both readability and performance.
2059 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2060 Collect SpamAssassin and rejection statistics.
2061 Don't display local sender or destination tables unless
2062 there is data to show.
2063 Added average volumes into the top table text output.
2065 SC/04 Eximstats V1.46
2066 Collect data on the number of addresses (recipients)
2067 as well as the number of messages.
2069 SC/05 Eximstats V1.47
2070 Added 'Message too big' to the list of mail rejection
2071 reasons (thanks to Marco Gaiarin).
2073 SC/06 Eximstats V1.48
2074 Mainlog lines which have GMT offsets and are too short to
2075 have a flag are now skipped.
2077 SC/07 Eximstats V1.49 (Alain Williams)
2078 Added the -emptyok flag.
2080 SC/08 Eximstats V1.50
2081 Fixes for obtaining the IP address from reject messages.
2083 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2084 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2085 whitesspace changes from 4.64-PH/27
2087 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2090 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2091 are found, allow negative numbers in numeric criteria)
2093 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2095 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2096 to match changes made in 4.64-PH/43
2098 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2100 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2101 "domain" is actually an IP address, and force a failure. This locks out
2102 those revolvers/nameservers that support "A-for-A" lookups, in
2103 contravention of the specifications.
2105 PH/48 When a host name was looked up from an IP address, and the subsequent
2106 forward lookup of the name timed out, the host name was left in
2107 $sender_host_name, contrary to the specification.
2109 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2110 restricted to single-key lookups, Exim was not diagnosing an error if
2111 * or *@ was used with a query-style lookup.
2113 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2115 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2116 long ago, but noone interested enough thought of it. Let's just say that
2117 the "1.1" means that there are some new functions that weren't there at
2118 some point in the past.
2120 PH/51 Error processing for expansion failure of helo_data from an smtp
2121 transport during callout processing was broken.
2123 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2124 tested/used via the -bh/-bhc/-bs options.
2126 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2127 bug, fixed in subsequent PCRE releases).
2129 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2130 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2132 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2138 SC/01 Use a glob alias rather than an array ref in eximstats generated
2139 parser. This improves both readability and performance.
2141 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2142 Don't display local sender or destination tables in eximstats unless
2143 there is data to show.
2144 Added average volumes into the eximstats top table text output.
2146 SC/03 Collect data on the number of addresses (recipients) as well
2147 as the number of messages in eximstats.
2149 TF/01 Correct an error in the documentation for the redirect router. Exim
2150 does (usually) call initgroups() when daemonizing.
2152 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2153 with consistent privilege compared to when running as a daemon.
2155 TF/03 Note in the spec that $authenticated_id is not set for local
2156 submissions from trusted users.
2158 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2159 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2161 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2162 by adding some example configuration directives to the default
2163 configuration file. A little bit of work is required to uncomment the
2164 directives and define how usernames and passwords are checked, but
2165 there is now a framework to start from.
2167 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2168 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2169 without this. I don't know how relevant this is to other LDAP libraries.
2171 PH/02 Add the verb name to the "unknown ACL verb" error.
2173 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2175 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2177 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2178 directory not expanded when it should be if an expanded home directory
2179 was set for the address (which is overridden by the transport).
2181 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2184 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2185 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2186 because it is too late at that time, and has no effect.
2188 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2189 security issue with \' (bugzilla #107). I could not use the
2190 PQescapeStringConn() function, because it needs a PGconn value as one of
2193 PH/08 When testing addresses using -bt, indicate those final addresses that
2194 are duplicates that would not cause an additional delivery. At least one
2195 person was confused, thinking that -bt output corresponded to deliveries.
2196 (Suppressing duplicates isn't a good idea as you lose the information
2197 about possibly different redirections that led to the duplicates.)
2199 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2200 systems where poll() doesn't work, in particular OS X.
2202 PH/10 Added more information to debugging output for retry time not reached.
2204 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2205 operations in malware.c.
2207 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2210 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2211 syslog_duplication set false, log lines that would normally be written
2212 both the the main log and to the reject log were not written to syslog at
2215 PH/14 In the default configuration, change the use of "message" in ACL warn
2216 statements to "add_header".
2218 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2219 not followed by a command (e.g. "seen endif").
2221 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2222 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2225 PH/17 Added extra conditions to the default value of delay_warning_condition
2229 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2230 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2231 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2234 The Auto-Submitted: and various List- headers are standardised, whereas I
2235 don't think Precedence: ever was.
2237 PH/18 Refactored debugging code in route_finduser() to show more information,
2238 in particular, the error code if getpwnam() issues one.
2240 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2241 This is apparently needed in addition to the PH/07 change above to avoid
2242 any possible encoding problems.
2244 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2245 but not after initializing Perl.
2247 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2248 output them only if debugging. By default they are written stderr,
2249 apparently, which is not desirable.
2251 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2254 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2257 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2259 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2260 authenticated or an ident call has been made. Suppress the default
2261 values for $authenticated_id and $authenticated_sender (but permit -oMai
2262 and -oMas) when testing with -bh.
2264 PH/24 Re-jigged the order of the tests in the default configuration so that the
2265 tests for valid domains and recipients precede the DNS black list and CSA
2266 tests, on the grounds that those ones are more expensive.
2268 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2269 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2270 This bug exists in every version of Exim that I still have, right back to
2273 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2274 However, an attempt to turn on case-sensitivity in a regex key by
2275 including (?-i) didn't work because the subject string was already
2276 lowercased, and the effects were non-intuitive. It turns out that a
2277 one-line patch can be used to allow (?-i) to work as expected.
2283 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2284 other effects) broke the use of negated acl sub-conditions.
2286 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2289 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2290 "Deny" causes Exim to reject the incoming connection with a 554 error.
2291 Unfortunately, if there is a major crisis, such as a disk failure,
2292 tcp-wrappers gives "deny", whereas what one would like would be some
2293 kind of temporary error. A kludge has been added to help with this.
2294 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2295 554 error is used if errno is still zero or contains ENOENT (which occurs
2296 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2299 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2301 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2302 errors. Otherwise a message that provokes a temporary error (when other
2303 messages do not) can cause a whole host to time out.
2305 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2306 addresses were routed directly to files or pipes from a redirect router.
2307 File deliveries just didn't batch; pipe deliveries might have suffered
2310 PH/06 A failure to get a lock for a hints database would erroneously always say
2311 "Failed to get write lock", even when it was really a read lock.
2313 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2314 of 0600. This has been changed to use the value of the lockfile_mode
2315 option (which defaults to 0600).
2317 PH/08 Applied small patch from the Sieve maintainer.
2319 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2320 folder from quota calculations, a direct delivery into this folder messed
2321 up the contents of the maildirsize file. This was because the regex was
2322 used only to exclude .Trash (or whatever) when the size of the mailbox
2323 was calculated. There was no check that a delivery was happening into an
2324 excluded directory. This bug has been fixed by ignoring all quota
2325 processing for deliveries into excluded directories.
2327 PH/10 Added the maildirfolder_create_regex option to appendfile.
2333 PH/01 The code for finding all the local interface addresses on a FreeBSD
2334 system running IPv6 was broken. This may well have applied to all BSD
2335 systems, as well as to others that have similar system calls. The broken
2336 code found IPv4 interfaces correctly, but gave incorrect values for the
2337 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2338 that it would not match correctly against @[] and not recognize the IPv6
2341 PH/02 The ipliteral router was not recognizing addresses of the form user@
2342 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2344 PH/03 Added disable_ipv6.
2346 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2347 decoded form, because it is most often used to construct To: headers
2348 lines in autoreplies, and the decoded form may well be syntactically
2349 invalid. However, $reply_address has leading white space removed, and all
2350 newlines turned into spaces so that the autoreply transport does not
2353 PH/05 If group was specified without a user on a router, and no group or user
2354 was specified on a transport, the group from the router was ignored.
2356 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2357 for visible compile-time settings that can be used to change these
2358 numbers, for those that want even more. Backwards compatibility with old
2359 spool files has been maintained. However, going back to a previous Exim
2360 release will lost any variables that are in spool files.
2362 PH/07 Two small changes when running in the test harness: increase delay when
2363 passing a TCP/IP connection to a new process, in case the original
2364 process has to generate a bounce, and remove special handling of
2365 127.0.0.2 (sic), which is no longer necessary.
2367 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2368 be the same on different OS.
2370 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2373 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2374 whether --show-vars was specified or not
2376 JJ/02 exipick: Added support for new ACL variable spool format introduced
2379 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2380 syntactically invalid From: or Reply-to: line, and a filter used this to
2381 generate an autoreply, and therefore failed to obtain an address for the
2382 autoreply, Exim could try to deliver to a non-existent relative file
2383 name, causing unrelated and misleading errors. What now happens is that
2384 it logs this as a hard delivery error, but does not attempt to create a
2387 PH/11 The exinext utility has a -C option for testing purposes, but although
2388 the given file was scanned by exinext itself; it wasn't being passed on
2389 when Exim was called.
2391 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2392 an end-of-file indication when reading a command response.
2394 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2395 compiled. In many other places in Exim, IPv6 addresses are always
2396 recognized, so I have changed this. It also means that IPv4 domain
2397 literals of the form [IPV4:n.n.n.n] are now always recognized.
2399 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2400 used if the router is not running as root, for example, when verifying at
2401 ACL time, or when using -bh. The debugging output from this situation was
2402 non-existent - all you got was a failure to exec. I have made two
2405 (a) Failures to set uid/gid, the current directory, or a process leader
2406 in a subprocess such as that created by queryprogram now generate
2407 suitable debugging ouput when -d is set.
2409 (b) The queryprogram router detects when it is not running as root,
2410 outputs suitable debugging information if -d is set, and then runs
2411 the subprocess without attempting to change uid/gid.
2413 PH/15 Minor change to Makefile for building test_host (undocumented testing
2416 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2417 additional section of a DNS packet that returns MX or SRV records.
2418 Instead, it always explicitly searches for A/AAAA records. This avoids
2419 major problems that occur when a DNS server includes only records of one
2420 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2421 fixed another bug: if SRV records were looked up and the corresponding
2422 address records were *not* found in the additional section, the port
2423 values from the SRV records were lost.
2425 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2426 using the correct key (the original address) when searching the retry
2427 rules in order to find which one to use for generating the retry hint.
2429 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2430 adding the default one. Similarly, if it contains a Reply-To: header, the
2431 errors_reply_to option, if set, is not used.
2433 PH/19 When calculating a retry time, Exim used to measure the "time since
2434 failure" by looking at the "first failed" field in the retry record. Now
2435 it does not use this if it is later than than the arrival time of the
2436 message. Instead it uses the arrival time. This makes for better
2437 behaviour in cases where some deliveries succeed, thus re-setting the
2438 "first failed" field. An example is a quota failure for a huge message
2439 when small messages continue to be delivered. Without this change, the
2440 "time since failure" will always be short, possible causing more frequent
2441 delivery attempts for the huge message than are intended.
2442 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2444 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2445 $1, $2, $3) because the numerical variables can be reset during some
2446 expansion items (e.g. "match"), thereby losing the authentication data.
2448 PH/21 Make -bV show the size of off_t variables so that the test suite can
2449 decide whether to run tests for quotas > 2G.
2451 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2452 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2453 filecount value is greater than 2G or if a quota value is greater than 2G
2454 on a system where the size of off_t is not greater than 4, a panic error
2457 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2458 never match. The debug and -bh output now contains an explicit error
2459 message indicating a malformed IPv4 address or mask.
2461 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2462 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2463 PH/23 above applies.
2465 PH/25 Do not write to syslog when running in the test harness. The only
2466 occasion when this arises is a failure to open the main or panic logs
2467 (for which there is an explicit test).
2469 PH/26 Added the /no_tell option to "control=freeze".
2471 PH/27 If a host name lookup failed very early in a connection, for example, if
2472 the IP address matched host_lookup and the reverse lookup yielded a name
2473 that did not have a forward lookup, an error message of the form "no IP
2474 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2475 could be logged. Now it outputs the IP address instead of "NULL".
2477 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2478 allows the sender and the authenticated sender to be set when
2479 submitting a message from within Exim. Since child_open_exim() is
2480 documented for local_scan(), the new function should be too.
2482 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2483 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2484 results in an empty string is now treated as unset.
2486 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2488 PH/31 Added sender_verify_fail logging option.
2490 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2491 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2492 tidied the source and removed it altogether.
2494 PH/33 When a queue run was abandoned because the load average was too high, a
2495 log line was always written; now it is written only if the queue_run log
2496 selector is set. In addition, the log line for abandonment now contains
2497 information about the queue run such as the pid. This is always present
2498 in "start" and "stop" lines but was omitted from the "abandon" line.
2500 PH/34 Omit spaces between a header name and the colon in the error message that
2501 is given when verify = headers_syntax fails (if there are lots of them,
2502 the message gets confusing).
2504 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2505 names, as there are now some PTR records that contain slashes. This check
2506 is only to protect against broken name servers that fall over on strange
2507 characters, so the fact that it applies to all lookups doesn't matter.
2509 PH/36 Now that the new test suite is complete, we can remove some of the
2510 special code in Exim that was needed for the old test suite. For example,
2511 sorting DNS records because real resolvers return them in an arbitrary
2512 order. The new test suite's fake resolver always returns records in the
2515 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2516 bounces) except when queue_only is set, to avoid logging races between
2517 the different processes.
2519 PH/38 Panic-die if .include specifies a non-absolute path.
2521 PH/39 A tweak to the "H" retry rule from its user.
2523 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2524 a label. They prevented compilation on older perls.
2526 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2527 a warning to be raised on newish perls.
2529 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2530 on queue. Changes to match documented behaviour of showing count of
2531 messages matching specified criteria.
2533 PH/40 Changed the default ident timeout from 30s to 5s.
2535 PH/41 Added support for the use of login_cap features, on those BSD systems
2536 that have them, for controlling the resources used by pipe deliveries.
2538 PH/42 The content-scanning code uses fopen() to create files in which to put
2539 message data. Previously it was not paying any attention to the mode of
2540 the files. Exim runs with umask(0) because the rest of the code creates
2541 files with open(), and sets the required mode explicitly. Thus, these
2542 files were ending up world-writeable. This was not a big issue, because,
2543 being within the spool directory, they were not world-accessible. I have
2544 created a function called modefopen, which takes an additional mode
2545 argument. It sets umask(777), creates the file, chmods it to the required
2546 mode, then resets the umask. All the relevant calls to fopen() in the
2547 content scanning code have been changed to use this function.
2549 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2550 to 24 hours. This avoids potential overflow problems when processing G
2551 and H retry rules. I suspect nobody ever tinkers with this value.
2553 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2555 PH/45 When the plaintext authenticator is running as a client, the server's
2556 challenges are checked to ensure they are valid base64 strings. By
2557 default, the authentication attempt is cancelled if an invalid string is
2558 received. Setting client_ignore_invalid_base64 true ignores these errors.
2559 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2560 they are received. Thus, the responses can be made to depend on the
2561 challenges. If an invalid string is ignored, an empty string is placed in
2564 PH/46 Messages that are created by the autoreply transport now contains a
2565 References: header, in accordance with RFCs 2822 and 3834.
2567 PH/47 Added authenticated_sender_force to the smtp transport.
2569 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2571 PH/49 Installed latest patch from the Sieve maintainer.
2573 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2574 was also set, the appendfile transport was unnecessarily scanning a
2575 directory of message files (e.g. for maildir delivery) to find the count
2576 of files (along with the size), even though it did not need this
2577 information. It now does the scan only if it needs to find either the
2578 size of the count of files.
2580 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2582 PH/52 Two bugs concerned with error handling when the smtp transport is
2585 (i) Exim was not creating retry information for temporary errors given
2586 for individual recipients after the DATA command when the smtp transport
2587 was used in LMTP mode. This meant that they could be retried too
2588 frequently, and not timed out correctly.
2590 (ii) Exim was setting the flag that allows error details to be returned
2591 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2592 recipients that were returned after the DATA command.
2594 PH/53 This is related to PH/52, but is more general: for any failing address,
2595 when detailed error information was permitted to be returned to the
2596 sender, but the error was temporary, then after the final timeout, only
2597 "retry timeout exceeded" was returned. Now it returns the full error as
2598 well as "retry timeout exceeded".
2600 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2601 do this, and (what is worse) MTAs that accept it.
2603 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2604 will now be deprecated.
2606 PH/56 New os.c-cygwin from the Cygwin maintainer.
2608 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2609 formats (previously only available in exim formats via -bpr, -bpru,
2610 and -bpra. Now also available in native and exiqgrep formats)
2612 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2613 with very large, slow to parse queues
2615 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2617 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2619 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2620 responses to authentication challenges, though it was showing the
2621 challenges; (ii) I've removed the CR characters from the debug output for
2624 PH/58 Allow for the insertion of a newline as well as a space when a string
2625 is turned into more than one encoded-word during RFC 2047 encoding. The
2626 Sieve code now uses this.
2628 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2629 data_4xx, lost_connection, tls_required.
2631 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2632 message was being sent as an SMTP response.
2634 PH/61 Add -l and -k options to exicyclog.
2636 PH/62 When verifying, if an address was redirected to one new address, so that
2637 verification continued, and the new address failed or deferred after
2638 having set something in $address_data, the value of $address_data was not
2639 passed back to the ACL. This was different to the case when no
2640 redirection occurred. The value is now passed back in both cases.
2642 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2643 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2644 use login_cap.h, so on its own it isn't the distinguishing feature. The
2645 new name refers directly to the setclassresources() function.
2647 PH/65 Added configuration files for NetBSD3.
2649 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2651 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2652 is preferred over IPv4.
2654 PH/68 The bounce_return_message and bounce_return_body options were not being
2655 honoured for bounces generated during the reception of non-SMTP messages.
2656 In particular, this applied to messages rejected by the ACL. This bug has
2657 been fixed. However, if bounce_return_message is true and bounce_return_
2658 body is false, the headers that are returned for a non-SMTP message
2659 include only those that have been read before the error was detected.
2660 (In the case of an ACL rejection, they have all been read.)
2662 PH/69 The HTML version of the specification is now built in a directory called
2663 spec_html instead of spec.html, because the latter looks like a path with
2664 a MIME-type, and this confuses some software.
2666 PH/70 Catch two compiler warnings in sieve.c.
2668 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2669 function verify_get_ident() calls ip_connect() to connect a socket, but
2670 if the "connect()" function timed out, ip_connect() used to close the
2671 socket. However, verify_get_ident() also closes the socket later, and in
2672 between Exim writes to the log, which may get opened at this point. When
2673 the socket was closed in ip_connect(), the log could get the same file
2674 descriptor number as the socket. This naturally causes chaos. The fix is
2675 not to close the socket in ip_connect(); the socket should be closed by
2676 the function that creates it. There was only one place in the code where
2677 this was missing, in the iplookup router, which I don't think anybody now
2678 uses, but I've fixed it anyway.
2680 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2681 well as to direct DNS lookups. Otherwise the handling of names in host
2682 lists is inconsistent and therefore confusing.
2688 PH/01 Two changes to the default runtime configuration:
2690 (1) Move the checks for relay_from_hosts and authenticated clients from
2691 after to before the (commented out) DNS black list checks.
2693 (2) Add control=submission to the relay_from_hosts and authenticated
2694 clients checks, on the grounds that messages accepted by these
2695 statements are most likely to be submissions.
2697 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2699 (1) Generate an error if the third argument for the ${prvs expansion is
2702 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2705 (3) Reset the variables that are obtained from the first argument of
2706 ${prvscheck and used in the second argument before leaving the code,
2707 because their memory is reclaimed, so using them afterwards may do
2710 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2711 one (it's much easier than Tom thought :-).
2713 (5) Because of (4), we can now allow for the use of $prvscheck_result
2714 inside the third argument.
2716 PH/03 For some reason, the default setting of PATH when running a command from
2717 a pipe transport was just "/usr/bin". I have changed it to
2720 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2721 anything to be listed in the output from -bV.
2723 PH/05 When a filter generated an autoreply, the entire To: header line was
2724 quoted in the delivery log line, like this:
2726 => >A.N.Other <ano@some.domain> <original@ddress> ...
2728 This has been changed so that it extracts the operative address. There
2729 may be more than one such address. If so, they are comma-separated, like
2732 => >ano@some.domain,ona@other.domain <original@ddress> ...
2734 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2735 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2736 not being looked up in the rDNS to get a host name, Exim was showing the
2737 IP address twice in Received: lines, even though the IP addresses were
2738 identical. For example:
2740 Received: from [1.2.3.4] (helo=[1.2.3.4])
2742 However, if the real host name was known, it was omitting the HELO data
2743 if it matched the actual IP address. This has been tidied up so that it
2744 doesn't show the same IP address twice.
2746 PH/07 When both +timestamp and +memory debugging was on, the value given by
2747 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2748 called by the debug printing, thereby overwriting the timestamp buffer.
2749 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2751 PH/08 When the original message was included in an autoreply transport, it
2752 always said "this is a copy of the message, including all the headers",
2753 even if body_only or headers_only was set. It now gives an appropriate
2756 PH/09 Applied a patch from the Sieve maintainer which:
2758 o fixes some comments
2759 o adds the (disabled) notify extension core
2760 o adds some debug output for the result of if/elsif tests
2761 o points to the current vacation draft in the documentation
2762 and documents the missing references header update
2766 o fixes a bug in processing the envelope test (when testing
2767 multiple envelope elements, the last element determinted the
2770 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2771 Electronic Mail") by including:
2773 Auto-submitted: auto-generated
2775 in the messages that it generates (bounce messages and others, such as
2776 warnings). In the case of bounce messages for non-SMTP mesages, there was
2777 also a typo: it was using "Auto_submitted" (underscore instead of
2778 hyphen). Since every message generated by Exim is necessarily in response
2779 to another message, thes have all been changed to:
2781 Auto-Submitted: auto-replied
2783 in accordance with these statements in the RFC:
2785 The auto-replied keyword:
2787 - SHOULD be used on messages sent in direct response to another
2788 message by an automatic process,
2790 - MUST NOT be used on manually-generated messages,
2792 - MAY be used on Delivery Status Notifications (DSNs) and Message
2793 Disposition Notifications (MDNs),
2795 - MUST NOT be used on messages generated by automatic or periodic
2796 processes, except for messages which are automatic responses to
2799 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2800 to the default Received: header definition.
2802 PH/12 Added log selector acl_warn_skipped (default on).
2804 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2805 variables because (a) they are in the wrong storage pool and (b) even if
2806 they were copied, it wouldn't work properly because of the caching.
2808 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2809 checking when decoding. Apparently there are clients that generate
2810 overlong encoded strings. Why am I not surprised?
2812 PH/15 If the first argument of "${if match_address" was not empty, but did not
2813 contain an "@" character, Exim crashed. Now it writes a panic log message
2814 and treats the condition as false.
2816 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2818 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2819 extension "envelope-auth". The code is finished and in agreement with
2820 other implementations, but there is no documentation so far and in fact,
2821 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2822 not changing the active code.
2824 Print executed "if" and "elsif" statements when debugging is used. This
2825 helps a great deal to understand what a filter does.
2827 Document more things not specified clearly in RFC3028. I had all this
2828 sorted out, when out of a sudden new issues came to my mind. Oops."
2830 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2833 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2834 canonical form (as documented). However, after a host name lookup from
2835 the IP address, check_host() was doing a simple string comparison with
2836 addresses acquired from the DNS when checking that the found name did
2837 have the original IP as one of its addresses. Since any found IPv6
2838 addresses are likely to be in abbreviated form, the comparison could
2839 fail. Luckily, there already exists a function for doing the comparison
2840 by converting both addresses to binary, so now that is used instead of
2841 the text comparison.
2843 PH/20 There was another similar case to PH/19, when a complete host name was
2844 given in a host list; looking up its IP address could give an abbreviated
2845 form, whereas the current host's name might or might not be abbreviated.
2846 The same fix has been applied.
2852 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2853 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2856 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2857 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2859 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2861 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2862 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2863 Also, for "Auto-Submitted", treat anything other than "no" as
2864 non-personal, in accordance with RFC 3834. (Previously it treated
2865 anything starting "auto-" as non-personal.)
2867 TF/01 The control=submission/name=... option had a problem with syntax
2868 errors if the name included a slash character. The /name= option
2869 now slurps the rest of the string, so it can include any characters
2870 but it must come last in the list of options (after /sender_retain
2873 PH/05 Some modifications to the interface to the fake nameserver for the new
2881 TK/01 Added the "success_on_redirect" address verification option. See
2882 NewStuff for rationale and an example.
2884 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2886 PH/02 Patch to exigrep to allow it to work on syslog lines.
2888 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2889 fread() to skip over the body file's header line, because in Cygwin the
2890 header line is locked and is inaccessible.
2892 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2893 co-exist for some time) to make it clear that it is the Exim ID that is
2894 referenced, not the Message-ID: header line.
2896 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2897 string_format() function, because snprintf() does not exist on all
2900 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2901 ${stat: expansion item.
2903 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2904 protocol synchronization error", to keep the pedants happy.
2906 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2907 well as for IRIX systems, when gcc is being used. See the host.c source
2910 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2912 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2915 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2916 transport and to the smtp transport in LMTP mode.
2918 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2920 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2921 run a filter in a subprocess. This could lead to confusion in subsequent
2922 lookups in the parent process. There should also be a search_tidyup() at
2923 the end of the subprocess.
2925 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2926 only if the host matched helo_try_verify_hosts, which caused the
2927 verification to occur when the EHLO/HELO command was issued. The ACL just
2928 tested the remembered result. Now, if a previous verification attempt has
2929 not happened, "verify = helo" does it there and then.
2931 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2933 TK/03 Fix log output including CR from clamd.
2935 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2936 exist provoked a memory error which could cause a segfault.
2938 PH/15 Installed PCRE 6.2
2940 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2942 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2943 of the problem. Specifically, suggested +O2 rather than +O1 for the
2946 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2948 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2949 record, the port was not correctly logged when the outgoing_port log
2950 selector was set (it logged the transort's default port).
2952 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2953 fallback_hosts, and "hosts" in the smtp transport.
2955 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2956 host errors such as "Connection refused".
2958 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2959 authentication with radiusclient 0.4.9:
2961 - Error returned from rc_read_config was caught wrongly
2962 - Username/password not passed on to radius server due to wrong length.
2964 The presumption is that some radiusclient API changes for 4.51/PH/17
2965 were not taken care of correctly. The code is still untested by me (my
2966 Linux distribution still has 0.3.2 of radiusclient), but it was
2967 contributed by a Radius user.
2969 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2970 expanding the "port" option of the smtp transport.
2972 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2973 while reading a MIME header. Thanks to Tom Hughes for a patch.
2975 PH/24 Include config.h inside local_scan.h so that configuration settings are
2978 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2979 that in an ACL for RCPT (for example), you can examine exactly what was
2982 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2983 commands, but it was not correctly comparing the address with the actual
2984 client host address. Thus, it would show the EHLO address in Received:
2985 header lines when this was not necessary.
2987 PH/27 Added the % operator to ${eval:}.
2989 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2990 it should be ignoring failures (because with -C, for example, it has lost
2991 privilege). It wasn't ignoring creation failures other than "already
2994 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2997 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2998 because an input line was too long, either on its own, or by virtue of
2999 too many continuations, the temporary file was not being removed, and the
3000 return code was incorrect.
3002 PH/31 Missing "BOOL" in function definition in filtertest.c.
3004 PH/32 Applied Sieve patches from the maintainer.
3006 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3008 PH/33 Added "verify = not_blind".
3010 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3011 Local/Makefile (with some defaults set). These are used in built scripts
3012 such as exicyclog, but they have never been used in the exim_install
3013 script (though there are many overriding facilities there). I have
3014 arranged that the exim_install script now takes note of these two
3017 PH/35 Installed configuration files for Dragonfly.
3019 PH/36 When a locally submitted message by a trusted user did not contain a
3020 From: header, and the sender address was obtained from -f or from an SMTP
3021 MAIL command, and the trusted user did not use -F to supply a sender
3022 name, $originator_name was incorrectly used when constructing a From:
3023 header. Furthermore, $originator_name was used for submission mode
3024 messages from external hosts without From: headers in a similar way,
3025 which is clearly wrong.
3027 PH/37 Added control=suppress_local_fixups.
3029 PH/38 When log_selector = +received_sender was set, and the addition of the
3030 sender made the log line's construction buffer exactly full, or one byte
3031 less than full, an overflow happened when the terminating "\n" was
3034 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3035 when the result of a list match is failure because a DNS lookup failed.
3037 PH/40 RM_COMMAND is now used in the building process.
3039 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3040 the "build-* directories that it finds.
3042 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3043 address in a domain literal was a prefix of an interface address.
3045 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3046 when verifying a sender address, unless rewrite_headers is false.
3048 PH/44 Wrote a long comment about why errors_to addresses are verified as
3049 recipients, not senders.
3051 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3052 the ratelimit ACL was added.
3054 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3056 PH/46 Added extra information about PostgreSQL errors to the error string.
3058 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3059 avoiding the need to install special zones in a real server. This is
3060 backwards compatible; if it can't find the fake resolver, it drops back.
3061 Thus, both old and new test suites can be run.
3063 TF/02 Added util/ratelimit.pl
3065 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3066 clock is set back in time.
3068 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3069 Candler <B.Candler@pobox.com>.
3071 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3072 OK for addresses that are the result of redirections.
3074 PH/48 A number of further additions for the benefit of the new test suite,
3075 including a fake gethostbyname() that interfaces to the fake DNS resolver
3078 TF/06 The fix for widen_domains has also been applied to qualify_single and
3079 search_parents which are the other dnslookup options that can cause
3082 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3085 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3087 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3088 service" if it encountered an MX record with an empty target hostname.
3089 The message is now "an MX or SRV record indicated no SMTP service".
3091 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3092 if the DNS of the sending site is misconfigured. This is quite a
3093 common situation. This change restores the behaviour of treating a
3094 helo verification defer as a failure.
3096 PH/51 If self=fail was set on a router, the bounce message did not include the
3097 actual error message.
3103 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3105 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3106 command itself ended in error, the underlying message about the transport
3107 filter timeout was being overwritten with the pipe command error. Now the
3108 underlying error message should be appended to the second error message.
3110 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3112 PH/02 Reduce the amount of output that "make" produces by default. Full output
3113 can still be requested.
3115 PH/03 The warning log line about a condition test deferring for a "warn" verb
3116 was being output only once per connection, rather than after each
3117 occurrence (because it was using the same function as for successful
3118 "warn" verbs). This seems wrong, so I have changed it.
3120 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3121 it should not have, which might have caused a crash in the right
3122 circumstances, but probably never did.
3124 PH/04 Installed a modified version of Tony Finch's patch to make submission
3125 mode fix the return path as well as the Sender: header line, and to
3126 add a /name= option so that you can make the user's friendly name appear
3129 TF/03 Added the control = fakedefer ACL modifier.
3131 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3132 Mark Lowes for thorough testing.
3134 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3136 TK/03 Merged latest SRS patch from Miles Wilton.
3138 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3139 with the definition in sysexits.h (which is #included earlier).
3140 Fortunately, Exim does not actually use EX_OK. The code used to try to
3141 preserve the sysexits.h value, by assumimg that macro definitions were
3142 scanned for macro replacements. I have been disabused of this notion,
3143 so now the code just undefines EX_OK before #including unistd.h.
3145 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3146 in the smtp transport. When a block could not be written in a single
3147 write() function, the timeout was being re-applied to each part-write.
3148 This seems wrong - if the receiver was accepting one byte at a time it
3149 would take for ever. The timeout is now adjusted when this happens. It
3150 doesn't have to be particularly precise.
3152 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3153 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3155 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3156 without a domain if the check to <postmaster@domain> fails.
3158 SC/01 Eximstats: added -xls and the ability to specify output files
3159 (patch written by Frank Heydlauf).
3161 SC/02 Eximstats: use FileHandles for outputing results.
3163 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3165 SC/04 Eximstats: fixed display of large numbers with -nvr option
3167 SC/05 Eximstats: fixed merging of reports with empty tables.
3169 SC/06 Eximstats: added the -include_original_destination flag
3171 SC/07 Eximstats: removed tabs and trailing whitespace.
3173 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3175 TK/06 MBOX spool code: Add real "From " MBOX separator line
3176 so the .eml file is really in mbox format (even though
3177 most programs do not really care). Patch from Alex Miller.
3179 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3180 The latter is generated from $received_to and is only set if the
3181 message has one envelope recipient. SA can use these headers,
3182 obviously out-of-the-box. Patch from Alex Miller.
3184 PH/08 The ${def test on a variable was returning false if the variable's
3185 value was "0", contrary to what the specification has always said!
3186 The result should be true unless the variable is empty.
3188 PH/09 The syntax error of a character other than { following "${if
3189 def:variable_name" (after optional whitespace) was not being diagnosed.
3190 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3191 accidental colon was present, for example, could give incorrect results.
3193 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3194 result is used (not including appendfile, where other changes are about
3197 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3198 This involved changing a lot of size variables from int to off_t. It
3199 should work with maildirs and everything.
3201 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3202 spamd dying while we are connected to it.
3204 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3207 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3208 function for BDB 4.3.
3210 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3212 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3213 the sources for PCRE have been re-arranged and more files are now
3216 PH/15 The code I had for printing potentially long long variables in PH/11
3217 above was not the best (it lost precision). The length of off_t variables
3218 is now inspected at build time, and an appropriate printing format (%ld
3219 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3220 to be "long long int" or "long int". This is needed for the internal
3221 formatting function string_vformat().
3223 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3224 the configuration file to be ":syslog", then the script "guesses" where
3225 the logs files are, rather than using the compiled in default. In our
3226 case the guess is not the same as the compiled default, so the script
3227 suddenly stopped working when I started to use syslog. The patch checks
3228 to see if log_file_path is "". If so, it attempts to read it from exim
3229 with no configuration file to get the compiled in version, before it
3230 falls back to the previous guessing code."
3232 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3233 implementing BATV in an Exim configuration. See NewStuff for the gory
3236 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3237 Makefile that are specific to HP-UX.
3239 PH/18 If the "use_postmaster" option was set for a recipient callout together
3240 with the "random" option, the postmaster address was used as the MAIL
3241 FROM address for the random test, but not for the subsequent recipient
3242 test. It is now used for both.
3244 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3245 patch removes a few documentation additions to RFC 3028, because the
3246 latest draft now contains them. It adds the new en;ascii-case comparator
3247 and a new error check for 8bit text in MIME parts. Comparator and
3248 require names are now matched exactly. I enabled the subaddress
3249 extension, but it is not well tested yet (read: it works for me)."
3251 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3252 rework some of the code of TK/09 above to avoid the hardwired use of
3253 "%lld" and "long long". Replaced the call to snprintf() with a call to
3256 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3257 records point to non-existent hosts", "retry timeout exceeded", and
3258 "retry time not reached for any host after a long failure period".
3260 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3261 experimental DomainKeys support:
3263 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3264 (2) On an error such as an illegally used "control", the wrong name for
3265 the control was given.
3267 These problems did NOT occur unless DomainKeys support was compiled.
3269 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3271 PH/24 Added ${if match_ip condition.
3273 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3274 databases so that it will be absolutely obvious if a crash occurs in the
3275 DB library. This is a regular occurrence (often caused by mis-matched
3278 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3279 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3280 picked up on a user's system that detects such things. There doesn't seem
3281 to be a gcc warning option for this - only an attribute that has to be
3282 put on the function's prototype. It seems that in Fedora Core 4 they have
3283 set this on a number of new functions. No doubt there will be more in due
3286 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3287 specify a transport. However, if an address that was verified by such a
3288 router was the subject of a callout, Exim crashed because it tried to
3289 read the rcpt_include_affixes from the non-existent transport. Now it
3290 just assumes that the setting of that option is false. This bug was
3291 introduced by 4.51/PH/31.
3293 PH/28 Changed -d+all to exclude +memory, because that information is very
3294 rarely of interest, but it makes the output a lot bigger. People tend to
3295 do -d+all out of habit.
3297 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3298 code in os-type was giving problems when libc.so lives in lib64, like on
3301 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3302 aren't the modern standard, and it seems that some systems' include files
3303 don't always have them. Exim was already checking for some of the newer
3304 ones like T_AAAA, and defining it itself. I've added checks for all the
3305 record types that Exim uses.
3307 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3308 not automatically generating a new one, as it is supposed to. This
3309 prevented TLS from working. If the file did exist, but contained invalid
3310 data, a new version was generated, as expected. It was only the case of a
3311 non-existent file that was broken.
3313 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3314 with a change in libdomainkeys > 0.64.
3316 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3317 from DNS. If the selector record carries the flag, it now has
3318 precedence over the domain-wide flag.
3320 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3322 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3323 the use of a port name that isn't defined in /etc/services) occurred, the
3324 message was deferred as in a normal delivery, and thus remained on the
3325 spool, instead of being failed because of the mua_wrapper setting. This
3326 is now fixed, and I tidied up some of the mua_wrapper messages at the
3329 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3330 the messages in a hash of arrays rather than using individual hashes.
3331 This is a bit cleaner and results in dramatic memory savings, albeit
3332 at a slight CPU cost.
3334 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3335 as requested by Marc Sherman.
3337 SC/10 Eximstats: added histograms for user specified patterns as requested
3340 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3342 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3343 fopen() in the content-scanning modules that did not already have it.
3349 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3350 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3352 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3354 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3356 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3357 to test Sieve filters that use "vacation".
3359 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3360 that changes the way the GnuTLS parameters are stored in the cache file.
3361 The new format can be generated externally. For backward compatibility,
3362 if the data in the cache doesn't make sense, Exim assumes it has read an
3363 old-format file, and it generates new data and writes a new file. This
3364 means that you can't go back to an older release without removing the
3367 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3368 work if there are any delivery delays because "one_time" forces the
3369 parent to be marked "delivered", so its unseen clone is never tried
3370 again. For this reason, Exim now forbids the simultaneous setting of
3373 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3374 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3375 entry if you want to know the details. The fix, however, neglected to
3376 consider the case when local delivery batching is involved. The test for
3377 "previously delivered" was not happening when checking to see if an
3378 address could be batched with a previous (undelivered) one; under
3379 certain circumstances this could lead to multiple deliveries to the same
3382 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3383 in its include files, and this causes problems building Exim.
3385 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3386 header_syntax) but Exim was just ignoring anything given after a slash.
3387 In particular, this caused confusion with an attempt to use "verify =
3388 reverse_host_lookup/defer_ok". An error is now given when options are
3389 supplied for verify items that do not have them. (Maybe reverse_host_
3390 lookup should have a defer_ok option, but that's a different point.)
3392 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3393 defined by RFC 821) to 2048, because there were problems with some AUTH
3394 commands, and RFC 1869 says the size should be increased for extended
3395 SMTP commands that take arguments.
3397 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3400 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3401 "unknown" error; now it says that the functionality isn't in the binary.
3403 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3404 an address' error message when a string expansion fails (syntax or
3405 whatever). Otherwise the password may appear in the log. Following change
3406 PH/42 below, there is no longer a chance of it appearing in a bounce
3409 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3411 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3412 of its addresses was ever tried. (Bugzilla bug #2.)
3414 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3415 the result incorrectly in the debug output. (It correctly added a newline
3416 to what was transported.)
3418 TF/01 Added $received_time.
3420 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3421 commented out examples of how to interface to a virus scanner and to
3422 SpamAssassin. Also added commented examples of av_scanner and
3423 spamd_address settings.
3425 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3426 and controls are allowed in which ACLs. There were a couple of minor
3427 errors. Some of the entries in the conditions table (which is a table of
3428 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3429 negation of where the condition IS allowed.
3431 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3433 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3434 header file does not have a version number, so I've had to invent a new
3435 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3436 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3437 radiusclient), but it was contributed by a Radius user.
3439 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3440 files or directories, for OpenSSL.
3442 PH/19 When an Exim process that is running as root has to create an Exim log
3443 file, it does so in a subprocess that runs as exim:exim so as to get the
3444 ownership right at creation (otherwise, other Exim processes might see
3445 the file with the wrong ownership). There was no test for failure of this
3446 fork() call, which would lead to the process getting stuck as it waited
3447 for a non-existent subprocess. Forks do occasionally fail when resources
3448 run out. I reviewed all the other calls to fork(); they all seem to check
3451 PH/20 When checking for unexpected SMTP input at connect time (before writing
3452 the banner), Exim was not dealing correctly with a non-positive return
3453 from the read() function. If the client had disconnected by this time,
3454 the result was a log entry for a synchronization error with an empty
3455 string after "input=" when read() returned zero. If read() returned -1
3456 (an event I could not check), uninitialized data bytes were printed.
3457 There were reports of junk text (parts of files, etc) appearing after
3460 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3462 PH/22 Added support for macro redefinition, and (re)definition in between
3463 driver and ACL definitions.
3465 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3466 forgetting to use the resulting value; it was using the unexpanded value.
3468 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3469 hadn't been configured. The fix is from Juergen Kreileder, who
3470 understands it better than I do:
3472 "Here's what I see happening with three configured cyrus_sasl
3473 authenticators configured (plain, login, cram-md5):
3475 On startup auth_cyrus_sasl_init() gets called for each of these.
3476 This means three calls to sasl_listmech() without a specified mech_list.
3477 => SASL tests which mechs of all available mechs actually work
3478 => three warnings about OTP not working
3479 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3481 With the patch, sasl_listmech() also gets called three times. But now
3482 SASL's mech_list option is set to the server_mech specified in the the
3483 authenticator. Or in other words, the answer from sasl_listmech()
3484 gets limited to just the mech you're testing for (which is different
3486 => the return list contains just 'plain' or 'login', 'cram-md5' or
3487 nothing depending on the value of ob->server_mech.
3489 I've just tested the patch: Authentication still works fine,
3490 unavailable mechs specified in the exim configuration are still
3491 caught, and the auth.log warnings about OTP are gone."
3493 PH/25 When debugging is enabled, the contents of the command line are added
3494 to the debugging output, even when log_selector=+arguments is not
3497 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3498 answer is "GNU", and only if the return is "GNU/something" is the answer
3501 PH/27 $acl_verify_message is now set immediately after the failure of a
3502 verification in an ACL, and so is available in subsequent modifiers. In
3503 particular, the message can be preserved by coding like this:
3505 warn !verify = sender
3506 set acl_m0 = $acl_verify_message
3508 Previously, $acl_verify_message was set only while expanding "message"
3509 and "log_message" when a very denied access.
3511 PH/28 Modified OS/os.c-Linux with
3513 -#ifndef OS_LOAD_AVERAGE
3514 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3516 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3517 nomenclature these days.)
3519 PH/29 Installed patch from the Sieve maintainer that adds the options
3520 sieve_useraddress and sieve_subaddress to the redirect router.
3522 PH/30 In these circumstances:
3523 . Two addresses routed to the same list of hosts;
3524 . First host does not offer TLS;
3525 . First host accepts first address;
3526 . First host gives temporary error to second address;
3527 . Second host offers TLS and a TLS session is established;
3528 . Second host accepts second address.
3529 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3530 and peerdn, if requested) that were in fact used only for the second
3533 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3534 attention to any local part prefix or suffix that was matched by the
3535 router that accepted the address. It now behaves in the same way as it
3536 does for delivery: the affixes are removed from the local part unless
3537 rcpt_include_affixes is set on the transport.
3539 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3540 timeout during the DATA phase of an incoming message.
3542 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3543 applied a patch sanctioned by the Sieve maintainer.
3545 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3546 the uid or gid is negative. A case of a negative gid caused this to be
3547 noticed. The fix allows for either to be negative.
3549 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3550 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3553 PH/36 The change PH/12 above was broken. Fixed it.
3555 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3556 the grounds that routing the same address twice would always produce the
3557 same answer. This might have been true once, but it is certainly no
3558 longer true now. Routing a child address may depend on the previous
3559 routing that produced that child. Some complicated redirection strategies
3560 went wrong when messages had multiple recipients, and made Exim's
3561 behaviour dependent on the order in which the addresses were given.
3563 I have moved the duplicate checking until after the routing is complete.
3564 Exim scans the addresses that are assigned to local and remote
3565 transports, and removes any duplicates. This means that more work will be
3566 done, as duplicates will always all be routed, but duplicates are
3567 presumably rare, so I don't expect this is of any significance.
3569 For deliveries to pipes, files, and autoreplies, the duplicate checking
3570 still happens during the routing process, since they are not going to be
3573 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3574 It corrects a timeout issue with spamd. This is Ian's comment: "The
3575 background is that sometimes spamd either never reads data from a
3576 connection it has accepted, or it never writes response data. The exiscan
3577 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3578 blindly assumes that writes won't block so it may never time out."
3580 PH/39 Allow G after quota size as well as K and M.
3582 PH/40 The value set for $authenticated_id in an authenticator may not contain
3583 binary zeroes or newlines because the value is written to log lines and
3584 to spool files. There was no check on this. Now the value is run through
3585 the string_printing() function so that such characters are converted to
3586 printable escape sequences.
3588 PH/41 $message_linecount is a new variable that contains the total number of
3589 lines in the message. Compare $body_linecount, which is the count for the
3592 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3593 bounce and delay warning messages, except in certain special cases, which
3596 (a) An SMTP error message from a remote host;
3597 (b) A message specified in a :fail: redirection;
3598 (c) A message specified in a "fail" command in a system filter;
3599 (d) A message specified in a FAIL return from the queryprogram router;
3600 (e) A message specified by the cannot_route_message router option.
3602 In these cases only, Exim does include the error details in bounce and
3603 warning messages. There are also a few cases where bland messages such
3604 as "unrouteable address" or "local delivery error" are given.
3606 PH/43 $value is now also set for the "else" part of a ${run expansion.
3608 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3609 being worked on, but at least Exim now implements the latest version to
3612 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3613 process to complete was treated as a delivery failure, a timeout while
3614 writing the message to the pipe was logged, but erroneously treated as a
3615 successful delivery. Such timeouts include transport filter timeouts. For
3616 consistency with the overall process timeout, these timeouts are now
3617 treated as errors, giving rise to delivery failures by default. However,
3618 there is now a new Boolean option for the pipe transport called
3619 timeout_defer, which, if set TRUE, converts the failures into defers for
3620 both kinds of timeout. A transport filter timeout is now identified in
3623 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3624 systems where "make" and "gmake" are different, calling "gmake" at top
3625 level broke things. I've arranged for the value of $(MAKE) to be passed
3626 from the Makefile to this script so that it can call the same version of
3630 A note about Exim versions 4.44 and 4.50
3631 ----------------------------------------
3633 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3634 changes of various kinds. As a consequence, a big documentation update was
3635 needed. This delayed the release for rather longer than seemed good, especially
3636 in the light of a couple of (minor) security issues. Therefore, the changes
3637 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3638 release. So 4.44 and 4.50 are in effect two different branches that both start
3641 I have left the 4.50 change log unchanged; it contains all the changes since
3642 4.43. The change log for 4.44 is below; many of its items are identical to
3643 those for 4.50. This seems to be the most sensible way to preserve the
3644 historical information.
3650 1. Minor wording change to the doc/README.SIEVE file.
3652 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3653 computation of the current number of files was incorrect.
3655 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3656 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3659 4. Give more explanation in the error message when the command for a transport
3660 filter fails to execute.
3662 5. There are several places where Exim runs a non-Exim command in a
3663 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3664 was being done only for the command run by the queryprogram router. It is
3665 now done for all such subprocesses. The other cases are: ${run, transport
3666 filters, and the commands run by the lmtp and pipe transports.
3668 6. Added CONFIGURE_GROUP build-time option.
3670 7. Some older OS have a limit of 256 on the maximum number of file
3671 descriptors. Exim was using setrlimit() to set 1000 as a large value
3672 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3673 systems. I've change it so that if it can't get 1000, it tries for 256.
3675 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3676 was an oversight, and furthermore, ever since the addition of extra
3677 controls (e.g. 4.43/32), the checks on when to allow different forms of
3678 "control" were broken. There should now be diagnostics for all cases when a
3679 control that does not make sense is encountered.
3681 9. Added the /retain_sender option to "control=submission".
3683 10. $recipients is now available in the predata ACL (oversight).
3685 11. Tidy the search cache before the fork to do a delivery from a message
3686 received from the command line. Otherwise the child will trigger a lookup
3687 failure and thereby defer the delivery if it tries to use (for example) a
3688 cached ldap connection that the parent has called unbind on.
3690 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3691 of $address_data from the recipient verification was clobbered by the
3692 sender verification.
3694 13. The value of address_data from a sender verification is now available in
3695 $sender_address_data in subsequent conditions in the ACL statement.
3697 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3699 15. Added a new option "connect=<time>" to callout options, to set a different
3702 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3703 was its contents. (It was OK if the option was not defined at all.)
3705 17. A "Completed" log line is now written for messages that are removed from
3706 the spool by the -Mrm option.
3708 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3709 information about exactly what failed.
3711 19. Added -dd to debug only the daemon process.
3713 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3714 handles timeouts, both on the server side and network timeouts. Renamed the
3715 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3717 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3718 It is now set to "smtps".
3720 22. $host_address is now set to the target address during the checking of
3721 ignore_target_hosts.
3723 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3724 being passed; this would have caused $sender_host_name to have been used if
3725 matching the list had actually called for a host name (not very likely,
3726 since this list is usually IP addresses). A host name is now passed as
3729 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3730 code that specifies a non-restarting handler (typically sigaction() in
3731 modern systems) in an attempt to fix a rare and obscure crash bug.
3733 25. Narrowed the window for a race in the daemon that could cause it to ignore
3734 SIGCHLD signals. This is not a major problem, because they are used only to
3735 wake it up if nothing else does.
3737 26. A malformed maildirsize file could cause Exim to calculate negative values
3738 for the mailbox size or file count. Odd effects could occur as a result.
3739 The maildirsize information is now recalculated if the size or filecount
3742 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3743 support for a long time. Removed HAVE_SYS_VFS_H.
3745 28. Installed the latest version of exipick from John Jetmore.
3747 29. In an address list, if the pattern was not a regular expression, an empty
3748 subject address (from a bounce message) matched only if the pattern was an
3749 empty string. Non-empty patterns were not even tested. This was the wrong
3750 because it is perfectly reasonable to use an empty address as part of a
3751 database query. An empty address is now tested by patterns that are
3752 lookups. However, all the other forms of pattern expect the subject to
3753 contain a local part and a domain, and therefore, for them, an empty
3754 address still always fails if the pattern is not itself empty.
3756 30. Exim went into a mad DNS loop when attempting to do a callout where the
3757 host was specified on an smtp transport, and looking it up yielded more
3758 than one IP address.
3760 31. Re-factored the code for checking spool and log partition space into a
3761 function that finds that data and another that does the check. The former
3762 is then used to implement four new variables: $spool_space, $log_space,
3763 $spool_inodes, and $log_inodes.
3765 32. The RFC2047 encoding function was originally intended for short strings
3766 such as real names; it was not keeping to the 75-character limit for
3767 encoded words that the RFC imposes. It now respects the limit, and
3768 generates multiple encoded words if necessary. To be on the safe side, I
3769 have increased the buffer size for the ${rfc2047: expansion operator from
3772 33. It is now permitted to omit both strings after an "if" condition; if the
3773 condition is true, the result is "true". As before, when the second string
3774 is omitted, a false condition yields an empty string. This makes it less
3775 cumbersome to write custom ACL and router conditions.
3777 34. Failure to deliver a bounce message always caused it to be frozen, even if
3778 there was an errors_to setting on the router. The errors_to setting is now
3781 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3782 canonical form (fully expanded) before being placed in
3783 $sender_host_address.
3785 36. The table in the code that translates DNS record types into text (T_A to
3786 "A" for instance) was missing entries for NS and CNAME. It is just possible
3787 that this could have caused confusion if both these types were looked up
3788 for the same domain, because the text type is used as part of Exim's
3789 per-process caching. But the chance of anyone hitting this buglet seems
3792 37. The dnsdb lookup has been extended in a number of ways.
3794 (1) There is a new type, "zns", which walks up the domain tree until it
3795 finds some nameserver records. It should be used with care.
3797 (2) There is a new type, "mxh", which is like "mx" except that it returns
3798 just the host names, not the priorities.
3800 (3) It is now possible to give a list of domains (or IP addresses) to be
3801 looked up. The behaviour when one of the lookups defers can be
3802 controlled by a keyword.
3804 (4) It is now possible to specify the separator character for use when
3805 multiple records are returned.
3807 38. The dnslists ACL condition has been extended: it is now possible to supply
3808 a list of IP addresses and/or domains to be looked up in a particular DNS
3811 39. Added log_selector=+queue_time_overall.
3813 40. When running the queue in the test harness, wait just a tad after forking a
3814 delivery process, to get repeatability of debugging output.
3816 41. Include certificate and key file names in error message when GnuTLS fails
3817 to set them up, because the GnuTLS error message doesn't include the name
3818 of the failing file when there is a problem reading it.
3820 42. Allow both -bf and -bF in the same test run.
3822 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3824 44. The "Exiscan patch" is now merged into the mainline Exim source.
3826 45. Sometimes the final signoff response after QUIT could fail to get
3827 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3828 < 0 before doing a fflush(). This bug looks as though it goes back to the
3829 introduction of TLS in release 3.20, but "sometimes" must have been rare
3830 because the tests only now provoked it.
3832 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3833 (this can affect the format of dates).
3835 47. exim_tidydb, when checking for the continued existence of a message for
3836 which it has found a message-specific retry record, was not finding
3837 messages that were in split spool directories. Consequently, it was
3838 deleting retry records that should have stayed in existence.
3840 48. Steve fixed some bugs in eximstats.
3842 49. The SPA authentication driver was not abandoning authentication and moving
3843 on to the next authenticator when an expansion was forced to fail,
3844 contradicting the general specification for all authenticators. Instead it
3845 was generating a temporary error. It now behaves as specified.
3847 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3848 (the order specifies the preference for clients). The order is now AES256,
3849 AES128, 3DES, ARCFOUR128.
3851 51. Small patch to Sieve code - explicitly set From: when generating an
3854 52. Exim crashed if a remote delivery caused a very long error message to be
3855 recorded - for instance if somebody sent an entire SpamAssassin report back
3856 as a large number of 550 error lines. This bug was coincidentally fixed by
3857 increasing the size of one of Exim's internal buffers (big_buffer) that
3858 happened as part of the Exiscan merge. However, to be on the safe side, I
3859 have made the code more robust (and fixed the comments that describe what
3862 53. Now that there can be additional text after "Completed" in log lines (if
3863 the queue_time_overall log selector is set), a one-byte patch to exigrep
3864 was needed to allow it to recognize "Completed" as not the last thing in
3867 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3868 patch that reportedly fixes this has been added. I am not expert enough to
3869 create a test for it. This is what the patch creator wrote:
3871 "I found a little strange behaviour of ldap code when working with
3872 Windows 2003 AD Domain, where users was placed in more than one
3873 Organization Units. When I tried to give exim partial DN, the exit code
3874 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3875 But simultaneously result of request was absolutely normal ldap result,
3876 so I produce this patch..."
3878 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3879 REFERENCE, so I have modified the code to exclude the patch when that macro
3882 55. Some experimental protocols are using DNS PTR records for new purposes. The
3883 keys for these records are domain names, not reversed IP addresses. The
3884 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3885 leaves it alone. Component reversal etc. now happens only for IP addresses.
3888 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3890 57. Double the size of the debug message buffer (to 2048) so that more of very
3891 long debug lines gets shown.
3893 58. The exicyclog utility now does better if the number of log files to keep
3894 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3896 59. Two changes related to the smtp_active_hostname option:
3898 (1) $smtp_active_hostname is now available as a variable.
3899 (2) The default for smtp_banner uses $smtp_active_hostname instead
3900 of $primary_hostname.
3902 60. The host_aton() function is supposed to be passed a string that is known
3903 to be a valid IP address. However, in the case of IPv6 addresses, it was
3904 not checking this. This is a hostage to fortune. Exim now panics and dies
3905 if the condition is not met. A case was found where this could be provoked
3906 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3907 components; fortuitously, this particular loophole had already been fixed
3908 by change 4.50/55 above.
3910 If there are any other similar loopholes, the new check in host_aton()
3911 itself should stop them being exploited. The report I received stated that
3912 data on the command line could provoke the exploit when Exim was running as
3913 exim, but did not say which command line option was involved. All I could
3914 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3915 running as the user.
3918 61. There was a buffer overflow vulnerability in the SPA authentication code
3919 (which came originally from the Samba project). I have added a test to the
3920 spa_base64_to_bits() function which I hope fixes it.
3923 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3924 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3926 63. The daemon start-up calls getloadavg() while still root for those OS that
3927 need the first call to be done as root, but it missed one case: when
3928 deliver_queue_load_max is set with deliver_drop_privilege. This is
3929 necessary for the benefit of the queue runner, because there is no re-exec
3930 when deliver_drop_privilege is set.
3932 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3933 This has been fixed.
3935 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3936 was in use, was not putting the data itself into the right store pool;
3937 consequently, it could be overwritten for a subsequent message in the same
3938 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3941 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3943 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3944 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3945 the function were treating the return as a boolean value, which happened to
3946 work because 0=false and not-0=true, but is not correct code.
3948 68. The host_aton() function was not handling scoped IPv6 addresses (those
3949 with, for example, "%eth0" on the end) correctly.
3951 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3952 negated items (that is, ~something) in unsigned ints. Some compilers
3953 apparently mutter when there is no cast.
3955 70. If an address verification called from an ACL failed, and did not produce a
3956 user-specific message (i.e. there was only a "system" message), nothing was
3957 put in $acl_verify_message. In this situation, it now puts the system
3960 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3961 session; change 4.31/43 added the unwanted input to the log line - except
3962 that it did not do this in the start of session case. It now does.
3964 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3965 This is wrong and can cause the other end to generate a synchronization
3966 error if it is another Exim or anything else that does the synchronization
3967 check. A QUIT command is no longer sent after a timeout.
3969 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3970 during host lookups.
3972 74. The defer_ok option of callout verification was not working if it was used
3973 when verifying addresses in header lines, that is, for this case:
3975 verify = header_sender/callout=defer_ok
3977 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3978 those file descriptors could be used for SMTP connections. If anything
3979 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3980 could be sent to the SMTP client, causing chaos. The daemon now opens
3981 stdin, stdout, and stderr to /dev/null when it puts itself into the
3984 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3985 log by default. The user can override this with suitable Perl magic.
3987 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3988 the log message when discard triggers, was not working for the DATA ACL or
3989 for the non-SMTP ACL.
3991 78. Error message wording change in sieve.c.
3993 79. If smtp_accept_max_per_host was set, the number of connections could be
3994 restricted to fewer than expected, because the daemon was trying to set up
3995 a new connection before checking whether the processes handling previous
3996 connections had finished. The check for completed processes is now done
3997 earlier. On busy systems, this bug wouldn't be noticed because something
3998 else would have woken the daemon, and it would have reaped the completed
4001 80. If a message was submitted locally by a user whose login name contained one
4002 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4003 It caused a spool format error. I have fixed the spool reading code. A
4004 related problem was that the "from" clause in the Received: line became
4005 illegal because of the space(s). It is now covered by ${quote_local_part.
4007 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4010 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4012 83. Patch from Sieve maintainer to fix unterminated string problem in
4013 "vacation" handling.
4015 84. Some minor changes to the Linux configuration files to help with other
4016 OS variants using glibc.
4018 85. One more patch for Sieve to update vacation handling to latest spec.
4021 ----------------------------------------------------
4022 See the note above about the 4.44 and 4.50 releases.
4023 ----------------------------------------------------
4029 1. Change 4.43/35 introduced a bug that caused file counts to be
4030 incorrectly computed when quota_filecount was set in an appendfile
4033 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4034 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4037 3. Give more explanation in the error message when the command for a transport
4038 filter fails to execute.
4040 4. There are several places where Exim runs a non-Exim command in a
4041 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4042 was being done only for the command run by the queryprogram router. It is
4043 now done for all such subprocesses. The other cases are: ${run, transport
4044 filters, and the commands run by the lmtp and pipe transports.
4046 5. Some older OS have a limit of 256 on the maximum number of file
4047 descriptors. Exim was using setrlimit() to set 1000 as a large value
4048 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4049 systems. I've change it so that if it can't get 1000, it tries for 256.
4051 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4052 was an oversight, and furthermore, ever since the addition of extra
4053 controls (e.g. 4.43/32), the checks on when to allow different forms of
4054 "control" were broken. There should now be diagnostics for all cases when a
4055 control that does not make sense is encountered.
4057 7. $recipients is now available in the predata ACL (oversight).
4059 8. Tidy the search cache before the fork to do a delivery from a message
4060 received from the command line. Otherwise the child will trigger a lookup
4061 failure and thereby defer the delivery if it tries to use (for example) a
4062 cached ldap connection that the parent has called unbind on.
4064 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4065 of $address_data from the recipient verification was clobbered by the
4066 sender verification.
4068 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4069 was its contents. (It was OK if the option was not defined at all.)
4071 11. A "Completed" log line is now written for messages that are removed from
4072 the spool by the -Mrm option.
4074 12. $host_address is now set to the target address during the checking of
4075 ignore_target_hosts.
4077 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4078 being passed; this would have caused $sender_host_name to have been used if
4079 matching the list had actually called for a host name (not very likely,
4080 since this list is usually IP addresses). A host name is now passed as
4083 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4084 code that specifies a non-restarting handler (typically sigaction() in
4085 modern systems) in an attempt to fix a rare and obscure crash bug.
4087 15. Narrowed the window for a race in the daemon that could cause it to ignore
4088 SIGCHLD signals. This is not a major problem, because they are used only to
4089 wake it up if nothing else does.
4091 16. A malformed maildirsize file could cause Exim to calculate negative values
4092 for the mailbox size or file count. Odd effects could occur as a result.
4093 The maildirsize information is now recalculated if the size or filecount
4096 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4097 support for a long time. Removed HAVE_SYS_VFS_H.
4099 18. Updated exipick to current release from John Jetmore.
4101 19. Allow an empty sender to be matched against a lookup in an address list.
4102 Previously the only cases considered were a regular expression, or an
4105 20. Exim went into a mad DNS lookup loop when doing a callout where the
4106 host was specified on the transport, if the DNS lookup yielded more than
4109 21. The RFC2047 encoding function was originally intended for short strings
4110 such as real names; it was not keeping to the 75-character limit for
4111 encoded words that the RFC imposes. It now respects the limit, and
4112 generates multiple encoded words if necessary. To be on the safe side, I
4113 have increased the buffer size for the ${rfc2047: expansion operator from
4116 22. Failure to deliver a bounce message always caused it to be frozen, even if
4117 there was an errors_to setting on the router. The errors_to setting is now
4120 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4121 canonical form (fully expanded) before being placed in
4122 $sender_host_address.
4124 24. Updated eximstats to version 1.33
4126 25. Include certificate and key file names in error message when GnuTLS fails
4127 to set them up, because the GnuTLS error message doesn't include the name
4128 of the failing file when there is a problem reading it.
4130 26. Expand error message when OpenSSL has problems setting up cert/key files.
4133 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4134 (this can affect the format of dates).
4136 28. exim_tidydb, when checking for the continued existence of a message for
4137 which it has found a message-specific retry record, was not finding
4138 messages that were in split spool directories. Consequently, it was
4139 deleting retry records that should have stayed in existence.
4141 29. eximstats updated to version 1.35
4142 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4143 1.35 - bugfix such that pie charts by volume are generated correctly
4145 30. The SPA authentication driver was not abandoning authentication and moving
4146 on to the next authenticator when an expansion was forced to fail,
4147 contradicting the general specification for all authenticators. Instead it
4148 was generating a temporary error. It now behaves as specified.
4150 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4151 (the order specifies the preference for clients). The order is now AES256,
4152 AES128, 3DES, ARCFOUR128.
4154 31. Small patch to Sieve code - explicitly set From: when generating an
4157 32. Exim crashed if a remote delivery caused a very long error message to be
4158 recorded - for instance if somebody sent an entire SpamAssassin report back
4159 as a large number of 550 error lines. This bug was coincidentally fixed by
4160 increasing the size of one of Exim's internal buffers (big_buffer) that
4161 happened as part of the Exiscan merge. However, to be on the safe side, I
4162 have made the code more robust (and fixed the comments that describe what
4165 33. Some experimental protocols are using DNS PTR records for new purposes. The
4166 keys for these records are domain names, not reversed IP addresses. The
4167 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4168 leaves it alone. Component reversal etc. now happens only for IP addresses.
4171 34. The host_aton() function is supposed to be passed a string that is known
4172 to be a valid IP address. However, in the case of IPv6 addresses, it was
4173 not checking this. This is a hostage to fortune. Exim now panics and dies
4174 if the condition is not met. A case was found where this could be provoked
4175 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4176 components; fortuitously, this particular loophole had already been fixed
4177 by change 4.50/55 or 4.44/33 above.
4179 If there are any other similar loopholes, the new check in host_aton()
4180 itself should stop them being exploited. The report I received stated that
4181 data on the command line could provoke the exploit when Exim was running as
4182 exim, but did not say which command line option was involved. All I could
4183 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4184 running as the user.
4187 35. There was a buffer overflow vulnerability in the SPA authentication code
4188 (which came originally from the Samba project). I have added a test to the
4189 spa_base64_to_bits() function which I hope fixes it.
4192 36. The daemon start-up calls getloadavg() while still root for those OS that
4193 need the first call to be done as root, but it missed one case: when
4194 deliver_queue_load_max is set with deliver_drop_privilege. This is
4195 necessary for the benefit of the queue runner, because there is no re-exec
4196 when deliver_drop_privilege is set.
4198 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4199 was in use, was not putting the data itself into the right store pool;
4200 consequently, it could be overwritten for a subsequent message in the same
4201 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4204 38. Sometimes the final signoff response after QUIT could fail to get
4205 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4206 < 0 before doing a fflush(). This bug looks as though it goes back to the
4207 introduction of TLS in release 3.20, but "sometimes" must have been rare
4208 because the tests only now provoked it.
4214 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4215 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4216 Now it returns an integer. A number of calls were still expecting a T/F
4217 return. Fortuitously, in all cases, the tests worked in OK situations,
4218 which is the norm. However, things would have gone wrong on any write
4219 failures on the smtp file descriptor. This function is used when sending
4220 messages over SMTP and also when doing verify callouts.
4222 2. When Exim is called to do synchronous delivery of a locally submitted
4223 message (the -odf or -odi options), it no longer closes stderr before doing
4226 3. Implemented the mua_wrapper option.
4228 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4230 5. Implemented the functions header_remove(), header_testname(),
4231 header_add_at_position(), and receive_remove_recipient(), and exported them
4234 6. If an ACL "warn" statement specified the addition of headers, Exim already
4235 inserted X-ACL-Warn: at the start if there was no header name. However, it
4236 was not making this test for the second and subsequent header lines if
4237 there were newlines in the string. This meant that an invalid header could
4238 be inserted if Exim was badly configured.
4240 7. Allow an ACL "warn" statement to add header lines at the start or after all
4241 the Received: headers, as well as at the end.
4243 8. Added the rcpt_4xx retry error code.
4245 9. Added postmaster_mailfrom=xxx to callout verification option.
4247 10. Added mailfrom=xxxx to the callout verification option, for verify=
4250 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4251 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4253 12. Inserted some casts to stop certain compilers warning when using pointer
4254 differences as field lengths or precisions in printf-type calls (mostly
4255 affecting debugging statements).
4257 13. Added optional readline() support for -be (dynamically loaded).
4259 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4260 same clock tick as a message's arrival, so that its received time was the
4261 same as the "first fail" time on the retry record, and that message
4262 remained on the queue past the ultimate address timeout, every queue runner
4263 would try a delivery (because it was past the ultimate address timeout) but
4264 after another failure, the ultimate address timeout, which should have then
4265 bounced the address, did not kick in. This was a "< instead of <=" error;
4266 in most cases the first failure would have been in the next clock tick
4267 after the received time, and all would be well.
4269 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4270 being recognized when the domain list was tested by the match_domain
4271 condition in an expansion string.
4273 16. Added the ${str2b64: operator.
4275 17. Exim was always calling setrlimit() to set a large limit for the number of
4276 processes, without checking whether the existing limit was already
4277 adequate. (It did check for the limit on file descriptors.) Furthermore,
4278 errors from getrlimit() and setrlimit() were being ignored. Now they are
4279 logged to the main and panic logs, but Exim does carry on, to try to do its
4280 job under whatever limits there are.
4282 18. Imported PCRE 5.0.
4284 19. Trivial typo in log message " temporarily refused connection" (the leading
4287 20. If the log selector return_path_on_delivery was set and an address was
4288 redirected to /dev/null, the delivery process crashed because it assumed
4289 that a return path would always be set for a "successful" delivery. In this
4290 case, the whole delivery is bypassed as an optimization, and therefore no
4293 21. Internal re-arrangement: the function for sending a challenge and reading
4294 a response while authentication was assuming a zero-terminated challenge
4295 string. It's now changed to take a pointer and a length, to allow for
4296 binary data in such strings.
4298 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4300 23. Exim was not respecting finduser_retries when seeking the login of the
4301 uid under which it was called; it was always trying 10 times. (The default
4302 setting of finduser_retries is zero.) Also, it was sleeping after the final
4303 failure, which is pointless.
4305 24. Implemented tls_on_connect_ports.
4307 25. Implemented acl_smtp_predata.
4309 26. If the domain in control=submission is set empty, Exim assumes that the
4310 authenticated id is a complete email address when it generates From: or
4311 Sender: header lines.
4313 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4314 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4315 chown and chgrp in /bin and hostname in /usr/bin.
4317 28. Exim was keeping the "process log" file open after each use, just as it
4318 does for the main log. This opens the possibility of it remaining open for
4319 long periods when the USR1 signal hits a daemon. Occasional processlog
4320 errors were reported, that could have been caused by this. Anyway, it seems
4321 much more sensible not to leave this file open at all, so that is what now
4324 29. The long-running daemon process does not normally write to the log once it
4325 has entered its main loop, and it closes the log before doing so. This is
4326 so that log files can straightforwardly be renamed and moved. However,
4327 there are a couple of unusual error situations where the daemon does write
4328 log entries, and I had neglected to close the log afterwards.
4330 30. The text of an SMTP error response that was received during a remote
4331 delivery was being truncated at 512 bytes. This is too short for some of
4332 the long messages that one sometimes sees. I've increased the limit to
4335 31. It is now possible to make retry rules that apply only when a message has a
4336 specific sender, in particular, an empty sender.
4338 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4339 it possible to be selective about when SMTP synchronization is enforced.
4341 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4343 32. Implemented hosts_connection_nolog.
4345 33. Added an ACL for QUIT.
4347 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4350 35. Added mailbox_size and mailbox_filecount to appendfile.
4352 36. Added control = no_multiline_responses to ACLs.
4354 37. There was a bug in the logic of the code that waits for the clock to tick
4355 in the case where the clock went backwards by a substantial amount such
4356 that the microsecond fraction of "now" was more than the microsecond
4357 fraction of "then" (but the whole seconds number was less).
4359 38. Added support for the libradius Radius client library this is found on
4360 FreeBSD (previously only the radiusclient library was supported).
4366 1. When certain lookups returned multiple values in the form name=value, the
4367 quoting of the values was not always being done properly. Specifically:
4368 (a) If the value started with a double quote, but contained no whitespace,
4370 (b) If the value contained whitespace other than a space character (i.e.
4371 tabs or newlines or carriage returns) it was not quoted.
4372 This fix has been applied to the mysql and pgsql lookups by writing a
4373 separate quoting function and calling it from the lookup code. The fix
4374 should probably also be applied to nisplus, ibase and oracle lookups, but
4375 since I cannot test any of those, I have not disturbed their existing code.
4377 2. A hit in the callout cache for a specific address caused a log line with no
4378 reason for rejecting RCPT. Now it says "Previous (cached) callout
4379 verification failure".
4381 3. There was an off-by-one bug in the queryprogram router. An over-long
4382 return line was truncated at 256 instead of 255 characters, thereby
4383 overflowing its buffer with the terminating zero. As well as fixing this, I
4384 have increased the buffer size to 1024 (and made a note to document this).
4386 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4387 when Exim is waiting for an SMTP response from a remote server, Exim
4388 restarts its select() call on the socket, thereby resetting its timeout.
4389 This is not a problem when such interrupts are rare. Somebody set up a cron
4390 job to run exiwhat every 2 minutes, which is less than the normal select()
4391 timeout (5 or 10 minutes). This meant that the select() timeout never
4392 kicked in because it was always reset. I have fixed this by comparing the
4393 time when an interrupt arrives with the time at the start of the first call
4394 to select(). If more time than the timeout has elapsed, the interrupt is
4395 treated as a timeout.
4397 5. Some internal re-factoring in preparation for the addition of Sieve
4398 extensions (by MH). In particular, the "personal" test is moved to a
4399 separate function, and given an option for scanning Cc: and Bcc: (which is
4400 not set for Exim filters).
4402 6. When Exim created an email address using the login of the caller as the
4403 local part (e.g. when creating a From: or Sender: header line), it was not
4404 quoting the local part when it contained special characters such as @.
4406 7. Installed new OpenBSD configuration files.
4408 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4409 try to make them clearer.
4411 9. Callout options, other than the timeout value, were being ignored when
4412 verifying sender addresses in header lines. For example, when using
4414 verify = header_sender/callout=no_cache
4416 the cache was (incorrectly) being used.
4418 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4419 only the Cygwin environment.
4421 11. When return_path_on_delivery was set as a log selector, if different remote
4422 addresses in the same message used different return paths and parallel
4423 remote delivery occurred, the wrong values would sometimes be logged.
4424 (Whenever a remote delivery process finished, the return path value from
4425 the most recently started remote delivery process was logged.)
4427 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4428 lines when AUTH and/or TLS are in use. This is the "received protocol"
4429 field. Exim used to use "asmtp" for authenticated SMTP, without any
4430 indication (in the protocol name) for TLS use. Now it follows the RFC and
4431 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4432 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4433 names appear in log lines as well as in Received: header lines.
4435 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4436 extensions, and comparison tests, and to fix some bugs.
4438 14. Changes to the "personal" filter test:
4440 (1) The test was buggy in that it was just doing the equivalent of
4441 "contains" tests on header lines. For example, if a user's address was
4442 anne@some.where, the "personal" test would incorrectly be true for
4444 To: susanne@some.where
4446 This test is now done by extracting each address from the header in turn,
4447 and checking the entire address. Other tests that are part of "personal"
4448 are now done using regular expressions (for example, to check local parts
4449 of addresses in From: header lines).
4451 (2) The list of non-personal local parts in From: addresses has been
4452 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4453 taken from the Sieve specification recommendations.
4455 (3) If the message contains any header line starting with "List-" it is
4456 treated as non-personal.
4458 (4) The test for "circular" in the Subject: header line has been removed
4459 because it now seems ill-conceived.
4461 15. Minor typos in src/EDITME comments corrected.
4463 16. Installed latest exipick from John Jetmore.
4465 17. If headers_add on a router specified a text string that was too long for
4466 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4467 of string_sprintf() is now avoided.
4469 18. $message_body_size was not set (it was always zero) when running the DATA
4470 ACL and the local_scan() function.
4472 19. For the "mail" command in an Exim filter, no default was being set for
4473 the once_repeat time, causing a random time value to be used if "once" was
4474 specified. (If the value happened to be <= 0, no repeat happened.) The
4475 default is now 0s, meaning "never repeat". The "vacation" command was OK
4476 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4477 (I found it when inspecting the code).
4479 20. There is now an overall timeout for performing a callout verification. It
4480 defaults to 4 times the callout timeout, which applies to individual SMTP
4481 commands during the callout. The overall timeout applies when there is more
4482 than one host that can be tried. The timeout is checked before trying the
4483 next host. This prevents very long delays if there are a large number of
4484 hosts and all are timing out (e.g. when the network connections are timing
4485 out). The value of the overall timeout can be changed by specifying an
4486 additional sub-option for "callout", called "maxwait". For example:
4488 verify = sender/callout=5s,maxwait=20s
4490 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4491 to the end before writing, but this should make it even safer).
4493 22. Exim was forgetting that it had advertised PIPELINING for the second and
4494 subsequent messages on an SMTP connection. It was also not resetting its
4495 memory on STARTTLS and an internal HELO.
4497 23. When Exim logs an SMTP synchronization error within a session, it now
4498 records whether PIPELINING has been advertised or not.
4500 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4501 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4502 rather than long int.
4504 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4506 26. Added the never_mail option to autoreply.
4512 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4513 crash if the getsockname() call failed; this can happen if a connection is
4514 closed very soon after it is established. The problem was simply in the
4515 order in which certain operations were done, causing Exim to try to write
4516 to the SMTP stream before it had set up the file descriptor. The bug has
4517 been fixed by making things happen in the correct order.
4523 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4524 before the connection was closed, thus losing the rejection response.
4526 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4527 some early Solaris releases, but causes trouble in current releases where
4528 socklen_t is defined.
4530 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4533 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4536 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4537 that releasing the top of it at the end releases what was used for sub-
4538 expansions (unless the block got too big). However, discard this block if
4539 the first thing is a variable or header, so that we can use its block when
4540 it is dynamic (useful for very large $message_headers, for example).
4542 6. Lookups now cache *every* query, not just the most recent. A new, separate
4543 store pool is used for this. It can be recovered when all lookup caches are
4544 flushed. Lookups now release memory at the end of their result strings.
4545 This has involved some general refactoring of the lookup sources.
4547 7. Some code has been added to the store_xxx() functions to reduce the amount
4548 of flapping under certain conditions.
4550 8. log_incoming_interface used to affect only the <= reception log lines. Now
4551 it causes the local interface and port to be added to several more SMTP log
4552 lines, for example "SMTP connection from", and rejection lines.
4554 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4556 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4558 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4559 could be overwritten at the end of the current message (or the start of a
4560 new message if it was set in a HELO ACL). The value is now preserved for
4561 the duration of the SMTP connection.
4563 12. If a transport had a headers_rewrite setting, and a matching header line
4564 contained an unqualified address, that address was qualified, even if it
4565 did not match any rewriting rules. The underlying bug was that the values
4566 of the flags that permit the existence of unqualified sender and recipient
4567 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4568 non-local messages, and by -bnq for local messages) were not being
4569 preserved with the message after it was received.
4571 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4572 "next input=" as part of the text comprising the host identity instead of
4573 the correct text. The code was using the same buffer for two different
4574 strings. However, depending on which order the printing function evaluated
4575 its arguments, the bug did not always show up. Under Linux, for example, my
4576 test suite worked just fine.
4578 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4579 doesn't work with some older versions of Perl. It has been changed to "my",
4580 which in any case is probably the better facility to use.
4582 15. A really picky compiler found some instances of statements for creating
4583 error messages that either had too many or two few arguments for the format
4586 16. The size of the buffer for calls to the DNS resolver has been increased
4587 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4588 for addresses that have a lot of PTR records. This alleviates a problem; it
4589 does not fully solve it.
4591 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4592 buffer now truncates the list and logs the incident, which is the same
4593 action as happens when Exim is looking up a host name and its aliases.
4594 Previously in this situation something unpredictable would happen;
4595 sometimes it was "internal error: store_reset failed".
4597 18. If a server dropped the connection unexpectedly when an Exim client was
4598 using GnuTLS and trying to read a response, the client delivery process
4599 crashed while trying to generate an error log message.
4601 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4602 string, for example:
4604 warn message = H1: something\nH2: something
4606 the text was added as a single header line from Exim's point of view
4607 though it ended up OK in the delivered message. However, searching for the
4608 second and subsequent header lines using $h_h2: did not work. This has been
4609 fixed. Similarly, if a system filter added multiple headers in this way,
4610 the routers could not see them.
4612 20. Expanded the error message when iplsearch is called with an invalid key to
4613 suggest using net-iplsearch in a host list.
4615 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4616 ACLs are no longer actually imposed (and a message to that effect is
4619 22. If a "gecos" field in a passwd entry contained escaped characters, in
4620 particular, if it contained a \" sequence, Exim got it wrong when building
4621 a From: or a Sender: header from that name. A second bug also caused
4622 incorrect handling when an unquoted " was present following a character
4623 that needed quoting.
4625 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4626 was not being matched caselessly.
4628 24. Arranged for all hyphens in the exim.8 source to be escaped with
4631 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4632 properly. Recipient callout cache records were still being keyed to include
4633 the sender, even when use_sender was set false. This led to far more
4634 callouts that were necessary. The sender is no longer included in the key
4635 when use_sender is false.
4637 26. Added "control = submission" modifier to ACLs.
4639 27. Added the ${base62d: operator to decode base 62 numbers.
4641 28. dnsdb lookups can now access SRV records.
4643 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4644 the configuration file.
4646 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4647 (-v) mode. This makes the output for a verbose queue run more intelligible.
4649 31. Added a use_postmaster feature to recipient callouts.
4651 32. Added the $body_zerocount variable, containing the number of binary zero
4652 bytes in the message body.
4654 33. The time of last modification of the "new" subdirectory is now used as the
4655 "mailbox time last read" when there is a quota error for a maildir
4658 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4660 35. Added +ignore_unknown as a special item in host lists.
4662 36. Code for decoding IPv6 addresses in host lists is now included, even if
4663 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4664 address was recognized as an IP address, but was then not correctly decoded
4665 into binary, causing unexpected and incorrect effects when compared with
4672 1. Very minor rewording of debugging text in manualroute to say "list of
4673 hosts" instead of "hostlist".
4675 2. If verify=header_syntax was set, and a header line with an unqualified
4676 address (no domain) and a large number of spaces between the end of the
4677 name and the colon was received, the reception process suffered a buffer
4678 overflow, and (when I tested it) crashed. This was caused by some obsolete
4679 code that should have been removed. The fix is to remove it!
4681 3. When running in the test harness, delay a bit after writing a bounce
4682 message to get a bit more predictability in the log output.
4684 4. Added a call to search_tidyup() just before forking a reception process. In
4685 theory, someone could use a lookup in the expansion of smtp_accept_max_
4686 per_host which, without the tidyup, could leave open a database connection.
4688 5. Added the variables $recipient_data and $sender_data which get set from a
4689 lookup success in an ACL "recipients" or "senders" condition, or a router
4690 "senders" option, similar to $domain_data and $local_part_data.
4692 6. Moved the writing of debug_print from before to after the "senders" test
4695 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4696 problems for message scanning, either using a data ACL, or using
4697 local_scan() because the Received: header was not generated till after they
4698 were called (in order to set the time as the time of reception completion).
4699 I have revised the way this works. The header is now generated after the
4700 body is received, but before the ACL or local_scan() are called. After they
4701 are run, the timestamp in the header is updated.
4707 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4708 before starting a queue runner without re-exec. This happened only when
4709 deliver_drop_privilege was set or when the Exim user was set to root. The
4710 effect of the bug was that timeouts during subsequent deliveries caused
4711 crashes instead of being properly handled. The handler is now left at its
4712 default (and expected) setting.
4714 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4715 message, again when deliver_drop_privilege is set or Exim is run as root.
4716 The bug described in (1) was not present in this case, but the tidying up
4717 of the other signals was missing. I have made the two cases consistent.
4719 3. The ignore_target_hosts setting on a manualroute router was being ignored
4720 for hosts that were looked up using the /MX notation.
4722 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4725 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4726 operated on the sender address. After changing the $sender_address to <>
4727 for the sender address verify, Exim was re-instated it as the original
4728 (before rewriting) address, but remembering that it had rewritten it, so it
4729 wasn't rewriting it again. This bug also had the effect of breaking the
4730 sender address verification caching when the sender address was rewritten.
4732 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4733 This has been changed so that if the ip literal address matches
4734 ignore_target_hosts, the router declines.
4736 7. Added expansion conditions match_domain, match_address, and match_local_
4737 part (NOT match_host).
4739 8. The placeholder for the Received: header didn't have a length field set.
4741 9. Added code to Exim itself and to exim_lock to test for a specific race
4742 condition that could lead to file corruption when using MBX delivery. The
4743 issue is with the lockfile that is created in /tmp. If this file is removed
4744 after a process has opened it but before that process has acquired a lock,
4745 there is the potential for a second process to recreate the file and also
4746 acquire a lock. This could lead to two Exim processes writing to the file
4747 at the same time. The added code performs the same test as UW imapd; it
4748 checks after acquiring the lock that its file descriptor still refers to
4749 the same named file.
4751 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4752 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4755 11. Added the smtp_active_hostname option. If used, this will typically be made
4756 to depend on the incoming interface address. Because $interface_address is
4757 not set up until the daemon has forked a reception process, error responses
4758 that can happen earlier (such as "too many connections") no longer contain
4761 12. If an expansion in a condition on a "warn" statement fails because a lookup
4762 defers, the "warn" statement is abandoned, and the next ACL statement is
4763 processed. Previously this caused the whole ACL to be aborted.
4765 13. Added the iplsearch lookup type.
4767 14. Added ident_timeout as a log selector.
4769 15. Added tls_certificate_verified as a log selector.
4771 16. Added a global option tls_require_ciphers (compare the smtp transport
4772 option of the same name). This controls incoming TLS connections.
4774 17. I finally figured out how to make tls_require_ciphers do a similar thing
4775 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4776 before starting the TLS session.
4778 18. Tabs are now shown as \t in -bP output.
4780 19. If the log selector return_path_on_delivery was set, Exim crashed when
4781 bouncing a message because it had too many Received: header lines.
4783 20. If two routers both had headers_remove settings, and the first one included
4784 a superfluous trailing colon, the final name in the first list and the
4785 first name in the second list were incorrectly joined into one item (with a
4786 colon in the middle).
4792 1. Added -C and -D options to the exinext utility, mainly to make it easier
4793 to include in the automated testing, but these could be helpful when
4794 multiple configurations are in use.
4796 2. The exinext utility was not formatting the output nicely when there was
4797 an alternate port involved in the retry record key, nor when there was a
4798 message id as well (for retries that were specific to a specific message
4799 and a specific host). It was also confused by IPv6 addresses, because of
4800 the additional colons they contain. I have fixed the IPv4 problem, and
4801 patched it up to do a reasonable job for IPv6.
4803 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4804 delivery, the log line now contains "pipelined" if PIPELINING was used.
4806 4. An SMTP transport process used to panic and die if the bind() call to set
4807 an explicit outgoing interface failed. This has been changed; it is now
4808 treated in the same way as a connect() failure.
4810 5. A reference to $sender_host_name in the part of a conditional expansion
4811 that was being skipped was still causing a DNS lookup. This no longer
4814 6. The def: expansion condition was not recognizing references to header lines
4815 that used bh_ and bheader_.
4817 7. Added the _cache feature to named lists.
4819 8. The code for checking quota_filecount in the appendfile transport was
4820 allowing one more file than it should have been.
4822 9. For compatibility with Sendmail, the command line option
4830 and sets the incoming protocol and host name (for trusted callers). The
4831 host name and its colon can be omitted when only the protocol is to be set.
4832 Note the Exim already has two private options, -pd and -ps, that refer to
4833 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4834 "s", but I don't think that's a major issue.
4836 10. A number of refactoring changes to the code, none of which should affect
4839 (a) The number of logging options was getting close to filling up the
4840 32-bit word that was used as a bit map. I have split them into two classes:
4841 those that are passed in the argument to log_write(), and those that are
4842 only ever tested independently outside of that function. These are now in
4843 separate 32-bit words, so there is plenty of room for expansion again.
4844 There is no change in the user interface or the logging behaviour.
4846 (b) When building, for example, log lines, the code previously used a
4847 macro that called string_cat() twice, in order to add two strings. This is
4848 not really sufficiently general. Furthermore, there was one instance where
4849 it was actually wrong because one of the argument was used twice, and in
4850 one call a function was used. (As it happened, calling the function twice
4851 did not affect the overall behaviour.) The macro has been replaced by a
4852 function that can join an arbitrary number of extra strings onto a growing
4855 (c) The code for expansion conditions now uses a table and a binary chop
4856 instead of a serial search (which was left over from when there were very
4857 few conditions). Also, it now recognizes conditions like "pam" even when
4858 the relevant support is not compiled in: a suitably worded error message is
4859 given if an attempt is made to use such a condition.
4861 11. Added ${time_interval:xxxxx}.
4863 12. A bug was causing one of the ddress fields not to be passed back correctly
4864 from remote delivery subprocesses. The field in question was not being
4865 subsequently used, so this caused to problems in practice.
4867 13. Added new log selectors queue_time and deliver_time.
4869 14. Might have fixed a bug in maildirsizefile handling that threw up
4870 "unexpected character" debug warnings, and recalculated the data
4871 unnecessarily. In any case, I expanded the warning message to give more
4874 15. Added the message "Restricted characters in address" to the statements in
4875 the default ACL that block characters like @ and % in local parts.
4877 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4878 Three changes have been made:
4880 (a) There was a serious bug; a negative response to MAIL caused the whole
4881 recipient domain to be cached as invalid, thereby blocking all messages
4882 to all local parts at the same domain, from all senders. This bug has
4883 been fixed. The domain is no longer cached after a negative response to
4884 MAIL if the sender used is not empty.
4886 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4889 (c) A new callout option, "use_sender" has been added for people who want
4890 the modified behaviour.
4896 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4899 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4900 indeed breaks things for older releases.
4902 3. Added additional logging to the case where there is a problem reading data
4903 from a filter that is running in a subprocess using a pipe, in order to
4904 try to track down a specific problem.
4906 4. Testing facility fudge: when running in the test harness and attempting
4907 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4908 getting "No route to host". Convert this to a timeout.
4910 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4913 6. Some OS don't have socklen_t but use size_t instead. This affects the
4914 fifth argument of getsockopt() amongst other things. This is now
4915 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4916 can be set for individual OS. I have set it for SunOS5, OSF1, and
4917 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4918 some earlier ones do not.
4920 7. Change 4.30/15 was not doing the test caselessly.
4922 8. The standard form for an IPv6 address literal was being rejected by address
4923 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4924 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4925 this, as well as the form without the "IPv6" on the front (but only when
4926 address literals are enabled, of course).
4928 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4930 10. Exim crashed if a message with an empty sender address specified by -f
4931 encountered a router with an errors_to setting. This could be provoked only
4932 by a command such as
4936 where an empty string was supplied; "<>" did not hit this bug.
4938 11. Installed PCRE release 4.5.
4940 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4941 remained set. It is now erased.
4943 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4944 times from message ids (which are base 36 rather than the normal 62).
4946 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4947 were being counted as actual protocol errors, and logged if the log
4948 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4949 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4950 and DATA following a set of rejected RCPTs do not count as protocol errors.
4951 In other words, Exim assumes they were pipelined, though this may not
4952 actually be the case. Of course, in all cases the client gets an
4953 appropriate error code.
4955 15. If a lookup fails in an ACL condition, a message about the failure may
4956 be available; it is used if testing the ACL cannot continue, because most
4957 such messages specify what the cause of the deferral is. However, some
4958 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4959 that caused an old message to be retained and used if a later statement
4960 caused a defer, replacing the real cause of the deferral.
4962 16. If an IP address had so many PTR records that the DNS lookup buffer
4963 was not large enough to hold them, Exim could crash while trying to process
4964 the truncated data. It now detects and logs this case.
4966 17. Further to 4.21/58, another change has been made: if (and only if) the
4967 first line of a message (the first header line) ends with CRLF, a bare LF
4968 in a subsequent header line has a space inserted after it, so as not to
4969 terminate the header.
4971 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4972 unnecessarily, used atoi() instead of strtol(), and didn't check the
4973 termination when getting file sizes from file names by regex.
4975 19. Completely re-implemented the support for maildirsize files, in the light
4976 of a number of problems with the previous contributed implementation
4977 (4.30/29). In particular:
4979 . If the quota is zero, the maildirsize file is maintained, but no quota is
4982 . If the maildir directory does not exist, it is created before any attempt
4983 to write a maildirsize file.
4985 . The quota value in the file is just a cache; if the quota is changed in
4986 the transport, the new value overrides.
4988 . A regular expression is available for excluding directories from the
4991 20. The autoreply transport checks the characters in options that define the
4992 message's headers; it allows continued headers, but it was checking with
4993 isspace() after an embedded newline instead of explicitly looking for a
4996 21. If all the "regular" hosts to which an address was routed had passed their
4997 expiry times, and had not reached their retry times, the address was
4998 bounced, even if fallback hosts were defined. Now Exim should go on to try
5001 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5002 equivalent code in the SMTP transport. Some hosts send humungous responses
5003 to HELO/EHLO, more than 1024 it seems.
5005 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5006 gives compiler warnings in some environments. I've now done it "properly",
5009 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5010 (because of problems with the built-in one) was declared to return uschar *
5011 instead of char *, causing compiler failure.
5013 25. Fixed a file descriptor leak when processing alias/forward files.
5015 26. Fixed a minor format string issue in dbfn.c.
5017 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5019 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5020 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5023 29. When a sender address is verified, it is cached, to save repeating the test
5024 when there is more than one recipient in a message. However, when the
5025 verification involves a callout, it is possible for different callout
5026 options to be set for different recipients. It is too complicated to keep
5027 track of this in the cache, so now Exim always runs a verification when a
5028 callout is required, relying on the callout cache for the optimization.
5029 The overhead is duplication of the address routing, but this should not be
5032 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5033 to be verified with callout=postmaster, and the main callout worked but the
5034 postmaster check failed, the verification correctly failed. However, if a
5035 subsequent RCPT command asked for sender verification *without* the
5036 postmaster check, incorrect caching caused this verification also to fail,
5039 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5040 it was not caching the DNS options (qualify_single, search_parents) that
5041 were used when the lookup failed. A subsequent lookup with different
5042 options therefore always gave the same answer, though there were cases
5043 where it should not have. (Example: a "domains = !$mx_any" option on a
5044 dnslookup router: the "domains" option is always processed without any
5045 widening, but the router might have qualify_single set.) Now Exim uses the
5046 cached value only when the same options are set.
5048 32. Added John Jetmore's "exipick" utility to the distribution.
5050 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5051 than a timeout (e.g. a certificate is required, and is not provided), an
5052 Exim server now closes the connection immediately. Previously it waited for
5053 the client to close - but if the client is SSL, it seems that they each
5054 wait for each other, leading to a delay before one of them times out.
5056 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5057 maintained 0.8.x compatibility because I don't think many are using it, and
5058 it is clearly obsolete.
5060 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5063 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5064 client certificate was expired. A simple patch fixes this, though I don't
5065 understand the full logic of why the verify callback is called multiple
5068 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5069 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5070 which causes problems with some clients (such as the Certicom SSL Plus
5071 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5072 disables the coutermeasure allowing Eudora to connect."
5074 38. Exim was not checking that a write() to a log file succeeded. This could
5075 lead to Bad Things if a log got too big, in particular if it hit a file
5076 size limit. Exim now panics and dies if it cannot write to a log file, just
5077 as it does if it cannot open a log file.
5079 39. Modified OS/Makefile-Linux so that it now contains
5081 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5083 The two -D definitions ensure that Exim is compiled with large file
5084 support, which makes it possible to handle log files that are bigger than
5087 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5088 instance) a domain was checked against a named list that involved a lookup,
5089 causing $domain_data to be set, then another domain was checked against the
5090 same list, then the first domain was re-checked, the value of $domain_data
5091 after the final check could be wrong. In particular, if the second check
5092 failed, it could be set empty. This bug probably also applied to
5095 41. The strip_trailing_dot option was not being applied to the address given
5096 with the -f command-line option.
5098 42. The code for reading a message's header from the spool was incrementing
5099 $received_count, but never initializing it. This meant that the value was
5100 incorrect (doubled) while delivering a message in the same process in which
5101 it was received. In the most common configuration of Exim, this never
5102 happens - a fresh exec is done - but it can happen when
5103 deliver_drop_privilege is set.
5105 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5106 it now includes up to 150 characters of the unexpected data in the log
5109 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5110 and building data strings. The size of both of these buffers was 10 000
5111 bytes - far larger than anybody would *ever* want, thought I. Needless to
5112 say, somebody hit the limit. I have increased the maximum line length to
5113 20 000 and the maximum data length of concatenated lines to 100 000. I have
5114 also fixed two bugs, because there was no checking on these buffers. Tsk,
5115 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5116 buffer is too small.
5118 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5119 lsearch lookups. Now it does.
5121 46. When parsing a route_list item in a manualroute router, a fixed-length
5122 buffer was used for the list of hosts. I made this 1024 bytes long,
5123 thinking that nobody would ever have a list of hosts that long. Wrong.
5124 Somebody had a whole pile of complicated expansion conditions, and the
5125 string was silently truncated, leading to an expansion error. It turns out
5126 that it is easier to change to an unlimited length (owing to other changes
5127 that have happened since this code was originally written) than to build
5128 structure for giving a limitation error. The length of the item that
5129 expands into the list of hosts is now unlimited.
5131 47. The lsearch lookup could not handle data where the length of text line was
5132 more than 4095 characters. Such lines were truncated, leading to shortened
5133 data being returned. It should now handle lines of any length.
5135 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5136 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5139 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5140 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5141 no longer happens in comment lines. A list of replacements is now placed
5142 at the head of all of the source files, except those whose only change is
5143 to replace PERL_COMMAND in the very first #! line.
5145 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5146 messages on the queue, with a bottom-up merge sort, using code contributed
5147 by Michael Haardt. This should make operations like -bp somewhat faster on
5148 large queues. It won't affect queue runners, except when queue_run_in_order
5151 51. Installed eximstats 1.31 in the distribution.
5153 52. Added support for SRV lookups to the dnslookup router.
5155 53. If an ACL referred to $message_body or $message_body_end, the value was not
5156 reset for any messages that followed in the same SMTP session.
5158 54. The store-handling optimization for building very long strings was not
5159 differentiating between the different store pools. I don't think this
5160 actually made any difference in practice, but I've tidied it.
5162 55. While running the routers to verify a sender address, $sender_address
5163 was still set to the sender address. This is wrong, because when routing to
5164 send a bounce to the sender, it would be empty. Therefore, I have changed
5165 it so that, while verifying a sender address, $sender_address is set to <>.
5166 (There is no change to what happens when verifying a recipient address.)
5168 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5169 target A or AAAA records (if not already returned) without resetting the
5170 qualify_single or search_parents options of the DNS resolver. These are
5171 inappropriate in this case because the targets of MX and SRV records must
5172 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5173 target that, when qualified, matched something in the local domain. These
5174 two options are now turned off when doing these lookups.
5176 57. It seems that at least some releases of Reiserfs (which does not have the
5177 concept of a fixed number of inodes) returns zero and not -1 for the
5178 number of available inodes. This interacted badly with check_spool_inodes,
5179 which assumed that -1 was the "no such thing" setting. What I have done is
5180 to check that the total number of inodes is greater than zero before doing
5181 the test of how many are available.
5183 58. When a "warn" ACL statement has a log_message modifier, the message is
5184 remembered, and not repeated. This is to avoid a lot of repetition when a
5185 message has many recipients that cause the same warning to be written.
5186 Howewer, Exim was preserving the list of already written lines for an
5187 entire SMTP session, which doesn't seem right. The memory is now reset if a
5188 new message is started.
5190 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5191 debugging output unless log_rewrite was also set.
5193 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5194 of (int)(handle) when we know that handle contains (void *)(-1).
5196 61. The Exim daemon panic-logs an error return when it closes the incoming
5197 connection. However "connection reset by peer" seems to be common, and
5198 isn't really an error worthy of noting specially, so that particular error
5201 62. When Exim is trying to find all the local interfaces, it used to panic and
5202 die if the ioctl to get the interface flags failed. However, it seems that
5203 on at least one OS (Solaris 9) it is possible to have an interface that is
5204 included in the list of interfaces, but for which you get a failure error
5205 for this call. This happens when the interface is not "plumbed" into a
5206 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5207 failure of the "get flags" call assumes that the interface is down.
5209 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5210 makes life easier for people who are doing arithmetic on fields extracted
5211 from dates, where you often get leading zeros that should not be
5212 interpreted as octal.
5214 64. Added qualify_domain to the redirect router, to override the global
5217 65. If a pathologically long header line contained very many addresses (the
5218 report of this problem mentioned 10 000) and each of them was rewritten,
5219 Exim could use up a very large amount of memory. (It kept on making new
5220 copies of the header line as it rewrote, and never released the old ones.)
5221 At the expense of a bit more processing, the header rewriting function has
5222 been changed so that it no longer eats memory in this way.
5224 66. The generation of the Received: header has been moved from the time that a
5225 message starts to be received, to the time that it finishes. The timestamp
5226 in the Received: header should now be very close to that of the <= log
5227 line. There are two side-effects of this change:
5229 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5230 logged header lines no longer include the local Received: line, because
5231 it has not yet been created. The same applies to a copy of the message
5232 that is returned to a non-SMTP sender when a message is rejected.
5234 (b) When a filter file is tested using -bf, no additional Received: header
5235 is added to the test message. After some thought, I decided that this
5238 This change does not affect the value of $received_for. It is still set
5239 after address rewriting, but before local_scan() is called.
5241 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5243 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5244 gave an unhelpful panic error message, and a defer error. I have managed to
5245 change this behaviour so that it now rejects any supplied certificate,
5246 which seems right, as the list of acceptable certificates is empty.
5248 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5249 gave an unhelpful defer error. I have not managed to make this reject any
5250 supplied certificates, but the error message it gives is "no certificate
5251 supplied", which is not helpful.
5253 70. exigrep's output now also includes lines that are not associated with any
5254 message, but which match the given pattern. Implemented by a patch from
5255 Martin Sluka, which also tidied up the Perl a bit.
5257 71. Recipient callout verification, like sender verification, was using <> in
5258 the MAIL FROM command. This isn't really the right thing, since the actual
5259 sender may affect whether the remote host accepts the recipient or not. I
5260 have changed it to use the actual sender in the callout; this means that
5261 the cache record is now keyed on a recipient/sender pair, not just the
5262 recipient address. There doesn't seem to be a real danger of callout loops,
5263 since a callout by the remote host to check the sender would use <>.
5264 [SEE ABOVE: changed after hitting problems.]
5266 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5267 temporary errors. However, in the case of such a code being given after
5268 the end of a data transmission (i.e. after ".") Exim was failing to write
5269 a retry record for the message. (Yes, there was some broken host that was
5270 actually sending 8xx at this point.)
5272 73. An unknown lookup type in a host list could cause Exim to panic-die when
5273 the list was checked. (An example that provoked this was putting <; in the
5274 middle of a list instead of at the start.) If this happened during a DATA
5275 ACL check, a -D file could be left lying around. This kind of configuration
5276 error no longer causes Exim to die; instead it causes a defer errror. The
5277 incident is still logged to the main and panic logs.
5279 74. Buglet left over from Exim 3 conversion. The message "too many messages
5280 in one connection" was written to the rejectlog but not the mainlog, except
5281 when address rewriting (yes!) was being logged.
5283 75. Added write_rejectlog option.
5285 76. When a system filter was run not as root (that is, when system_filter_user
5286 was set), the values of the $n variables were not being returned to the
5287 main process; thus, they were not subsequently available in the $sn
5290 77. Added +return_path_on_delivery log selector.
5292 78. A connection timeout was being treated differently from recipients deferred
5293 when testing hosts_max_try with a message that was older than the host's
5294 retry timeout. (The host should not be counted, thus allowing all hosts to
5295 be tried at least once before bouncing.) This may have been the cause of an
5296 occasionally reported bug whereby a message would remain on the queue
5297 longer than the retry timeout, but would be bounced if a delivery was
5298 forced. I say "may" because I never totally pinned down the problem;
5299 setting up timeout/retry tests is difficult. See also the next item.
5301 79. The ultimate address timeout was not being applied to errors that involved
5302 a combination of host plus message (for example, a timeout on a MAIL
5303 command). When an address resolved to a number of possible hosts, and they
5304 were not all tried for each delivery (e.g. because of hosts_max_try), a
5305 message could remain on the queue longer than the retry timeout.
5307 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5310 81. Fixed an obscure SMTP outgoing bug which required at least the following
5311 conditions: (a) there was another message waiting for the same server;
5312 (b) the server returned 5xx to all RCPT commands in the first message so
5313 that the message was not completed; (c) the server dropped the connection
5314 or gave a negative response to the RSET that Exim sends to abort the
5315 transaction. The observed case was a dropped connection after DATA that had
5316 been sent in pipelining mode. That is, the server had advertised PIPELINING
5317 but was not implementing it correctly. The effect of the bug was incorrect
5318 behaviour, such as trying another host, and this could lead to a crash.
5324 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5325 and daemon.c were passed as pointers to ints; they should have been
5326 pointers to socklen_t variables (which are typically unsigned ints).
5328 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5331 3. Fixed a really odd bug that affected only the testing scheme; patching a
5332 certain fixed string in the binary changed the value of another string that
5333 happened to be identical to the end of the original first string.
5335 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5336 name", it returns that address as the IP address. On some operating
5337 systems (e.g. Solaris), it also passes back the IP address string as the
5338 "host name". However, on others (e.g. Linux), it passes back an empty
5339 string. Exim wasn't checking for this, and was changing the host name to an
5340 empty string, assuming it had been canonicized.
5342 5. Although rare, it is permitted to have more than one PTR record for a given
5343 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5344 all the names associated with an address, because they do in Solaris.
5345 However, it seems that they do not in Linux for data that comes from the
5346 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5347 I found this out when I moved to a new Linux workstation and tried to run
5348 the Exim test suite.
5350 To get round this problem I have changed the code so that it now does its
5351 own call to the DNS to look up PTR records when searching for a host name.
5352 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5353 addresses that are only in /etc/hosts are still found.
5355 This behaviour is, however, controlled by an option called host_lookup_
5356 order, which defaults to "bydns:byaddr". If people want to use the other
5357 order, or indeed, just use one or the other means of lookup, they can
5358 specify it in this variable.
5360 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5361 some operating systems, this comes back from gethostbyaddr() as an empty
5362 string, and this is what Exim used to test for. However, it seems that in
5363 other systems, "." is yielded. Exim now tests for this case too.
5365 7. The values of check_spool_space and check_log_space are now held internally
5366 as a number of kilobytes instead of an absolute number of bytes. If a
5367 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5368 kilobyte. This means that much larger values can be stored.
5370 8. Exim monitor: an attempt to get the action menu when not actually pointing
5371 at a message produces an empty menu entitled "No message selected". This
5372 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5373 no entries in it ("Shell widget menu has zero width and/or height"). So I
5374 have added a single, blank menu entry in this case.
5376 9. Added ${quote_local_part.
5378 10. MIME decoding is now applied to the contents of Subject: header lines when
5381 11. Now that a reference to $sender_host_address automatically causes a reverse
5382 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5383 host lookup before query-style lookups in lists that might use this
5384 variable. This has therefore been abolished, and the "net-" prefix is no
5385 longer necessary for query-style lookups.
5387 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5388 have been a typo for LFLAGS, so it has been changed.
5390 13. The install script calls Exim with "-C /dev/null" in order to find the
5391 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5392 to be output. Howeve, since Exim outputs its version number before the
5393 error, it didn't break the script. It just looked ugly. I fixed this by
5394 always allowing "-C /dev/null" if the caller is root.
5396 14. Ignore overlarge ACL variable number when reading spool file - insurance
5397 against a later release with more variables having written the file.
5399 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5400 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5401 this, as well as the form without the "IPv6" on the front.
5403 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5404 OS/Makefile-Darwin file.
5406 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5407 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5409 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5410 variable. However, some OS use ints for inodes. I've added cast to long int
5411 to get rid of the compiler warning.
5413 19. I had forgotten to lock out "/../" in configuration file names when
5414 ALT_CONFIG_PREFIX was set.
5416 20. Routers used for verification do not need to specify transports. However,
5417 if such a router generated a host list, and callout was configured, Exim
5418 crashed, because it could not find a port number from the (non-existent)
5419 transport. It now assumes port 25 in this circumstance.
5421 21. Added the -t option to exigrep.
5423 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5424 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5425 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5426 them is set, LOOKUP_LSEARCH is forced.
5428 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5430 24. Added sender and host information to the "rejected by local_scan()" log
5431 line; previously there was no indication of these.
5433 25. Added .include_if_exists.
5435 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5436 Linux. It turns out that not all file systems support this. Apparently some
5437 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5438 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5439 not supported on the file descriptor, is now ignored when Exim is trying to
5440 sync a directory. This applies only to Linux.
5442 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5444 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5445 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5448 29. Added support for maildirsize files from supplied patch (modified a bit).
5450 30. The use of :fail: followed by an empty string could lead Exim to respond to
5451 sender verification failures with (e.g.):
5453 550 Verification failed for <xxx>
5454 550 Sender verify failed
5456 where the first response line was missing the '-' that indicates it is not
5457 the final line of the response.
5459 31. The loop for finding the name of the user that called Exim had a hardwired
5460 limit of 10; it now uses the value of finduser_retries, which is used for
5461 all other user lookups.
5463 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5466 33. Exim was neglecting to zero errno before one call of strtol() when
5467 expanding a string and expecting an integer value. On some systems this
5468 resulted in spurious "integer overflow" errors. Also, it was casting the
5469 result into an int without checking.
5471 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5472 did not work. The code looks as if it has *never* worked, though it appears
5473 to have been documented since at least releast 1.62. I have made it work.
5475 35. The "timeout_DNS" error in retry rules, also documented since at least
5476 1.62, also never worked. As it isn't clear exactly what this means, and
5477 clearly it isn't a major issue, I have abolished the feature by treating it
5478 as "timeout", and writing a warning to the main and panic logs.
5480 36. The display of retry rules for -brt wasn't always showing the error code
5483 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5484 timeout_connect_A, timeout_connect_MX.
5486 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5487 to the empty sender.
5489 39. The daemon was not analysing the content of -oX till after it had closed
5490 stderr and disconnected from the controlling terminal. This meant that any
5491 syntax errors were only noted on the panic log, and the return code from
5492 the command was 0. By re-arranging the code a little, I've made the
5493 decoding happen first, so such errors now appear on stderr, and the return
5494 code is 1. However, the actual setting up of the sockets still happens in
5495 the disconnected process, so errors there are still only recorded on the
5498 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5499 connections (as happens on some IP stacks) was logged at start up time as
5500 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5501 it from "IPv6 and IPv4", which means that two separate sockets are being
5504 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5505 says whether AF_INET or AF_INET6 was passed as an argument.
5507 42. Exiwhat output was messed up when time zones were included in log
5510 43. Exiwhat now gives more information about the daemon's listening ports,
5511 and whether -tls-on-connect was used.
5513 44. The "port" option of the smtp transport is now expanded.
5515 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5516 silently ignored. Now an error message is written to the main and panic
5519 46. There's a new ACL modifier called "logwrite" which writes to a log file
5520 as soon as it is encountered.
5522 47. Added $local_user_uid and $local_user_gid at routing time.
5524 48. Exim crashed when trying to verify a sender address that was being
5527 49. Exim was recognizing only a space character after ".include". It now also
5528 recognizes a tab character.
5530 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5531 extracting the relevant information from the specification. The man page no
5532 longer contains scrambled data for the -d option, and I've added a section
5533 at the front about calling Exim under different names.
5535 51. Added "extra_headers" argument to the "mail" command in filter files.
5537 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5540 53. Installed eximstats 1.29.
5542 54. Added transport_filter_timeout as a generic transport option.
5544 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5545 Cc: header lines. This was required by RFC 822, but it not required by RFC
5548 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5549 incoming messages that did not have them. Now it does so only if the
5550 message originates locally, that is, if there is no associated remote host
5551 address. When Resent- header lines are present, this applies to the Resent-
5552 lines rather than the non-Resent- lines.
5554 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5555 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5557 58. Messages for configuration errors now include the name of the main
5558 configuration file - useful now that there may be more than one file in a
5559 list (.included file names were always shown).
5561 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5562 for those rare installations that do not start the daemon as root or run it
5563 setuid root. I've cut out the call to initgroups() if the daemon is not
5566 60. The Exim user and group can now be bound into the binary as text strings
5567 that are looked up at the start of Exim's processing.
5569 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5571 62. Added $mailstore_basename variable.
5573 63. Installed patch to sieve.c from Michael Haardt.
5575 64. When Exim failed to open the panic log after failing to open the main log,
5576 the original message it was trying to log was written to stderr and debug
5577 output, but if they were not available (the usual case in production), it
5578 was lost. Now it is written to syslog before the two lines that record the
5579 failures to open the logs.
5581 65. Users' Exim filters run in subprocesses under the user's uid. It is
5582 possible for a "deliver" command or an alias in a "personal" command to
5583 provoke an address rewrite. If logging of address rewriting is configured,
5584 this fails because the process is not running as root or exim. There may be
5585 a better way of dealing with this, but for the moment (because 4.30 needs
5586 to be released), I have disabled address rewrite logging when running a
5587 filter in a non-root, non-exim process.
5593 1. The buildconfig auxiliary program wasn't quoting the value set for
5594 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5595 not defined. This bug was masked in 4.22 by the effect that was fixed in
5598 2. Some messages that were rejected after a message id was allocated were
5599 shown as "incomplete" by exigrep. It no longer does this for messages that
5600 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5602 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5603 have allow_domain_literals set, the ID did not get logged in the <= line.
5604 Domain literals are now always recognized in Message-ID: header lines.
5606 4. The first argument for a ${extract expansion item is the key name or field
5607 number. Leading and trailing spaces in this item were not being ignored,
5608 causing some misleading effects.
5610 5. When deliver_drop_privilege was set, single queue runner processes started
5611 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5612 same command in the process it spins off) were not dropping privilege.
5614 6. When the daemon running as "exim" started a queue runner, it always
5615 re-executed Exim in the spun-off process. This is a waste of effort when
5616 deliver_drop_privilege is set. The new process now just calls the
5617 queue-runner function directly.
5623 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5626 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5627 ignored. Though the use of -oP was forcing the writing of a pid file, it
5628 was always written to the default place.
5630 3. If the message "no IP address found for host xxxx" is generated during
5631 incoming verification, it is now followed by identification of the incoming
5632 connection (so you can more easily find what provoked it).
5634 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5636 5. Added some features to "harden" Exim a bit more against certain attacks:
5638 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5639 be put in Local/Makefile. This is like the never_users runtime option,
5640 but it cannot be overridden. The default setting is "root".
5642 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5643 prefix string with which any file named in a -C command line option
5646 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5647 is retained for -C and -D only if the caller of Exim is root. Without
5648 it, the exim user may also use -C and -D and retain privilege.
5650 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5651 command line option is disabled.
5653 6. Macro names set by the -D option must start with an upper case letter, just
5654 like macro names defined in the configuration file.
5656 7. Added "dereference=" facility to LDAP.
5658 8. Two instances of the typo "uknown" in the source files are fixed.
5660 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5661 the Configure-Makefile script screwed up while processing it.
5663 10. Incorporated PCRE 4.4.
5665 11. The SMTP synchronization check was not operating right at the start of an
5666 SMTP session. For example, it could not catch a HELO sent before the client
5667 waited for the greeting. There is now a check for outstanding input at the
5668 point when the greeting is written. Because of the duplex, asynchronous
5669 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5670 way, but not yet received, when the check is performed.
5672 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5673 on TCP/IP sockets, because this apparently causes some broken clients to
5676 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5677 unchanged) from the Cygwin maintainer.
5679 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5680 was supported instead of testing for mbx. Effectively a typo.
5682 15. The spa authenticator server code was not checking that the input it
5683 received was valid base64.
5685 16. The debug output line for the "set" modifier in ACLs was not showing the
5686 name of the variable that was being set.
5688 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5690 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5691 lookup on its own. Something else was needed to trigger the lookup. For
5692 example, a match in host_lookup or the need for a host name in a host list.
5693 Now, if $sender_host_name is referenced and the host name has not yet been
5694 looked up, a lookup is performed. If the lookup fails, the variable remains
5695 empty, and $host_lookup_failed is set to "1".
5697 19. Added "eqi" as a case-independent comparison operator.
5699 20. The saslauthd authentication condition could segfault if neither service
5700 nor realm was specified.
5702 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5703 error message that was logged was misleading, and incoming SMTP
5704 connections were dropped. The message is now more accurate, and temporary
5705 errors are given to SMTP connections.
5707 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5708 (except RSET) with a 421 error, until QUIT is received. However, it was
5709 failing to send a response to QUIT.
5711 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5712 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5713 ACL is now run after the helo_try_verify_hosts code.
5715 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5716 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5717 case-independent, so other case variants are also recognized). Apparently
5718 some systems use these upper case variants.
5720 25. If more than two messages were waiting for the same host, and a transport
5721 filter was specified for the transport, Exim sent two messages over the
5722 same TCP/IP connection, and then failed with "socket operation on non-
5723 socket" when it tried to send the third.
5725 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5727 27. The extern definition of crypt16() in expand.c was not being excluded when
5728 the OS had its own crypt16() function.
5730 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5731 as a preferred synonym for return_size_limit, both as an option and as an
5734 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5736 30. Changed the default configuration ACL to relax the local part checking rule
5737 for addresses that are not in any local domains. For these addresses,
5738 slashes and pipe symbols are allowed within local parts, but the sequence
5739 /../ is explicitly forbidden.
5741 31. SPA server authentication was not clearing the challenge buffer before
5744 32. log_message in a "warn" ACL statement was writing to the reject log as
5745 well as to the main log, which contradicts the documentation and doesn't
5746 seem right (because no rejection is happening). So I have stopped it.
5748 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5749 However, I am unable to do any testing of this.
5751 34. Fixed an infelicity in the appendfile transport. When checking directories
5752 for a mailbox, to see if any needed to be created, it was accidentally
5753 using path names with one or more superfluous leading slashes; tracing
5754 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5756 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5757 contents are added to the log line that is written for every discarded
5758 recipient. (Previously a log_message setting was ignored.)
5760 36. The ${quote: operator now quotes the string if it is empty.
5762 37. The install script runs exim in order to find its version number. If for
5763 some reason other than non-existence or emptiness, which it checks, it
5764 could not run './exim', it was installing it with an empty version number,
5765 i.e. as "exim-". This error state is now caught, and the installation is
5768 38. An argument was missing from the function that creates an error message
5769 when Exim fails to connect to the socket for saslauthd authentication.
5770 This could cause Exim to crash, or give a corrupted message.
5772 39. Added isip, isip4, and isip6 to ${if conditions.
5774 40. The ACL variables $acl_xx are now saved with the message, and can be
5775 accessed later in routers, transports, and filters.
5777 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5778 strings in the file are not string-expanded.
5780 42. If a MAIL command specified a SIZE value that was too large to fit into an
5781 int variable, the check against message_size_limit failed. Such values are
5782 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5783 day this will have to be increased, but I don't think I want to be around
5784 when emails are that large.
5791 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5792 iconv() is not standard in FreeBSD.
5794 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5795 IPv6 enabled. The observed symptom was a segmentation fault on return from
5796 the function os_common_find_running_interfaces() in src/os.c.
5798 3. In the check_special_case() function in daemon.c I had used "errno" as an
5799 argument name, which causes warnings on some systems. This was basically a
5800 typo, since it was named "eno" in the comments!
5802 4. The code that waits for the clock to tick (at a resolution of some fraction
5803 of a second) so as to ensure message-id uniqueness was always waiting for
5804 at least one whole tick, when it could have waited for less. [This is
5805 almost certainly not relevant at current processor speeds, where it is
5806 unlikely to ever wait at all. But we try to future-proof.]
5808 5. The function that sleeps for a time interval that includes fractions of a
5809 second contained a race. It did not block SIGALRM between setting the
5810 timer, and suspending (a couple of lines later). If the interval was short
5811 and the sigsuspend() was delayed until after it had expired, the suspension
5812 never ended. On busy systems this could lead to processes getting stuck for
5815 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5816 process, before it forks any delivery processes. The open lookup caching
5817 mechanism meant that the open file or database connection was passed into
5818 the delivery process. The problem was that delivery processes always tidy
5819 up cached lookup data. This could cause a problem for the next delivery
5820 process started by the queue runner, because the external queue runner
5821 process does not know about the closure. So the next delivery process
5822 still has data in the lookup cache. In the case of a file lookup, there was
5823 no problem because closing a file descriptor in a subprocess doesn't affect
5824 the parent. However, if the lookup was caching a connection to a database,
5825 the connection was closed, and the second delivery process was likely to
5826 see errors such as "PGSQL: query failed: server closed the connection
5827 unexpectedly". The problem has been fixed by closing all cached lookups
5828 in a queue runner before running a delivery process.
5830 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5831 seem to have the "const" qualifier which it has on other OS. I've
5834 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5835 *of the same type* (client or server) with the same public name that an
5836 error should be diagnosed.
5838 9. When Exim looked up a host name for an IP address, but failed to find the
5839 original IP address when looking up the host name (a safety check), it
5840 output the message "<ip address> does not match any IP for NULL", which was
5841 confusing, to say the least. The bug was that the host name should have
5842 appeared instead of "NULL".
5844 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5845 user that is built into the binary, and the -C or -D options is used, root
5846 privilege is dropped before the configuration file is read. In addition,
5847 logging is switched to stderr instead of the normal log files. If the
5848 configuration then re-defines the Exim user, the unprivileged environment
5849 is probably not what is expected, so Exim logs a panic warning message (but
5852 However, if deliver_drop_privilege is set, the unprivileged state may well
5853 be exactly what is intended, so the warning has been cut out in that case,
5854 and Exim is allowed to try to write to its normal log files.
5860 1. smtp_return_error_details was not giving details for temporary sender
5861 or receiver verification errors.
5863 2. Diagnose a configuration error if two authenticators have the same public
5866 3. Exim used not to create the message log file for a message until the first
5867 delivery attempt. This could be confusing when incoming messages were held
5868 for policy or load reasons. The message log file is now created at the time
5869 the message is received, and an initial "Received" line is written to it.
5871 4. The automatically generated man page for command line options had a minor
5872 bug that caused no ill effects; however, a more serious problem was that
5873 the procedure for building the man page automatically didn't always
5874 operate. Consequently, release 4.20 contains an out-of-date version. This
5875 shouldn't happen again.
5877 5. When building Exim with embedded Perl support, the script that builds the
5878 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5879 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5881 6. The freeze_tell option was not being used for messages that were frozen on
5882 arrival, either by an ACL or by local_scan().
5884 7. Added the smtp_incomplete_transaction log selector.
5886 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5887 was accepting AUTH without a new EHLO.
5889 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5890 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5893 10. Logging of TCP/IP connections (when configured) now happens in the main
5894 daemon process instead of the child process, so that the TCP/IP connection
5895 count is more accurate (but it can never be perfect).
5897 11. The use of "drop" in a nested ACL was not being handled correctly in the
5898 outer ACL. Now, if condition failure induced by the nested "drop" causes
5899 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5900 or "require"), the connection is dropped.
5902 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5903 that yield "discard" can now be used with an "accept" or a "discard" verb,
5904 but an error is generated for any others (because I can't see a useful way
5905 to define what should happen).
5907 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5908 are now processed in the same way as lines in the Exim configuration file.
5909 In particular, continuation lines are supported.
5911 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5913 15. Added -ti meaning -t -i.
5915 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5916 domains, and warn by logging if others are found.
5918 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5919 returned from GIFCONF when Exim is trying to find the list of interfaces on
5920 a host. The code in os.c has been modified to copy each ifreq to an aligned
5921 structure in all cases.
5923 Also, in some cases, the returned ifreq's were being copied to a 'struct
5924 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5925 means the last couple of bytes of an IPv6 address could be chopped if the
5926 ifreq contained only a normal sockaddr (14 bytes storage).
5928 18. Named domain lists were not supported in the hosts_treat_as_local option.
5929 An entry such as +xxxx was not recognized, and was treated as a literal
5932 19. Ensure that header lines added by a DATA ACL are included in the reject log
5933 if the ACL subsequently rejects the message.
5935 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5936 MD5 (which is deprecated).
5938 21. When testing a filter file using -bf, Exim was writing a message when it
5939 took the sender from a "From " line in the message, but it was not doing so
5940 when it took $return_path from a Return-Path: header line. It now does.
5942 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5943 with a valid header line field name (a series of printing characters
5944 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5946 23. Changed "disc" in the source to "disk" to conform to the documentation and
5947 the book and for uniformity.
5949 24. Ignore Sendmail's -Ooption=value command line item.
5951 25. When execve() failed while trying to run a command in a pipe transport,
5952 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5953 could be confused with a return value of 69 from the command itself. This
5954 has been changed to 127, the value the shell returns if it is asked to run
5955 a non-existent command. The wording for the related log line suggests a
5956 non-existent command as the problem.
5958 26. If received_header_text expands to an empty string, do not add a Received:
5959 header line to the message. (Well, it adds a token one on the spool, but
5960 marks it "old" so that it doesn't get used or transmitted.)
5962 27. Installed eximstats 1.28 (addition of -nt option).
5964 28. There was no check for failure on the call to getsockname() in the daemon
5965 code. This can fail if there is a shortage of resources on the system, with
5966 ENOMEM, for example. A temporary error is now given on failure.
5968 29. Contrary to the C standard, it seems that in some environments, the
5969 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5970 program. Exim now does this explicitly; it affects the formatting of
5971 timestamps using strftime().
5973 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5974 complaints. I've now initialized all the variables, to avoid this.
5976 32. Header lines added by a system filter were not being "seen" during
5977 transport-time rewrites.
5979 33. The info_callback() function passed to OpenSSL is set up with type void
5980 (*)(SSL *, int, int), as described somewhere. However, when calling the
5981 function (actually a macro) that sets it up, the type void(*)() is
5982 expected. I've put in a cast to prevent warnings from picky compilers.
5984 34. If a DNS black list lookup found a CNAME record, but there were no A
5985 records associated with the domain it pointed at, Exim crashed.
5987 35. If a DNS black list lookup returned more than one A record, Exim ignored
5988 all but the first. It now scans all returned addresses if a particular IP
5989 value is being sought. In this situation, the contents of the
5990 $dnslist_value variable are a list of all the addresses, separated by a
5993 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5994 to accept a host name and all its aliases if the forward lookup for any of
5995 them yielded the IP address of the incoming connection. Now it accepts only
5996 those names whose forward lookup yields the correct IP address. Any other
5997 names are discarded. This closes a loophole whereby a rogue DNS
5998 administrator could create reverse DNS records to break through a
5999 wildcarded host restriction in an ACL.
6001 37. If a user filter or a system filter that ran in a subprocess used any of
6002 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6003 the wrong values were passed to the pipe command ($thisaddress had the
6004 value of $0, $0 had the value of $1, etc). This bug was introduced by
6005 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6007 38. Improved the line breaking for long SMTP error messages from ACLs.
6008 Previously, if there was no break point between 40 and 75 characters, Exim
6009 left the rest of the message alone. Two changes have been made: (a) I've
6010 reduced the minimum length to 35 characters; (b) if it can't find a break
6011 point between 35 and 75 characters, it looks ahead and uses the first one
6012 that it finds. This may give the occasional overlong line, but at least the
6013 remaining text gets split now.
6015 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6016 file descriptors might be low, and that setting 1000 would always raise it.
6017 It turns out that in some environments, the limit is already over 1000 and
6018 that lowering it causes trouble. So now Exim takes care not to decrease it.
6020 40. When delivering a message, the value of $return_path is set to $sender_
6021 address at the start of routing (routers may change the value). By an
6022 oversight, this default was not being set up when an address was tested by
6023 -bt or -bv, which affected the outcome if any router or filter referred to
6026 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6027 log only when "message" or "log_message" are set. However, if one of the
6028 conditions was an address verification, or a call to a nested ACL, the
6029 messages generated by the underlying test were being passed through. This
6030 no longer happens. The underlying message is available in $acl_verify_
6031 message for both "message" and "log_message" expansions, so it can be
6032 passed through if needed.
6034 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6035 new expansion $bh_ to give the encoded byte string without charset
6036 translation. Translation happens only if iconv() is available; HAVE_ICONV
6037 indicates this at build time. HEADERS_CHARSET gives the charset to
6038 translate to; headers_charset can change it in the configuration, and
6039 "headers charset" can change it in an individual filter file.
6041 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6042 that creates RFC 2047 encoded "words" labels them as that charset instead
6043 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6044 expansion operator; (ii) when Exim creates a From: line for a local
6045 message; (iii) when a header line is rewritten to include a "phrase" part.
6047 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6048 buggy, causing it to skip the first lines of messages whose message ID
6049 ended in 'D'. This would not have bitten before Exim release 4.14, because
6050 message IDs were unlikely to end in 'D' before then. The effect was to have
6051 incorrect size information for certain domains.
6053 45. #include "config.h" was missing at the start of the crypt16.c module. This
6054 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6057 46. If there was a timeout during a "random" callout check, Exim treated it as
6058 a failure of the random address, and carried on sending RSET and the real
6059 address. If the delay was just some slowness somewhere, the response to the
6060 original RCPT would be taken as a response to RSET and so on, causing
6061 mayhem of various kinds.
6063 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6064 when I implemented it. It didn't allow for the fact that some option values
6065 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6066 the right test for positive values.
6068 48. Domain names in DNS records are case-independent. Exim always looks them up
6069 in lower case. Some resolvers return domain names in exactly the case they
6070 appear in the zone file, that is, they may contain uppercase letters. Not
6071 all resolvers do this - some return always lower case. Exim was treating a
6072 change of case by a resolver as a change of domain, similar to a widening
6073 of a domain abbreviation. This triggered its re-routing code and so it was
6074 trying to route what was effectively the same domain again. This normally
6075 caused routing to fail (because the router wouldn't handle the domain
6076 twice). Now Exim checks for this case specially, and just changes the
6077 casing of the domain that it ultimately uses when it transmits the message
6080 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6083 50. If a filter generated a file delivery with a non-absolute name (possible if
6084 no home directory exists for the router), the forbid_file option was not
6087 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6088 the existing equality matching.
6090 52. Exim was using ints instead of ino_t variables in some places where it was
6091 dealing with inode numbers.
6093 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6094 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6095 called TMPDIR, and if it finds it is different, it changes its value.
6097 54. The smtp_printf() function is now made available to local_scan() so
6098 additional output lines can be written before returning. There is also an
6099 smtp_fflush() function to enable the detection of a dropped connection.
6100 The variables smtp_input and smtp_batched_input are exported to
6103 55. Changed the default runtime configuration: the message "Unknown user"
6104 has been removed from the ACL, and instead placed on the localuser router,
6105 using the cannot_route_message feature. This means that any verification
6106 failures that generate their own messages won't get overridden. Similarly,
6107 the "Unrouteable address" message that was in the ACL for unverifiable
6108 relay addresses has also been removed.
6110 56. Added hosts_avoid_esmtp to the smtp transport.
6112 57. The exicyclog script was not checking for the esoteric option
6113 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6114 will work only if exicyclog is run under the appropriate euid.
6116 58. Following a discussion on the list, the rules by which Exim recognises line
6117 endings on incoming messages have been changed. The -dropcr and drop_cr
6118 options are now no-ops, retained only for backwards compatibility. The
6119 following line terminators are recognized: LF CRLF CR. However, special
6120 processing applies to CR:
6122 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6123 nor a local message in the state where . is a terminator.
6125 (ii) If a bare CR is encountered in a header line, an extra space is added
6126 after the line terminator so as not to end the header. The reasoning
6127 behind this is that bare CRs in header lines are most likely either
6128 to be mistakes, or people trying to play silly games.
6130 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6131 was being incorrectly given as 18 bytes larger than it should have been.
6132 This is a VOB (very old bug).
6134 60. This may never have affected anything current, but just in case it has:
6135 When the local host is found other than at the start of a list of hosts,
6136 the local host, those with the same MX, and any that follow, are discarded.
6137 When the list in question was part of a longer list of hosts, the following
6138 hosts (not currently being processed) were also being discarded. This no
6139 longer happens. I'm not sure if this situation could ever has previously
6142 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6145 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6146 header. This is something that is recommended in a new Internet Draft, and
6147 is something that is documented as being done by Sendmail. There are two
6148 possible values. For messages generated by the autoreply transport, Exim
6151 Auto-Submitted: auto-replied
6153 whereas for all other generated messages (e.g. bounces) it adds
6155 Auto-Submitted: auto-generated
6157 63. The "personal" condition in filters now includes a test for the
6158 Auto-Submitted: header. If it contains the string "auto-" the message it
6159 not considered personal.
6161 64. Added rcpt_include_affixes as a generic transport option.
6163 65. Added queue_only_override (default true).
6165 66. Added the syslog_duplication option.
6167 67. If what should have been the first header line of a message consisted of
6168 a space followed by a colon, Exim was mis-interpreting it as a header line.
6169 It isn't of course - it is syntactically invalid and should therefore be
6170 treated as the start of the message body. The misbehaviour could have
6171 caused a number of strange effects, including loss of data in subsequent
6172 header lines, and spool format errors.
6174 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6175 client host had authenticated. This control can now be exercised by an ACL
6176 for more flexibility.
6178 69. By default, callouts do not happen when testing with -bh. There is now a
6179 variant, -bhc, which does actually run the callout code, including
6180 consulting and updating the callout cache.
6182 70. Added support for saslauthd authentication, courtesy of Alexander
6185 71. If statvfs() failed on the spool or log directories while checking their
6186 size for availability, Exim confusingly gave the error "space shortage".
6187 Furthermore, in debugging mode it crashed with a floating point exception.
6188 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6189 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6190 serious problem, Exim now writes to the main and panic logs when this
6191 happens, with details of the failure. It then refuses to accept the
6192 incoming message, giving the message "spool directory problem" or "log
6193 directory problem" with a 421 code for SMTP messages.
6195 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6196 0, 1, and 2 exist, because some OS complain for execs without them (see
6197 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6198 descriptors. However, the code omitted to check that the open succeeded,
6199 causing mysterious errors if for some reason the permissions on /dev/null
6200 got screwed. Now Exim writes a message to the main and panic logs, and
6201 bombs out if it can't open /dev/null.
6203 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6204 interact so that it is all more flexible. It is supposed to remain
6205 backwards compatible. Also added extra_local_interfaces.
6207 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6208 to bomb out with an assertion failure - to the client this appears as a
6209 connection drop. This problem occurs in the part of the code that was taken
6210 from the Samba project. Fortunately, the assertion is in a very simple
6211 function, so I have fixed this by reproducing the function inline in the
6212 one place where it is called, and arranging for authentication to fail
6213 instead of killing the process with assert().
6215 75. The SPA client code was not working when the server requested OEM rather
6216 than Unicode encoding.
6218 76. Added code to make require_files with a specific uid setting more usable in
6219 the case where statting the file as root fails - usually a non-root-mounted
6220 NFS file system. When this happens and the failure is EACCES, Exim now
6221 forks a subprocess and does the per-uid checking as the relevant uid.
6223 77. Added process_log_path.
6225 78. If log_file_path was not explicitly set, a setting of check_log_space or
6226 check_log_inodes was ignored.
6228 79. If a space check for the spool or log partitions fails, the incident is now
6229 logged. Of course, in the latter case the data may get lost...
6231 80. Added the %p formatting code to string_format() so that it can be used to
6232 print addresses in debug_print(). Adjusted all the address printing in the
6233 debugging in store.c to use %p rather than %d.
6235 81. There was a concern that a line of code in smtp_in.c could overflow a
6236 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6237 initially expressed, the concern was not well-founded, because trailing
6238 spaces are removed early. However, if the trailing spaces were followed by
6239 a NULL, they did not get removed, so the overflow was possible. Two fixes
6242 (a) I re-wrote the offending code in a cleaner fashion.
6243 (b) If an incoming SMTP command contains a NULL character, it is rejected
6246 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6247 runs initgroups(), so that if the Exim user is in any additional groups,
6248 they will be used during message reception.
6254 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob