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18 EXPERIMENTAL_CERTNAMES is withdrawn.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
73 JH/20 BUG 1571: Ensure that $tls_in_peerdn is set, when verification fails
80 TL/01 When running the test suite, the README says that variables such as
81 no_msglog_check are global and can be placed anywhere in a specific
82 test's script, however it was observed that placement needed to be near
83 the beginning for it to behave that way. Changed the runtest perl
84 script to read through the entire script once to detect and set these
85 variables, reset to the beginning of the script, and then run through
86 the script parsing/test process like normal.
88 TL/02 The BSD's have an arc4random API. One of the functions to induce
89 adding randomness was arc4random_stir(), but it has been removed in
90 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
91 function when detected.
93 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
94 cause callback expansion.
96 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
97 syntax errors in an expansion can be treated as a string instead of
98 logging or causing an error, due to the internal use of bool_lax
99 instead of bool when processing it.
101 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
102 server certificates when making smtp deliveries.
104 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
106 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
108 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
110 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
111 Merged patch from Sebastian Wiedenroth.
113 JH/05 Fix results-pipe from transport process. Several recipients, combined
114 with certificate use, exposed issues where response data items split
115 over buffer boundaries were not parsed properly. This eventually
116 resulted in duplicates being sent. This issue only became common enough
117 to notice due to the introduction of conection certificate information,
118 the item size being so much larger. Found and fixed by Wolfgang Breyha.
120 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
121 size buffer was used, resulting in syntax errors when an expansion
124 JH/07 Add support for directories of certificates when compiled with a GnuTLS
125 version 3.3.6 or later.
127 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
128 is EXPERIMENTAL_EVENT, the main-configuration and transport options
129 both become "event_action", the variables become $event_name, $event_data
130 and $event_defer_errno. There is a new variable $verify_mode, usable in
131 routers, transports and related events. The tls:cert event is now also
132 raised for inbound connections, if the main configuration event_action
135 TL/06 In test suite, disable OCSP for old versions of openssl which contained
136 early OCSP support, but no stapling (appears to be less than 1.0.0).
138 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
139 server certificate names available under the smtp transport option
140 "tls_verify_cert_hostname" now do not permit multi-component wildcard
143 JH/10 Time-related extraction expansions from certificates now use the main
144 option "timezone" setting for output formatting, and are consistent
145 between OpenSSL and GnuTLS compilations. Bug 1541.
147 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
148 encoded parameter in the incoming message. Bug 1558.
150 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
151 include certificate info, eximon was claiming there were spoolfile
154 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
156 JH/14 Log delivery-related information more consistently, using the sequence
157 "H=<name> [<ip>]" wherever possible.
159 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
160 are problematic for Debian distribution, omit them from the release
163 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
165 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
168 JH/17 Fix a null-indirection in certextract expansions when a nondefault
169 output list separator was used.
174 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
175 checkers that were complaining about end of non-void function with no
178 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
179 This was a regression intruduced in 4.83 by another bugfix.
181 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
183 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
184 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
190 TF/01 Correctly close the server side of TLS when forking for delivery.
192 When a message was received over SMTP with TLS, Exim failed to clear up
193 the incoming connection properly after forking off the child process to
194 deliver the message. In some situations the subsequent outgoing
195 delivery connection happened to have the same fd number as the incoming
196 connection previously had. Exim would try to use TLS and fail, logging
197 a "Bad file descriptor" error.
199 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
200 utilities have not been installed.
202 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
203 temporary space as the ACL may create new global variables.
205 TL/01 LDAP support uses per connection or global context settings, depending
206 upon the detected version of the libraries at build time.
208 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
209 to extract and use the src ip:port in logging and expansions as if it
210 were a direct connection from the outside internet. PPv2 support was
211 updated based on HAProxy spec change in May 2014.
213 JH/02 Add ${listextract {number}{list}{success}{fail}}.
215 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
216 Properly escape header and check for NULL return.
218 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
221 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
223 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
224 characters in header names, implemented as a verify condition.
225 Contributed by Michael Fischer v. Mollard.
227 TL/05 Rename SPF condition results err_perm and err_temp to standardized
228 results permerror and temperror. Previous values are deprecated but
229 still accepted. In a future release, err_perm and err_temp will be
230 completely removed, which will be a backward incompatibility if the
231 ACL tests for either of these two old results. Patch contributed by
232 user bes-internal on the mailing list.
234 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
236 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
237 selectors, in both main and reject logs.
239 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
242 JH/07 Add malware type "sock" for talking to simple daemon.
244 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
246 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
247 routers/transports under cutthrough routing.
249 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
250 numbers. Touch up "bool" conditional to keep the same definition.
252 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
254 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
256 JH/12 Expand items in router/transport headers_add or headers_remove lists
257 individually rather than the list as a whole. Bug 1452.
259 Required for reasonable handling of multiple headers_ options when
260 they may be empty; requires that headers_remove items with embedded
261 colons must have them doubled (or the list-separator changed).
263 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
264 view the policy declared in the DMARC record. Currently, $dmarc_status
265 is a combined value of both the record presence and the result of the
268 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
270 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
271 dnslookup router and the smtp transport (applying to the forward
274 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
275 of ldap servers used for a specific lookup. Patch provided by Heiko
278 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
279 New variable $lookup_dnssec_authenticated for observability.
281 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
282 Patch submitted by Lars Timman.
284 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
286 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
287 Requires trusted mode and valid format message id, aborts otherwise.
288 Patch contributed by Heiko Schlichting.
290 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
291 certextract with support for various fields. Bug 1358.
293 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
294 is requested by default, modifiable by smtp transport option
297 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
298 operate on certificate variables to give certificate fingerprints
299 Also new ${sha256:cert_variable}.
301 JH/23 The PRDR feature is moved from being Experimental into the mainline.
303 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
304 Christian Aistleitner.
306 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
308 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
309 file. Patch from Wolfgang Breyha.
311 JH/25 Expand the coverage of the delivery $host and $host_address to
312 client authenticators run in verify callout. Bug 1476.
314 JH/26 Port service names are now accepted for tls_on_connect_ports, to
315 align with daemon_smtp_ports. Bug 72.
317 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
318 support and error reporting did not work properly.
320 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
321 and is readable. Patch from Andrew Colin Kissa.
323 TL/14 Enhance documentation of ${run expansion and how it parses the
324 commandline after expansion, particularly in the case when an
325 unquoted variable expansion results in an empty value.
327 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
329 PP/02 Fix internal collision of T_APL on systems which support RFC3123
330 by renaming away from it. Addresses GH issue 15, reported by
333 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
335 TL/15 SECURITY: prevent double expansion in math comparison functions
336 (can expand unsanitized data). Not remotely exploitable.
343 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
345 PP/02 Make -n do something, by making it not do something.
346 When combined with -bP, the name of an option is not output.
348 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
351 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
352 $sender_host_name and config options to manage this, and basic check
355 PP/05 DSCP support for outbound connections and control modifier for inbound.
357 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
358 (Only plugin which currently uses this is kerberos4, which nobody should
359 be using, but we should make it available and other future plugins might
360 conceivably use it, even though it would break NAT; stuff *should* be
361 using channel bindings instead).
363 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
364 name; added for Sendmail compatibility; requires admin caller.
365 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
366 just ignore it); requires trusted caller.
367 Also parse but ignore: -Ac -Am -X<logfile>
370 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
372 TL/02 Add +smtp_confirmation as a default logging option.
374 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
375 Patch by Magnus Holmgren from 2007-02-20.
377 TL/04 Bugzilla 1281 - Spec typo.
378 Bugzilla 1283 - Spec typo.
379 Bugzilla 1290 - Spec grammar fixes.
381 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
383 TL/06 Add Experimental DMARC support using libopendmarc libraries.
385 TL/07 Fix an out of order global option causing a segfault. Reported to dev
386 mailing list by by Dmitry Isaikin.
388 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
390 JH/02 Support "G" suffix to numbers in ${if comparisons.
392 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
394 NM/01 Bugzilla 1197 - Spec typo
395 Bugzilla 1196 - Spec examples corrections
397 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
399 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
400 gnutls_enable_pkcs11, but renamed to more accurately indicate its
403 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
404 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
406 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
407 "acl {{name}{arg}...}", and optional args on acl condition
410 JH/05 Permit multiple router/transport headers_add/remove lines.
412 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
414 JH/07 Avoid using a waiting database for a single-message-only transport.
415 Performance patch from Paul Fisher. Bugzilla 1262.
417 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
420 JH/09 Add $headers_added variable, with content from use of ACL modifier
421 add_header (but not yet added to the message). Bugzilla 199.
423 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
424 Pulled from Bugzilla 817 by Wolfgang Breyha.
426 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
428 (nb: this is the same fix as in Exim 4.80.1)
430 JH/11 Add A= logging on delivery lines, and a client_set_id option on
433 JH/12 Add optional authenticated_sender logging to A= and a log_selector
436 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
438 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
439 advertise SMTP AUTH mechanism to us, instead of a generic
440 protocol violation error. Also, make Exim more robust to bad
441 data from the Dovecot auth socket.
443 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
445 When a queue runner is handling a message, Exim first routes the
446 recipient addresses, during which it prunes them based on the retry
447 hints database. After that it attempts to deliver the message to
448 any remaining recipients. It then updates the hints database using
451 So if a recipient address works intermittently, it can get repeatedly
452 deferred at routing time. The retry hints record remains fresh so the
453 address never reaches the final cutoff time.
455 This is a fairly common occurrence when a user is bumping up against
456 their storage quota. Exim had some logic in its local delivery code
457 to deal with this. However it did not apply to per-recipient defers
458 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
460 This change adds a proper retry rule check during routing so that the
461 final cutoff time is checked against the message's age. We only do
462 this check if there is an address retry record and there is not a
463 domain retry record; this implies that previous attempts to handle
464 the address had the retry_use_local_parts option turned on. We use
465 this as an approximation for the destination being like a local
466 delivery, as in LMTP.
468 I suspect this new check makes the old local delivery cutoff check
469 redundant, but I have not verified this so I left the code in place.
471 TF/02 Correct gecos expansion when From: is a prefix of the username.
473 Test 0254 submits a message to Exim with the header
477 When I ran the test suite under the user fanf2, Exim expanded
478 the header to contain my full name, whereas it should have added
479 a Resent-Sender: header. It erroneously treats any prefix of the
480 username as equal to the username.
482 This change corrects that bug.
484 GF/01 DCC debug and logging tidyup
485 Error conditions log to paniclog rather than rejectlog.
486 Debug lines prefixed by "DCC: " to remove any ambiguity.
488 TF/03 Avoid unnecessary rebuilds of lookup-related code.
490 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
491 Bug spotted by Jeremy Harris; was flawed since initial commit.
492 Would have resulted in OCSP responses post-SNI triggering an Exim
493 NULL dereference and crash.
495 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
497 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
498 Bug detection, analysis and fix by Samuel Thibault.
499 Bugzilla 1331, Debian bug #698092.
501 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
503 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
504 Server implementation by Todd Lyons, client by JH.
505 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
506 config variable "prdr_enable" controls whether the server
507 advertises the facility. If the client requests PRDR a new
508 acl_data_smtp_prdr ACL is called once for each recipient, after
509 the body content is received and before the acl_smtp_data ACL.
510 The client is controlled by bolth of: a hosts_try_prdr option
511 on the smtp transport, and the server advertisement.
512 Default client logging of deliveries and rejections involving
513 PRDR are flagged with the string "PRDR".
515 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
516 fclose(). Diagnosis by Todd Lyons.
518 PP/17 Update configure.default to handle IPv6 localhost better.
519 Patch by Alain Williams (plus minor tweaks).
522 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
523 This is now consistent with GnuTLS, and is now documented: the
524 previous undocumented portable approach to treating the option as
525 unset was to force an expansion failure. That still works, and
526 an empty string is now equivalent.
528 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
529 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
530 not performing validation itself.
532 PP/20 Added force_command boolean option to pipe transport.
533 Patch from Nick Koston, of cPanel Inc.
535 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
538 TF/04 Added udpsend ACL modifer and hexquote expansion operator
540 PP/21 Fix eximon continuous updating with timestamped log-files.
541 Broken in a format-string cleanup in 4.80, missed when I repaired the
542 other false fix of the same issue.
543 Report and fix from Heiko Schlichting.
546 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
547 Report from Prashanth Katuri.
549 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
550 It's SecureTransport, so affects any MacOS clients which use the
551 system-integrated TLS libraries, including email clients.
553 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
554 using a MIME ACL for non-SMTP local injection.
555 Report and assistance in diagnosis by Warren Baker.
557 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
559 JH/16 Fix comparisons for 64b. Bugzilla 1385.
561 TL/09 Add expansion variable $authenticated_fail_id to keep track of
562 last id that failed so it may be referenced in subsequent ACL's.
564 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
567 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
568 ldap library initialization, allowing self-signed CA's to be
569 used. Also properly sets require_cert option later in code by
570 using NULL (global ldap config) instead of ldap handle (per
571 session). Bug diagnosis and testing by alxgomz.
573 TL/12 Enhanced documentation in the ratelimit.pl script provided in
574 the src/util/ subdirectory.
576 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
577 renamed to Transport Post Delivery Action by Jeremy Harris, as
580 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
581 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
582 redis_servers = needs to be configured which will be used by the redis
583 lookup. Patch from Warren Baker, of The Packet Hub.
585 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
587 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
588 hostname or reverse DNS when processing a host list. Used suggestions
589 from multiple comments on this bug.
591 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
593 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
594 Missed a few lines, added it to make the runtest require no keyboard
597 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
598 contains upper case chars. Make router use caseful_local_part.
600 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
601 support when GnuTLS has been built with p11-kit.
607 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
609 This, or similar/improved, will also be change PP/11 of 4.82.
615 PP/01 Handle short writes when writing local log-files.
616 In practice, only affects FreeBSD (8 onwards).
617 Bugzilla 1053, with thanks to Dmitry Isaikin.
619 NM/01 Bugzilla 949 - Documentation tweak
621 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
624 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
626 PP/02 Implemented gsasl authenticator.
628 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
630 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
631 `pkg-config foo` for cflags/libs.
633 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
634 with rest of GSASL and with heimdal_gssapi.
636 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
637 `pkg-config foo` for cflags/libs for the TLS implementation.
639 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
640 properties get this fed in as external SSF. A number of robustness
641 and debugging improvements to the cyrus_sasl authenticator.
643 PP/08 cyrus_sasl server now expands the server_realm option.
645 PP/09 Bugzilla 1214 - Log authentication information in reject log.
646 Patch by Jeremy Harris.
648 PP/10 Added dbmjz lookup type.
650 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
652 PP/12 MAIL args handles TAB as well as SP, for better interop with
653 non-compliant senders.
654 Analysis and variant patch by Todd Lyons.
656 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
657 Bug report from Lars Müller <lars@samba.org> (via SUSE),
658 Patch from Dirk Mueller <dmueller@suse.com>
660 PP/13 tls_peerdn now print-escaped for spool files.
661 Observed some $tls_peerdn in wild which contained \n, which resulted
662 in spool file corruption.
664 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
665 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
666 or write after TLS renegotiation, which otherwise led to messages
669 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
670 as a tracking header (ie: a signed header comes before the signature).
671 Patch from Wolfgang Breyha.
673 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
674 comma-sep list; embedded commas doubled.
676 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
678 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
680 Report and patch from Dmitry Banschikov.
682 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
683 Removed SSL_clear() after SSL_new() which led to protocol negotiation
684 failures. We appear to now support TLS1.1+ with Exim.
686 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
687 lets Exim select keys and certificates based upon TLS SNI from client.
688 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
689 before an outbound SMTP session. New log_selector, +tls_sni.
691 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
692 NULL dereference. Report and patch from Alun Jones.
694 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
695 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
696 Not seeing resolver debug output on NetBSD, but suspect this is a
697 resolver implementation change.
699 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
700 Left warnings. Added "eximon gdb" invocation mode.
702 PP/21 Defaulting "accept_8bitmime" to true, not false.
704 PP/22 Added -bw for inetd wait mode support.
706 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
707 locate the relevant includes and libraries. Made this the default.
709 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
710 Bugzilla 1246, report and most of solution from Tomasz Kusy.
712 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
713 This may cause build issues on older platforms.
715 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
716 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
717 gnutls_require_mac & gnutls_require_protocols (no longer supported).
718 Added SNI support via GnuTLS too.
719 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
721 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
723 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
724 Applied second patch from Janne, implementing suggestion to default
725 multiple-strings-in-record handling to match SPF spec.
727 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
729 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
730 read-only, out of scope).
731 Patch from Wolfgang Breyha, report from Stuart Northfield.
733 PP/29 Fix three issues highlighted by clang analyser static analysis.
734 Only crash-plausible issue would require the Cambridge-specific
735 iplookup router and a misconfiguration.
736 Report from Marcin Mirosław.
738 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
740 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
741 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
742 As part of this, removing so much warning spew let me fix some minor
743 real issues in debug logging.
745 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
746 assignment on my part. Fixed.
748 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
749 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
750 Janne Snabb (who went above and beyond: thank you).
752 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
753 string otherwise requires a connection and a bunch more work and it's
754 relatively easy to get wrong. Should also expose TLS library linkage
757 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
758 64-bit ${eval} (JH/03).
760 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
761 GNU libc to support some of the 64-bit stuff, should not lead to
762 conflicts. Defined before os.h is pulled in, so if a given platform
763 needs to override this, it can.
765 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
766 protection layer was required, which is not implemented.
767 Bugzilla 1254, patch from Wolfgang Breyha.
769 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
770 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
771 tls_dhparam take prime identifiers. Also unbreak combination of
772 OpenSSL+DH_params+TLSSNI.
774 PP/39 Disable SSLv2 by default in OpenSSL support.
780 PP/01 Solaris build fix for Oracle's LDAP libraries.
781 Bugzilla 1109, patch from Stephen Usher.
783 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
785 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
788 TF/02 Fix a couple more cases where we did not log the error message
789 when unlink() failed. See also change 4.74-TF/03.
791 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
792 lock up or crash if it happened to be inside a call to libc when it
793 got a SIGUSR1 from exiwhat.
795 The SIGUSR1 handler appends the current process status to the process
796 log which is later printed by exiwhat. It used to use the general
797 purpose logging code to do this, but several functions it calls are
798 not safe for signals.
800 The new output code in the SIGUSR1 handler is specific to the process
801 log, and simple enough that it's easy to inspect for signal safety.
802 Removing some special cases also simplifies the general logging code.
803 Removing the spurious timestamps from the process log simplifies
806 TF/04 Improved ratelimit ACL condition.
808 The /noupdate option has been deprecated in favour of /readonly which
809 has clearer semantics. The /leaky, /strict, and /readonly update modes
810 are mutually exclusive. The update mode is no longer included in the
811 database key; it just determines when the database is updated. (This
812 means that when you upgrde Exim will forget old rate measurements.)
814 Exim now checks that the per_* options are used with an update mode that
815 makes sense for the current ACL. For example, when Exim is processing a
816 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
817 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
818 must specify per_mail/readonly. If you omit the update mode it defaults to
819 /leaky where that makes sense (as before) or /readonly where required.
821 The /noupdate option is now undocumented but still supported for
822 backwards compatibility. It is equivalent to /readonly except that in
823 ACLs where /readonly is required you may specify /leaky/noupdate or
824 /strict/noupdate which are treated the same as /readonly.
826 A useful new feature is the /count= option. This is a generalization
827 of the per_byte option, so that you can measure the throughput of other
828 aggregate values. For example, the per_byte option is now equivalent
829 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
831 The per_rcpt option has been generalized using the /count= mechanism
832 (though it's more complicated than the per_byte equivalence). When it is
833 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
834 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
835 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
836 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
837 Note that using per_rcpt with a non-readonly update mode in more than
838 one ACL will cause the recipients to be double-counted. (The per_mail
839 and per_byte options don't have this problem.)
841 The handling of very low rates has changed slightly. If the computed rate
842 is less than the event's count (usually one) then this event is the first
843 after a long gap. In this case the rate is set to the same as this event's
844 count, so that the first message of a spam run is counted properly.
846 The major new feature is a mechanism for counting the rate of unique
847 events. The new per_addr option counts the number of different
848 recipients that someone has sent messages to in the last time period. It
849 behaves like per_rcpt if all the recipient addresses are different, but
850 duplicate recipient addresses do not increase the measured rate. Like
851 the /count= option this is a general mechanism, so the per_addr option
852 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
853 example, measure the rate that a client uses different sender addresses
854 with the options per_mail/unique=$sender_address. There are further
855 details in the main documentation.
857 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
859 TF/06 Removed a few PCRE remnants.
861 TF/07 Automatically extract Exim's version number from tags in the git
862 repository when doing development or release builds.
864 PP/02 Raise smtp_cmd_buffer_size to 16kB.
865 Bugzilla 879. Patch from Paul Fisher.
867 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
868 Heavily based on revision 40f9a89a from Simon Arlott's tree.
871 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
873 PP/05 Variable $av_failed, true if the AV scanner deferred.
874 Bugzilla 1078. Patch from John Horne.
876 PP/06 Stop make process more reliably on build failure.
877 Bugzilla 1087. Patch from Heiko Schlittermann.
879 PP/07 Make maildir_use_size_file an _expandable_ boolean.
880 Bugzilla 1089. Patch from Heiko Schlittermann.
882 PP/08 Handle ${run} returning more data than OS pipe buffer size.
883 Bugzilla 1131. Patch from Holger Weiß.
885 PP/09 Handle IPv6 addresses with SPF.
886 Bugzilla 860. Patch from Wolfgang Breyha.
888 PP/10 GnuTLS: support TLS 1.2 & 1.1.
890 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
893 PP/11 match_* no longer expand right-hand-side by default.
894 New compile-time build option, EXPAND_LISTMATCH_RHS.
895 New expansion conditions, "inlist", "inlisti".
897 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
899 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
901 PP/14 fix log_write() format string regression from TF/03.
902 Bugzilla 1152. Patch from Dmitry Isaikin.
908 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
910 PP/02 Harmonised TLS library version reporting; only show if debugging.
911 Layout now matches that introduced for other libraries in 4.74 PP/03.
913 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
915 PP/04 New "dns_use_edns0" global option.
917 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
920 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
921 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
923 TK/01 Updated PolarSSL code to 0.14.2.
924 Bugzilla 1097. Patch from Andreas Metzler.
926 PP/07 Catch divide-by-zero in ${eval:...}.
929 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
932 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
933 format-string attack -- SECURITY: remote arbitrary code execution.
935 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
936 time unintentionally subject to list matching rules, letting the header
937 cause arbitrary Exim lookups (of items which can occur in lists, *not*
938 arbitrary string expansion). This allowed for information disclosure.
940 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
941 INT_MIN/-1 -- value coerced to INT_MAX.
947 NM/01 Workround for PCRE version dependency in version reporting
950 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
951 This fixes portability to compilers other than gcc, notably
952 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
954 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
955 makefiles for portability to HP-UX and POSIX correctness.
957 PP/01 Permit LOOKUP_foo enabling on the make command-line.
958 Also via indented variable definition in the Makefile.
959 (Debugging by Oliver Heesakkers).
961 PP/02 Restore caching of spamd results with expanded spamd_address.
962 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
964 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
965 Improves build reliability. Fix from: Frank Elsner
967 NM/02 Fix wide character breakage in the rfc2047 coding
968 Fixes bug 1064. Patch from Andrey N. Oktyabrski
970 NM/03 Allow underscore in dnslist lookups
971 Fixes bug 1026. Patch from Graeme Fowler
973 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
974 Code patches from Adam Ciarcinski of NetBSD.
976 NM/04 Fixed exiqgrep to cope with mailq missing size issue
979 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
980 is logged, to avoid truncation. Patch from John Horne.
982 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
983 Patch from Jakob Hirsch.
985 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
986 SQL string expansion failure details.
987 Patch from Andrey Oktyabrski.
989 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
990 Patch from Simon Arlott.
992 PP/09 New lookups functionality failed to compile on old gcc which rejects
993 extern declarations in function scope.
994 Patch from Oliver Fleischmann
996 PP/10 Use sig_atomic_t for flags set from signal handlers.
997 Check getgroups() return and improve debugging.
998 Fixed developed for diagnosis in bug 927 (which turned out to be
1001 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1002 Patch from Mark Zealey.
1004 PP/12 Bugzilla 1056: Improved spamd server selection.
1005 Patch from Mark Zealey.
1007 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1008 Based on patch from Heiko Schlittermann.
1010 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1011 Patch from Uwe Doering, sign-off by Michael Haardt.
1013 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1014 variable declaration deep within a block. Bug and patch from
1017 PP/15 lookups-Makefile IRIX compatibilty coercion.
1019 PP/16 Make DISABLE_DKIM build knob functional.
1021 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1022 Patch by Simon Arlott
1024 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1025 variable argument macros. Our copy now differs from upstream.
1031 TF/01 Failure to get a lock on a hints database can have serious
1032 consequences so log it to the panic log.
1034 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1035 controlled using the smtp_confirmation log selector.
1037 TF/03 Include the error message when we fail to unlink a spool file.
1039 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1040 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1041 for maintaining out-of-tree patches for some time.
1043 PP/01 Bugzilla 139: Documentation and portability issues.
1044 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1045 Handle per-OS dynamic-module compilation flags.
1047 PP/02 Let /dev/null have normal permissions.
1048 The 4.73 fixes were a little too stringent and complained about the
1049 permissions on /dev/null. Exempt it from some checks.
1050 Reported by Andreas M. Kirchwitz.
1052 PP/03 Report version information for many libraries, including
1053 Exim version information for dynamically loaded libraries. Created
1054 version.h, now support a version extension string for distributors
1055 who patch heavily. Dynamic module ABI change.
1057 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1058 privilege escalation vulnerability whereby the Exim run-time user
1059 can cause root to append content of the attacker's choosing to
1062 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1065 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1066 If dropping privileges for untrusted macros, we disabled normal logging
1067 on the basis that it would fail; for the Exim run-time user, this is not
1068 the case, and it resulted in successful deliveries going unlogged.
1069 Fixed. Reported by Andreas Metzler.
1075 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1076 only prepend for the Resent-* case. Fixes regression introduced in
1077 Exim 4.70 by NM/22 for Bugzilla 607.
1079 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1080 increasing numbers of administrators be bitten by this.
1082 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1084 PP/03 Bugzilla 994: added openssl_options main configuration option.
1086 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1088 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1090 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1092 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1093 without a peer certificate, leading to a segfault because of an
1094 assumption that peers always have certificates. Be a little more
1095 paranoid. Problem reported by Martin Tscholak.
1097 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1098 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1099 NB: ClamAV planning to remove STREAM in "middle of 2010".
1100 CL also introduces -bmalware, various -d+acl logging additions and
1101 more caution in buffer sizes.
1103 PP/09 Implemented reverse_ip expansion operator.
1105 PP/10 Bugzilla 937: provide a "debug" ACL control.
1107 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1109 PP/12 Bugzilla 973: Implement --version.
1111 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1113 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1115 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1117 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1118 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1119 ignore trailing whitespace.
1121 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1123 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1126 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1127 Notification from Dr Andrew Aitchison.
1129 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1130 ExtendedDetectionInfo response format.
1131 Notification from John Horne.
1133 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1136 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1137 XSL and documented dependency on system catalogs, with examples of how
1140 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1143 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1144 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1145 configuration file which is writeable by the Exim user or group.
1147 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1148 of configuration files to cover files specified with the -C option if
1149 they are going to be used with root privileges, not just the default
1152 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1153 option (effectively making it always true).
1155 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1156 files to be used while preserving root privileges.
1158 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1159 that rogue child processes cannot use them.
1161 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1162 run-time user, instead of root.
1164 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1165 Exim run-time user without dropping privileges.
1167 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1168 result string, instead of calling string_vformat() twice with the same
1171 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1172 for other users. Others should always drop root privileges if they use
1173 -C on the command line, even for a whitelisted configure file.
1175 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1177 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1183 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1184 $data_path, and $header_path variables; fixed documentation bugs and
1187 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1188 exipick to access non-standard spools, including the "frozen" queue
1191 NM/01 Bugzilla 965: Support mysql stored procedures.
1192 Patch from Alain Williams
1194 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1196 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1197 Patch from Andreas Metzler
1199 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1200 Patch from Kirill Miazine
1202 NM/05 Bugzilla 671: Added umask to procmail example.
1204 JJ/03 installed exipick 20100323.0, fixing doc bug
1206 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1207 directory. Notification and patch from Dan Rosenberg.
1209 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1211 TK/02 Improve log output when DKIM signing operation fails.
1213 MH/01 Treat the transport option dkim_domain as a colon separated
1214 list, not as a single string, and sign the message with each element,
1215 omitting multiple occurences of the same signer.
1217 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1218 Bugzilla 985, 986. Patch by Simon Arlott
1220 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1221 Patch by Simon Arlott
1223 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1224 MBX locking. Notification from Dan Rosenberg.
1230 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1232 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1234 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1236 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1238 NM/04 Bugzilla 915: Flag broken perl installation during build.
1244 TK/01 Added patch by Johannes Berg that expands the main option
1245 "spamd_address" if it starts with a dollar sign.
1247 TK/02 Write list of recipients to X-Envelope-Sender header when building
1248 the mbox-format spool file for content scanning (suggested by Jakob
1251 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1252 (http://www.dcc-servers.net/) support via dccifd. Activated by
1253 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1255 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1256 by Mark Daniel Reidel <mr@df.eu>.
1258 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1259 When building exim an external PCRE library is now needed -
1260 PCRE is a system library on the majority of modern systems.
1261 See entry on PCRE_LIBS in EDITME file.
1263 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1264 conversation. Added nologin parameter to request.
1265 Patch contributed by Kirill Miazine.
1267 TF/01 Do not log submission mode rewrites if they do not change the address.
1269 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1271 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1272 log files in place. Contributed by Roberto Lima.
1274 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1276 TF/03 Bugzilla 615: When checking the local_parts router precondition
1277 after a local_part_suffix or local_part_prefix option, Exim now
1278 does not use the address's named list lookup cache, since this
1279 contains cached lookups for the whole local part.
1281 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1282 Robert Millan. Documentation is in experimental-spec.txt.
1284 TF/04 Bugzilla 668: Fix parallel build (make -j).
1286 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1288 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1289 Patch provided by Jan Srzednicki.
1291 TF/05 Leading white space used to be stripped from $spam_report which
1292 wrecked the formatting. Now it is preserved.
1294 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1295 that they are available at delivery time.
1297 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1299 TF/08 TLS error reporting now respects the incoming_interface and
1300 incoming_port log selectors.
1302 TF/09 Produce a more useful error message if an SMTP transport's hosts
1303 setting expands to an empty string.
1305 NM/06 Bugzilla 744: EXPN did not work under TLS.
1306 Patch provided by Phil Pennock.
1308 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1309 Patch provided by Richard Godbee.
1311 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1312 acl_smtp_notquit, added index entry.
1314 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1315 Patch provided by Eugene Bujak.
1317 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1318 accept(). Patch provided by Maxim Dounin.
1320 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1321 Patch provided by Phil Pennock.
1323 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1325 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1326 Patch provided by Brad "anomie" Jorsch.
1328 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1329 Patch provided by Dean Brooks.
1331 TK/05 Add native DKIM support (does not depend on external libraries).
1333 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1334 Patch provided by Graeme Fowler.
1336 NM/16 Bugzilla 851: Documentation example syntax fix.
1338 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1340 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1343 NM/19 Bugzilla 745: TLS version reporting.
1344 Patch provided by Phil Pennock.
1346 NM/20 Bugzilla 167: bool: condition support.
1347 Patch provided by Phil Pennock.
1349 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1350 clients. Patch provided by Phil Pennock.
1352 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1353 Patch provided by Brad "anomie" Jorsch.
1355 NM/23 Bugzilla 687: Fix misparses in eximstats.
1356 Patch provided by Heiko Schlittermann.
1358 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1359 Patch provided by Heiko Schlittermann.
1361 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1362 plus update to original patch.
1364 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1366 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1367 Patch provided by David Brownlee.
1369 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1371 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1373 NM/30 Bugzilla 888: TLS documentation bugfixes.
1375 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1377 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1378 Unlike the original bugzilla I have changed all shell scripts in src tree.
1380 NM/33 Bugzilla 898: Transport filter timeout fix.
1381 Patch by Todd Rinaldo.
1383 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1384 Patch by Serge Demonchaux.
1386 NM/35 Bugzilla 39: Base64 decode bug fixes.
1387 Patch by Jakob Hirsch.
1389 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1391 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1393 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1395 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1401 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1402 ALT-N's libdkim that I have put here:
1403 http://duncanthrax.net/exim-experimental/
1405 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1406 were called 'true' and it seems that C99 defines that as a reserved
1407 keyword to be used with 'bool' variable types. That means you could
1408 not include C99-style headers which use bools without triggering
1409 build errors in sieve.c.
1411 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1412 as mailq or other aliases. Changed the --help handling significantly
1413 to do whats expected. exim_usage() emits usage/help information.
1415 SC/01 Added the -bylocaldomain option to eximstats.
1417 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1419 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1421 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1427 PH/01 Another patch from the Sieve maintainer.
1429 PH/02 When an IPv6 address is converted to a string for single-key lookup
1430 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1431 dots are used instead of colons so that keys in lsearch files need not
1432 contain colons. This was done some time before quoting was made available
1433 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1434 (notated using the quote facility) so as to distinguish them from IPv4
1435 keys. This meant that lookups for IP addresses in host lists did not work
1436 for iplsearch lookups.
1438 This has been fixed by arranging for IPv6 addresses to be expressed with
1439 colons if the lookup type is iplsearch. This is not incompatible, because
1440 previously such lookups could never work.
1442 The situation is now rather anomolous, since one *can* have colons in
1443 ordinary lsearch keys. However, making the change in all cases is
1444 incompatible and would probably break a number of configurations.
1446 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1449 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1450 conversion specification without a maximum field width, thereby enabling
1451 a rogue spamd server to cause a buffer overflow. While nobody in their
1452 right mind would setup Exim to query an untrusted spamd server, an
1453 attacker that gains access to a server running spamd could potentially
1454 exploit this vulnerability to run arbitrary code as the Exim user.
1456 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1457 $primary_hostname instead of what libspf2 thinks the hosts name is.
1459 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1460 a directory entry by the name of the lookup key. Previously, if a
1461 symlink pointed to a non-existing file or a file in a directory that
1462 Exim lacked permissions to read, a lookup for a key matching that
1463 symlink would fail. Now it is enough that a matching directory entry
1464 exists, symlink or not. (Bugzilla 503.)
1466 PH/03 The body_linecount and body_zerocount variables are now exported in the
1469 PH/04 Added the $dnslist_matched variable.
1471 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1472 This means they are set thereafter only if the connection becomes
1475 PH/06 Added the client_condition to authenticators so that some can be skipped
1476 by clients under certain conditions.
1478 PH/07 The error message for a badly-placed control=no_multiline_responses left
1479 "_responses" off the end of the name.
1481 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1483 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1484 (without spaces) instead of just copying the configuration text.
1486 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1488 PH/11 Added $max_received_linelength.
1490 PH/12 Added +ignore_defer and +include_defer to host lists.
1492 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1493 way in which PCRE > 7.0 is built.
1495 PH/14 Implemented queue_only_load_latch.
1497 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1498 MAIL command. The effect was to mangle the value on 64-bit systems.
1500 PH/16 Another patch from the Sieve maintainer.
1502 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1504 PH/18 If a system quota error occurred while trying to create the file for
1505 a maildir delivery, the message "Mailbox is full" was not appended to the
1506 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1507 only to a quota excession during the actual writing of the file.
1509 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1510 characters?) which causes problems in log lines. The DN values are now
1511 passed through string_printing() before being added to log lines.
1513 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1514 and InterBase are left for another time.)
1516 PH/21 Added message_body_newlines option.
1518 PH/22 Guard against possible overflow in moan_check_errorcopy().
1520 PH/23 POSIX allows open() to be a macro; guard against that.
1522 PH/24 If the recipient of an error message contained an @ in the local part
1523 (suitably quoted, of course), incorrect values were put in $domain and
1524 $local_part during the evaluation of errors_copy.
1530 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1531 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1534 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1535 issue a MAIL command.
1537 PH/02 In an ACL statement such as
1539 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1541 if a client was not listed at all, or was listed with a value other than
1542 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1543 the condition was not true (as it should be), so access was not denied.
1544 The bug was that the ! inversion was incorrectly passed on to the second
1545 item. This has been fixed.
1547 PH/03 Added additional dnslists conditions == and =& which are different from
1548 = and & when the dns lookup returns more than one IP address.
1550 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1551 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1553 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1554 FSYNC, which compiles an option called disable_fsync that allows for
1555 bypassing fsync(). The documentation is heavily laced with warnings.
1557 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1559 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1560 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1561 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1562 including adding "make clean"; (3) Added -fPIC when compiling the test
1563 dynamically loaded module, to get rid of a warning.
1565 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1566 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1567 = 0s. The bug is otherwise harmless.
1569 PH/07 There was a bug in the dovecot authenticator such that the value of
1570 $auth1 could be overwritten, and so not correctly preserved, after a
1571 successful authentication. This usually meant that the value preserved by
1572 the server_setid option was incorrect.
1574 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1576 PH/09 Installed PCRE release 7.0.
1578 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1579 run for batched SMTP input. It is now run at the start of every message
1580 in the batch. While fixing this I discovered that the process information
1581 (output by running exiwhat) was not always getting set for -bs and -bS
1582 input. This is fixed, and it now also says "batched" for BSMTP.
1584 PH/11 Added control=no_pipelining.
1586 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1587 patch, slightly modified), and move the expansion of helo_data till after
1588 the connection is made in the smtp transport (so it can use these
1591 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1593 PH/14 Added log_selector = +pid.
1595 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1597 PH/16 Add ${if forany and ${if forall.
1599 PH/17 Added dsn_from option to vary the From: line in DSNs.
1601 PH/18 Flush SMTP output before performing a callout, unless control =
1602 no_callout_flush is set.
1604 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1605 was true (the default) a successful delivery failed to delete the retry
1606 item, thus causing premature timeout of the address. The bug is now
1609 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1611 PH/21 Long custom messages for fakedefer and fakereject are now split up
1612 into multiline reponses in the same way that messages for "deny" and
1613 other ACL rejections are.
1615 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1616 with slight modification.
1618 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1619 draft, changing the syntax and factoring some duplicate code".
1621 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1622 for deliveries of the second and subsequent messages over the same SMTP
1625 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1626 ${reduce, with only minor "tidies".
1628 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1630 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1631 expansion side effects.
1633 PH/27 When a message times out after an over-quota error from an Exim-imposed
1634 quota, the bounce message says "mailbox is full". This message was not
1635 being given when it was a system quota that was exceeded. It now should
1638 MH/03 Made $recipients available in local_scan(). local_scan() already has
1639 better access to the recipient list through recipients_list[], but
1640 $recipients can be useful in postmaster-provided expansion strings.
1642 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1643 in the case of a MAIL command with additional options following the
1644 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1645 were accidentally chopped off.
1647 PH/29 SMTP synchronization checks are implemented when a command is read -
1648 there is a check that no more input is waiting when there shouldn't be
1649 any. However, for some commands, a delay in an ACL can mean that it is
1650 some time before the response is written. In this time, more input might
1651 arrive, invalidly. So now there are extra checks after an ACL has run for
1652 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1653 pipelining has not been advertised.
1655 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1657 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1658 returned in the SMTP response when smtp_return_error_details was false.
1659 This has been fixed.
1661 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1662 instead of the C I/O that was originally supplied, because problems were
1663 reported on Solaris.
1665 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1666 Exim which did not show up earlier: it was assuming that a call to
1667 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1668 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1669 was a macro that became an assignment, so it seemed to work. This has
1670 changed to a proper function call with a void return, hence the compile
1671 error. Exim's code has been fixed.
1673 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1676 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1678 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1680 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1681 because it turns out to be too big or there is a timeout) there is no
1682 "Completed" line in the log. When some messages of this type were
1683 selected by exigrep, they were listed as "not completed". Others were
1684 picked up by some special patterns. I have improved the selection
1685 criteria to be more general.
1687 PH/38 The host_find_failed option in the manualroute router can now be set
1688 to "ignore", to completely ignore a host whose IP address cannot be
1689 found. If all hosts are ignored, the behaviour is controlled by the new
1690 host_all_ignored option.
1692 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1693 homing or because of multiple MX records with /mx) generated more than
1694 one IP address, and the following item turned out to be the local host,
1695 all the secondary addresses of the first item were incorrectly removed
1696 from the list, along with the local host and any following hosts (which
1697 is what is supposed to happen).
1699 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1700 whoever called Exim into the -H file. In the case of the daemon it was
1701 behaving confusingly. When first started, it used values for whoever
1702 started the daemon, but after a SIGHUP it used the Exim user (because it
1703 calls itself on a restart). I have changed the code so that it now always
1706 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1707 message are rejected with the same error (e.g. no authentication or bad
1708 sender address), and a DATA command is nevertheless sent (as can happen
1709 with PIPELINING or a stupid MUA), the error message that was given to the
1710 RCPT commands is included in the rejection of the DATA command. This is
1711 intended to be helpful for MUAs that show only the final error to their
1714 PH/42 Another patch from the Sieve maintainer.
1716 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1717 Eximstats - Fixed some broken HTML links and added missing column headers
1719 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1720 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1722 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1724 PH/43 Yet another patch from the Sieve maintainer.
1726 PH/44 I found a way to check for a TCP/IP connection going away before sending
1727 the response to the final '.' that terminates a message, but only in the
1728 case where the client has not sent further data following the '.'
1729 (unfortunately, this is allowed). However, in many cases there won't be
1730 any further data because there won't be any more messages to send. A call
1731 to select() can be used: if it shows that the input is "ready", there is
1732 either input waiting, or the socket has been closed. An attempt to read
1733 the next input character can distinguish the two cases. Previously, Exim
1734 would have sent an OK response which the client would never have see.
1735 This could lead to message repetition. This fix should cure that, at
1736 least in a lot of common cases.
1738 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1739 advertised in response to EHLO.
1745 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1746 fixed by 4.65/MH/01 (is this a record?) are fixed:
1748 (i) An empty string was always treated as zero by the numeric comparison
1749 operators. This behaviour has been restored.
1751 (ii) It is documented that the numeric comparison operators always treat
1752 their arguments as decimal numbers. This was broken in that numbers
1753 starting with 0 were being interpreted as octal.
1755 While fixing these problems I realized that there was another issue that
1756 hadn't been noticed. Values of message_size_limit (both the global option
1757 and the transport option) were treated as octal if they started with 0.
1758 The documentation was vague. These values are now always treated as
1759 decimal, and I will make that clear in the documentation.
1765 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1766 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1769 MH/01 Don't check that the operands of numeric comparison operators are
1770 integers when their expansion is in "skipping" mode (fixes bug
1771 introduced by 4.64-PH/07).
1773 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1774 child addresses, Exim now panics and dies. Previously, because the count
1775 is held in a short int, deliveries were likely to be lost. As such a
1776 large number of recipients for a single message is ridiculous
1777 (performance will be very, very poor), I have chosen to impose a limit
1778 rather than extend the field.
1784 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1785 leftover -K file (the existence of which was triggered by #402).
1786 While we were at it, introduced process PID as part of the -K
1787 filename. This should rule out race conditions when creating
1790 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1791 processing considerably. Previous code took too long for large mails,
1792 triggering a timeout which in turn triggers #401.
1794 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1795 in the DK code in transports.c. sendfile() is not really portable,
1796 hence the _LINUX specificness.
1798 TF/01 In the add_headers option to the mail command in an Exim filter,
1799 there was a bug that Exim would claim a syntax error in any
1800 header after the first one which had an odd number of characters
1803 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1804 callout verification, Exim cached a "reject" for the entire domain. This
1805 is correct for most verifications, but it is not correct for a recipient
1806 verification with use_sender or use_postmaster set, because in that case
1807 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1808 case of MAIL FROM:<> rejection from other early rejections (e.g.
1809 rejection of HELO). When verifying a recipient using a non-null MAIL
1810 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1811 Whatever the result of the callout, the value of the domain cache is
1812 left unchanged (for any other kind of callout, getting as far as trying
1813 RCPT means that the domain itself is ok).
1815 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1818 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1819 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1820 session, the daemon ignores ECONNECT errors and logs others; it now
1821 ignores EPIPE as well.
1823 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1824 (quoted-printable decoding).
1826 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1827 later the small subsequent patch to fix an introduced bug.
1829 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1831 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1833 PH/08 An error is now given if message_size_limit is specified negative.
1835 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1836 to be given (somewhat) arbitrary names.
1838 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1841 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1842 miscellaneous code fixes
1844 PH/10 Added the log_reject_target ACL modifier to specify where to log
1847 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1848 hostname. This is wrong, because it relates to the incoming message (and
1849 probably the interface on which it is arriving) and not to the outgoing
1850 callout (which could be using a different interface). This has been
1851 changed to use the value of the helo_data option from the smtp transport
1852 instead - this is what is used when a message is actually being sent. If
1853 there is no remote transport (possible with a router that sets up host
1854 addresses), $smtp_active_hostname is used.
1856 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1857 tweaks were necessary in order to get it to work (see also 21 below):
1858 (a) The code assumed that strncpy() returns a negative number on buffer
1859 overflow, which isn't the case. Replaced with Exim's string_format()
1861 (b) There were several signed/unsigned issues. I just did the minimum
1862 hacking in of casts. There is scope for a larger refactoring.
1863 (c) The code used strcasecmp() which is not a standard C function.
1864 Replaced with Exim's strcmpic() function.
1865 (d) The code set only $1; it now sets $auth1 as well.
1866 (e) A simple test gave the error "authentication client didn't specify
1867 service in request". It would seem that Dovecot has changed its
1868 interface. Fortunately there's a specification; I followed it and
1869 changed what the client sends and it appears to be working now.
1871 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1874 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1875 address A is aliased to B and C, where B exists and C does not. Without
1876 -v the output is "A verified" because verification stops after a
1877 successful redirection if more than one address is generated. However,
1878 with -v the child addresses are also verified. Exim was outputting "A
1879 failed to verify" and then showing the successful verification for C,
1880 with its parentage. It now outputs "B failed to verify", showing B's
1881 parentage before showing the successful verification of C.
1883 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1884 look up a TXT record in a specific list after matching in a combined
1887 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1888 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1889 they consult the DNS. I had assumed they would set it the way they
1890 wanted; and indeed my experiments on Linux seem to show that in some
1891 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1892 To be on the safe side, however, I have now made the interface to
1893 host_find_byname() similar to host_find_bydns(), with an argument
1894 containing the DNS resolver options. The host_find_byname() function now
1895 sets these options at its start, just as host_find_bydns() does. The smtp
1896 transport options dns_qualify_single and dns_search_parents are passed to
1897 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1898 of host_find_byname() use the default settings of RES_DEFNAMES
1899 (qualify_single) but not RES_DNSRCH (search_parents).
1901 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1902 spool_read_header() do less string testing, by means of a preliminary
1903 switch on the second character of optional "-foo" lines. (This is
1904 overdue, caused by the large number of possibilities that now exist.
1905 Originally there were few.) While I was there, I also converted the
1906 str(n)cmp tests so they don't re-test the leading "-" and the first
1907 character, in the hope this might squeeze out yet more improvement.
1909 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1910 flag allowing group syntax was set by the header_syntax check but not
1911 turned off, possible causing trouble later; (2) The flag was not being
1912 set at all for the header_verify test, causing "group"-style headers to
1913 be rejected. I have now set it in this case, and also caused header_
1914 verify to ignore an empty address taken from a group. While doing this, I
1915 came across some other cases where the code for allowing group syntax
1916 while scanning a header line wasn't quite right (mostly, not resetting
1917 the flag correctly in the right place). These bugs could have caused
1918 trouble for malformed header lines. I hope it is now all correct.
1920 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1921 with the "reply" argument non-NULL. The code, however (which originally
1922 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1923 but it didn't always do it. This confused somebody who was copying the
1924 code for some other use. I have removed all the tests.
1926 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1927 feature that was used to support insecure browsers during the U.S. crypto
1928 embargo. It requires special client support, and Exim is probably the
1929 only MTA that supported it -- and would never use it because real RSA is
1930 always available. This code has been removed, because it had the bad
1931 effect of slowing Exim down by computing (never used) parameters for the
1932 RSA_EXPORT functionality.
1934 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1935 authenticator to fail if there's a tab character in the incoming data
1936 (there should never be unless someone is messing about, as it's supposed
1937 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1938 if the connection is using TLS or if the remote IP is the same as the
1939 local IP, and the "valid-client-cert option" if a client certificate has
1942 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1943 authenticators. This can be used for authorization after authentication
1944 succeeds. (In the case of plaintext, it servers for both authentication
1947 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1948 if any retry times were supplied.
1950 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1951 connection, where there is no client IP address to check. In this
1952 situation, the verify now always succeeds.
1954 PH/25 Applied John Jetmore's -Mset patch.
1956 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1958 PH/27 In a string expansion for a processed (not raw) header when multiple
1959 headers of the same name were present, leading whitespace was being
1960 removed from all of them, but trailing whitespace was being removed only
1961 from the last one. Now trailing whitespace is removed from each header
1962 before concatenation. Completely empty headers in a concatenation (as
1963 before) are ignored.
1965 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1966 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1968 PH/29 [Removed. This was a change that I later backed out, and forgot to
1969 correct the ChangeLog entry (that I had efficiently created) before
1970 committing the later change.]
1972 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1973 address errors (4xx response to RCPT) over the same connection as other
1974 messages routed to the same hosts. Such deliveries are always "forced",
1975 so retry times are not inspected. This resulted in far too many retries
1976 for the affected addresses. The effect occurred only when there were more
1977 hosts than the hosts_max_try setting in the smtp transport when it had
1978 the 4xx errors. Those hosts that it had tried were not added to the list
1979 of hosts for which the message was waiting, so if all were tried, there
1980 was no problem. Two fixes have been applied:
1982 (i) If there are any address or message errors in an SMTP delivery, none
1983 of the hosts (tried or untried) are now added to the list of hosts
1984 for which the message is waiting, so the message should not be a
1985 candidate for sending over the same connection that was used for a
1986 successful delivery of some other message. This seems entirely
1987 reasonable: after all the message is NOT "waiting for some host".
1988 This is so "obvious" that I'm not sure why it wasn't done
1989 previously. Hope I haven't missed anything, but it can't do any
1990 harm, as the worst effect is to miss an optimization.
1992 (ii) If, despite (i), such a delivery is accidentally attempted, the
1993 routing retry time is respected, so at least it doesn't keep
1994 hammering the server.
1996 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1997 in ${readsocket because some servers need this prod.
1999 PH/32 Added some extra debug output when updating a wait-xxx database.
2001 PH/33 The hint "could be header name not terminated by colon", which has been
2002 given for certain expansion errors for a long time, was not being given
2003 for the ${if def:h_colon_omitted{... case.
2005 PH/34 The spec says: "With one important exception, whenever a domain list is
2006 being scanned, $domain contains the subject domain." There was at least
2007 one case where this was not true.
2009 PH/35 The error "getsockname() failed: connection reset by peer" was being
2010 written to the panic log as well as the main log, but it isn't really
2011 panic-worthy as it just means the connection died rather early on. I have
2012 removed the panic log writing for the ECONNRESET error when getsockname()
2015 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2016 runs only) independently of the message's sender address. This meant
2017 that, if the 4xx error was in fact related to the sender, a different
2018 message to the same recipient with a different sender could confuse
2019 things. In particualar, this can happen when sending to a greylisting
2020 server, but other circumstances could also provoke similar problems.
2021 I have changed the default so that the retry time for these errors is now
2022 based a combination of the sender and recipient addresses. This change
2023 can be overridden by setting address_retry_include_sender=false in the
2026 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2027 remote server are returned as part of bounce messages. This was not
2028 happening for LMTP over a pipe (the lmtp transport), but now it is the
2029 same for both kinds of LMTP.
2031 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2032 in header lines that were in fact CNAMEs. This is no longer the case.
2034 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2035 and queue runs started by the daemon processed all messages. This has
2036 been fixed so that -R and -S can now usefully be given with -q<time>.
2038 PH/40 Import PCRE release 6.7 (fixes some bugs).
2040 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2042 PH/42 Give an error if -q is specified more than once.
2044 PH/43 Renamed the variables $interface_address and $interface_port as
2045 $received_ip_address and $received_port, to make it clear that these
2046 values apply to message reception, and not to the outgoing interface when
2047 a message is delivered. (The old names remain recognized, of course.)
2049 PH/44 There was no timeout on the connect() call when using a Unix domain
2050 socket in the ${readsocket expansion. There now is.
2052 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2053 be meaningful with "accept".
2055 SC/01 Eximstats V1.43
2056 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2058 SC/02 Eximstats V1.44
2059 Use a glob alias rather than an array ref in the generated
2060 parser. This improves both readability and performance.
2062 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2063 Collect SpamAssassin and rejection statistics.
2064 Don't display local sender or destination tables unless
2065 there is data to show.
2066 Added average volumes into the top table text output.
2068 SC/04 Eximstats V1.46
2069 Collect data on the number of addresses (recipients)
2070 as well as the number of messages.
2072 SC/05 Eximstats V1.47
2073 Added 'Message too big' to the list of mail rejection
2074 reasons (thanks to Marco Gaiarin).
2076 SC/06 Eximstats V1.48
2077 Mainlog lines which have GMT offsets and are too short to
2078 have a flag are now skipped.
2080 SC/07 Eximstats V1.49 (Alain Williams)
2081 Added the -emptyok flag.
2083 SC/08 Eximstats V1.50
2084 Fixes for obtaining the IP address from reject messages.
2086 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2087 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2088 whitesspace changes from 4.64-PH/27
2090 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2093 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2094 are found, allow negative numbers in numeric criteria)
2096 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2098 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2099 to match changes made in 4.64-PH/43
2101 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2103 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2104 "domain" is actually an IP address, and force a failure. This locks out
2105 those revolvers/nameservers that support "A-for-A" lookups, in
2106 contravention of the specifications.
2108 PH/48 When a host name was looked up from an IP address, and the subsequent
2109 forward lookup of the name timed out, the host name was left in
2110 $sender_host_name, contrary to the specification.
2112 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2113 restricted to single-key lookups, Exim was not diagnosing an error if
2114 * or *@ was used with a query-style lookup.
2116 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2118 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2119 long ago, but noone interested enough thought of it. Let's just say that
2120 the "1.1" means that there are some new functions that weren't there at
2121 some point in the past.
2123 PH/51 Error processing for expansion failure of helo_data from an smtp
2124 transport during callout processing was broken.
2126 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2127 tested/used via the -bh/-bhc/-bs options.
2129 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2130 bug, fixed in subsequent PCRE releases).
2132 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2133 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2135 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2141 SC/01 Use a glob alias rather than an array ref in eximstats generated
2142 parser. This improves both readability and performance.
2144 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2145 Don't display local sender or destination tables in eximstats unless
2146 there is data to show.
2147 Added average volumes into the eximstats top table text output.
2149 SC/03 Collect data on the number of addresses (recipients) as well
2150 as the number of messages in eximstats.
2152 TF/01 Correct an error in the documentation for the redirect router. Exim
2153 does (usually) call initgroups() when daemonizing.
2155 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2156 with consistent privilege compared to when running as a daemon.
2158 TF/03 Note in the spec that $authenticated_id is not set for local
2159 submissions from trusted users.
2161 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2162 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2164 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2165 by adding some example configuration directives to the default
2166 configuration file. A little bit of work is required to uncomment the
2167 directives and define how usernames and passwords are checked, but
2168 there is now a framework to start from.
2170 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2171 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2172 without this. I don't know how relevant this is to other LDAP libraries.
2174 PH/02 Add the verb name to the "unknown ACL verb" error.
2176 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2178 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2180 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2181 directory not expanded when it should be if an expanded home directory
2182 was set for the address (which is overridden by the transport).
2184 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2187 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2188 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2189 because it is too late at that time, and has no effect.
2191 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2192 security issue with \' (bugzilla #107). I could not use the
2193 PQescapeStringConn() function, because it needs a PGconn value as one of
2196 PH/08 When testing addresses using -bt, indicate those final addresses that
2197 are duplicates that would not cause an additional delivery. At least one
2198 person was confused, thinking that -bt output corresponded to deliveries.
2199 (Suppressing duplicates isn't a good idea as you lose the information
2200 about possibly different redirections that led to the duplicates.)
2202 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2203 systems where poll() doesn't work, in particular OS X.
2205 PH/10 Added more information to debugging output for retry time not reached.
2207 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2208 operations in malware.c.
2210 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2213 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2214 syslog_duplication set false, log lines that would normally be written
2215 both the the main log and to the reject log were not written to syslog at
2218 PH/14 In the default configuration, change the use of "message" in ACL warn
2219 statements to "add_header".
2221 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2222 not followed by a command (e.g. "seen endif").
2224 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2225 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2228 PH/17 Added extra conditions to the default value of delay_warning_condition
2232 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2233 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2234 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2237 The Auto-Submitted: and various List- headers are standardised, whereas I
2238 don't think Precedence: ever was.
2240 PH/18 Refactored debugging code in route_finduser() to show more information,
2241 in particular, the error code if getpwnam() issues one.
2243 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2244 This is apparently needed in addition to the PH/07 change above to avoid
2245 any possible encoding problems.
2247 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2248 but not after initializing Perl.
2250 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2251 output them only if debugging. By default they are written stderr,
2252 apparently, which is not desirable.
2254 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2257 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2260 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2262 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2263 authenticated or an ident call has been made. Suppress the default
2264 values for $authenticated_id and $authenticated_sender (but permit -oMai
2265 and -oMas) when testing with -bh.
2267 PH/24 Re-jigged the order of the tests in the default configuration so that the
2268 tests for valid domains and recipients precede the DNS black list and CSA
2269 tests, on the grounds that those ones are more expensive.
2271 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2272 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2273 This bug exists in every version of Exim that I still have, right back to
2276 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2277 However, an attempt to turn on case-sensitivity in a regex key by
2278 including (?-i) didn't work because the subject string was already
2279 lowercased, and the effects were non-intuitive. It turns out that a
2280 one-line patch can be used to allow (?-i) to work as expected.
2286 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2287 other effects) broke the use of negated acl sub-conditions.
2289 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2292 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2293 "Deny" causes Exim to reject the incoming connection with a 554 error.
2294 Unfortunately, if there is a major crisis, such as a disk failure,
2295 tcp-wrappers gives "deny", whereas what one would like would be some
2296 kind of temporary error. A kludge has been added to help with this.
2297 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2298 554 error is used if errno is still zero or contains ENOENT (which occurs
2299 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2302 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2304 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2305 errors. Otherwise a message that provokes a temporary error (when other
2306 messages do not) can cause a whole host to time out.
2308 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2309 addresses were routed directly to files or pipes from a redirect router.
2310 File deliveries just didn't batch; pipe deliveries might have suffered
2313 PH/06 A failure to get a lock for a hints database would erroneously always say
2314 "Failed to get write lock", even when it was really a read lock.
2316 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2317 of 0600. This has been changed to use the value of the lockfile_mode
2318 option (which defaults to 0600).
2320 PH/08 Applied small patch from the Sieve maintainer.
2322 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2323 folder from quota calculations, a direct delivery into this folder messed
2324 up the contents of the maildirsize file. This was because the regex was
2325 used only to exclude .Trash (or whatever) when the size of the mailbox
2326 was calculated. There was no check that a delivery was happening into an
2327 excluded directory. This bug has been fixed by ignoring all quota
2328 processing for deliveries into excluded directories.
2330 PH/10 Added the maildirfolder_create_regex option to appendfile.
2336 PH/01 The code for finding all the local interface addresses on a FreeBSD
2337 system running IPv6 was broken. This may well have applied to all BSD
2338 systems, as well as to others that have similar system calls. The broken
2339 code found IPv4 interfaces correctly, but gave incorrect values for the
2340 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2341 that it would not match correctly against @[] and not recognize the IPv6
2344 PH/02 The ipliteral router was not recognizing addresses of the form user@
2345 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2347 PH/03 Added disable_ipv6.
2349 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2350 decoded form, because it is most often used to construct To: headers
2351 lines in autoreplies, and the decoded form may well be syntactically
2352 invalid. However, $reply_address has leading white space removed, and all
2353 newlines turned into spaces so that the autoreply transport does not
2356 PH/05 If group was specified without a user on a router, and no group or user
2357 was specified on a transport, the group from the router was ignored.
2359 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2360 for visible compile-time settings that can be used to change these
2361 numbers, for those that want even more. Backwards compatibility with old
2362 spool files has been maintained. However, going back to a previous Exim
2363 release will lost any variables that are in spool files.
2365 PH/07 Two small changes when running in the test harness: increase delay when
2366 passing a TCP/IP connection to a new process, in case the original
2367 process has to generate a bounce, and remove special handling of
2368 127.0.0.2 (sic), which is no longer necessary.
2370 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2371 be the same on different OS.
2373 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2376 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2377 whether --show-vars was specified or not
2379 JJ/02 exipick: Added support for new ACL variable spool format introduced
2382 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2383 syntactically invalid From: or Reply-to: line, and a filter used this to
2384 generate an autoreply, and therefore failed to obtain an address for the
2385 autoreply, Exim could try to deliver to a non-existent relative file
2386 name, causing unrelated and misleading errors. What now happens is that
2387 it logs this as a hard delivery error, but does not attempt to create a
2390 PH/11 The exinext utility has a -C option for testing purposes, but although
2391 the given file was scanned by exinext itself; it wasn't being passed on
2392 when Exim was called.
2394 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2395 an end-of-file indication when reading a command response.
2397 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2398 compiled. In many other places in Exim, IPv6 addresses are always
2399 recognized, so I have changed this. It also means that IPv4 domain
2400 literals of the form [IPV4:n.n.n.n] are now always recognized.
2402 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2403 used if the router is not running as root, for example, when verifying at
2404 ACL time, or when using -bh. The debugging output from this situation was
2405 non-existent - all you got was a failure to exec. I have made two
2408 (a) Failures to set uid/gid, the current directory, or a process leader
2409 in a subprocess such as that created by queryprogram now generate
2410 suitable debugging ouput when -d is set.
2412 (b) The queryprogram router detects when it is not running as root,
2413 outputs suitable debugging information if -d is set, and then runs
2414 the subprocess without attempting to change uid/gid.
2416 PH/15 Minor change to Makefile for building test_host (undocumented testing
2419 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2420 additional section of a DNS packet that returns MX or SRV records.
2421 Instead, it always explicitly searches for A/AAAA records. This avoids
2422 major problems that occur when a DNS server includes only records of one
2423 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2424 fixed another bug: if SRV records were looked up and the corresponding
2425 address records were *not* found in the additional section, the port
2426 values from the SRV records were lost.
2428 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2429 using the correct key (the original address) when searching the retry
2430 rules in order to find which one to use for generating the retry hint.
2432 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2433 adding the default one. Similarly, if it contains a Reply-To: header, the
2434 errors_reply_to option, if set, is not used.
2436 PH/19 When calculating a retry time, Exim used to measure the "time since
2437 failure" by looking at the "first failed" field in the retry record. Now
2438 it does not use this if it is later than than the arrival time of the
2439 message. Instead it uses the arrival time. This makes for better
2440 behaviour in cases where some deliveries succeed, thus re-setting the
2441 "first failed" field. An example is a quota failure for a huge message
2442 when small messages continue to be delivered. Without this change, the
2443 "time since failure" will always be short, possible causing more frequent
2444 delivery attempts for the huge message than are intended.
2445 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2447 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2448 $1, $2, $3) because the numerical variables can be reset during some
2449 expansion items (e.g. "match"), thereby losing the authentication data.
2451 PH/21 Make -bV show the size of off_t variables so that the test suite can
2452 decide whether to run tests for quotas > 2G.
2454 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2455 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2456 filecount value is greater than 2G or if a quota value is greater than 2G
2457 on a system where the size of off_t is not greater than 4, a panic error
2460 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2461 never match. The debug and -bh output now contains an explicit error
2462 message indicating a malformed IPv4 address or mask.
2464 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2465 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2466 PH/23 above applies.
2468 PH/25 Do not write to syslog when running in the test harness. The only
2469 occasion when this arises is a failure to open the main or panic logs
2470 (for which there is an explicit test).
2472 PH/26 Added the /no_tell option to "control=freeze".
2474 PH/27 If a host name lookup failed very early in a connection, for example, if
2475 the IP address matched host_lookup and the reverse lookup yielded a name
2476 that did not have a forward lookup, an error message of the form "no IP
2477 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2478 could be logged. Now it outputs the IP address instead of "NULL".
2480 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2481 allows the sender and the authenticated sender to be set when
2482 submitting a message from within Exim. Since child_open_exim() is
2483 documented for local_scan(), the new function should be too.
2485 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2486 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2487 results in an empty string is now treated as unset.
2489 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2491 PH/31 Added sender_verify_fail logging option.
2493 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2494 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2495 tidied the source and removed it altogether.
2497 PH/33 When a queue run was abandoned because the load average was too high, a
2498 log line was always written; now it is written only if the queue_run log
2499 selector is set. In addition, the log line for abandonment now contains
2500 information about the queue run such as the pid. This is always present
2501 in "start" and "stop" lines but was omitted from the "abandon" line.
2503 PH/34 Omit spaces between a header name and the colon in the error message that
2504 is given when verify = headers_syntax fails (if there are lots of them,
2505 the message gets confusing).
2507 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2508 names, as there are now some PTR records that contain slashes. This check
2509 is only to protect against broken name servers that fall over on strange
2510 characters, so the fact that it applies to all lookups doesn't matter.
2512 PH/36 Now that the new test suite is complete, we can remove some of the
2513 special code in Exim that was needed for the old test suite. For example,
2514 sorting DNS records because real resolvers return them in an arbitrary
2515 order. The new test suite's fake resolver always returns records in the
2518 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2519 bounces) except when queue_only is set, to avoid logging races between
2520 the different processes.
2522 PH/38 Panic-die if .include specifies a non-absolute path.
2524 PH/39 A tweak to the "H" retry rule from its user.
2526 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2527 a label. They prevented compilation on older perls.
2529 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2530 a warning to be raised on newish perls.
2532 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2533 on queue. Changes to match documented behaviour of showing count of
2534 messages matching specified criteria.
2536 PH/40 Changed the default ident timeout from 30s to 5s.
2538 PH/41 Added support for the use of login_cap features, on those BSD systems
2539 that have them, for controlling the resources used by pipe deliveries.
2541 PH/42 The content-scanning code uses fopen() to create files in which to put
2542 message data. Previously it was not paying any attention to the mode of
2543 the files. Exim runs with umask(0) because the rest of the code creates
2544 files with open(), and sets the required mode explicitly. Thus, these
2545 files were ending up world-writeable. This was not a big issue, because,
2546 being within the spool directory, they were not world-accessible. I have
2547 created a function called modefopen, which takes an additional mode
2548 argument. It sets umask(777), creates the file, chmods it to the required
2549 mode, then resets the umask. All the relevant calls to fopen() in the
2550 content scanning code have been changed to use this function.
2552 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2553 to 24 hours. This avoids potential overflow problems when processing G
2554 and H retry rules. I suspect nobody ever tinkers with this value.
2556 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2558 PH/45 When the plaintext authenticator is running as a client, the server's
2559 challenges are checked to ensure they are valid base64 strings. By
2560 default, the authentication attempt is cancelled if an invalid string is
2561 received. Setting client_ignore_invalid_base64 true ignores these errors.
2562 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2563 they are received. Thus, the responses can be made to depend on the
2564 challenges. If an invalid string is ignored, an empty string is placed in
2567 PH/46 Messages that are created by the autoreply transport now contains a
2568 References: header, in accordance with RFCs 2822 and 3834.
2570 PH/47 Added authenticated_sender_force to the smtp transport.
2572 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2574 PH/49 Installed latest patch from the Sieve maintainer.
2576 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2577 was also set, the appendfile transport was unnecessarily scanning a
2578 directory of message files (e.g. for maildir delivery) to find the count
2579 of files (along with the size), even though it did not need this
2580 information. It now does the scan only if it needs to find either the
2581 size of the count of files.
2583 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2585 PH/52 Two bugs concerned with error handling when the smtp transport is
2588 (i) Exim was not creating retry information for temporary errors given
2589 for individual recipients after the DATA command when the smtp transport
2590 was used in LMTP mode. This meant that they could be retried too
2591 frequently, and not timed out correctly.
2593 (ii) Exim was setting the flag that allows error details to be returned
2594 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2595 recipients that were returned after the DATA command.
2597 PH/53 This is related to PH/52, but is more general: for any failing address,
2598 when detailed error information was permitted to be returned to the
2599 sender, but the error was temporary, then after the final timeout, only
2600 "retry timeout exceeded" was returned. Now it returns the full error as
2601 well as "retry timeout exceeded".
2603 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2604 do this, and (what is worse) MTAs that accept it.
2606 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2607 will now be deprecated.
2609 PH/56 New os.c-cygwin from the Cygwin maintainer.
2611 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2612 formats (previously only available in exim formats via -bpr, -bpru,
2613 and -bpra. Now also available in native and exiqgrep formats)
2615 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2616 with very large, slow to parse queues
2618 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2620 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2622 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2623 responses to authentication challenges, though it was showing the
2624 challenges; (ii) I've removed the CR characters from the debug output for
2627 PH/58 Allow for the insertion of a newline as well as a space when a string
2628 is turned into more than one encoded-word during RFC 2047 encoding. The
2629 Sieve code now uses this.
2631 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2632 data_4xx, lost_connection, tls_required.
2634 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2635 message was being sent as an SMTP response.
2637 PH/61 Add -l and -k options to exicyclog.
2639 PH/62 When verifying, if an address was redirected to one new address, so that
2640 verification continued, and the new address failed or deferred after
2641 having set something in $address_data, the value of $address_data was not
2642 passed back to the ACL. This was different to the case when no
2643 redirection occurred. The value is now passed back in both cases.
2645 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2646 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2647 use login_cap.h, so on its own it isn't the distinguishing feature. The
2648 new name refers directly to the setclassresources() function.
2650 PH/65 Added configuration files for NetBSD3.
2652 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2654 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2655 is preferred over IPv4.
2657 PH/68 The bounce_return_message and bounce_return_body options were not being
2658 honoured for bounces generated during the reception of non-SMTP messages.
2659 In particular, this applied to messages rejected by the ACL. This bug has
2660 been fixed. However, if bounce_return_message is true and bounce_return_
2661 body is false, the headers that are returned for a non-SMTP message
2662 include only those that have been read before the error was detected.
2663 (In the case of an ACL rejection, they have all been read.)
2665 PH/69 The HTML version of the specification is now built in a directory called
2666 spec_html instead of spec.html, because the latter looks like a path with
2667 a MIME-type, and this confuses some software.
2669 PH/70 Catch two compiler warnings in sieve.c.
2671 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2672 function verify_get_ident() calls ip_connect() to connect a socket, but
2673 if the "connect()" function timed out, ip_connect() used to close the
2674 socket. However, verify_get_ident() also closes the socket later, and in
2675 between Exim writes to the log, which may get opened at this point. When
2676 the socket was closed in ip_connect(), the log could get the same file
2677 descriptor number as the socket. This naturally causes chaos. The fix is
2678 not to close the socket in ip_connect(); the socket should be closed by
2679 the function that creates it. There was only one place in the code where
2680 this was missing, in the iplookup router, which I don't think anybody now
2681 uses, but I've fixed it anyway.
2683 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2684 well as to direct DNS lookups. Otherwise the handling of names in host
2685 lists is inconsistent and therefore confusing.
2691 PH/01 Two changes to the default runtime configuration:
2693 (1) Move the checks for relay_from_hosts and authenticated clients from
2694 after to before the (commented out) DNS black list checks.
2696 (2) Add control=submission to the relay_from_hosts and authenticated
2697 clients checks, on the grounds that messages accepted by these
2698 statements are most likely to be submissions.
2700 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2702 (1) Generate an error if the third argument for the ${prvs expansion is
2705 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2708 (3) Reset the variables that are obtained from the first argument of
2709 ${prvscheck and used in the second argument before leaving the code,
2710 because their memory is reclaimed, so using them afterwards may do
2713 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2714 one (it's much easier than Tom thought :-).
2716 (5) Because of (4), we can now allow for the use of $prvscheck_result
2717 inside the third argument.
2719 PH/03 For some reason, the default setting of PATH when running a command from
2720 a pipe transport was just "/usr/bin". I have changed it to
2723 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2724 anything to be listed in the output from -bV.
2726 PH/05 When a filter generated an autoreply, the entire To: header line was
2727 quoted in the delivery log line, like this:
2729 => >A.N.Other <ano@some.domain> <original@ddress> ...
2731 This has been changed so that it extracts the operative address. There
2732 may be more than one such address. If so, they are comma-separated, like
2735 => >ano@some.domain,ona@other.domain <original@ddress> ...
2737 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2738 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2739 not being looked up in the rDNS to get a host name, Exim was showing the
2740 IP address twice in Received: lines, even though the IP addresses were
2741 identical. For example:
2743 Received: from [1.2.3.4] (helo=[1.2.3.4])
2745 However, if the real host name was known, it was omitting the HELO data
2746 if it matched the actual IP address. This has been tidied up so that it
2747 doesn't show the same IP address twice.
2749 PH/07 When both +timestamp and +memory debugging was on, the value given by
2750 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2751 called by the debug printing, thereby overwriting the timestamp buffer.
2752 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2754 PH/08 When the original message was included in an autoreply transport, it
2755 always said "this is a copy of the message, including all the headers",
2756 even if body_only or headers_only was set. It now gives an appropriate
2759 PH/09 Applied a patch from the Sieve maintainer which:
2761 o fixes some comments
2762 o adds the (disabled) notify extension core
2763 o adds some debug output for the result of if/elsif tests
2764 o points to the current vacation draft in the documentation
2765 and documents the missing references header update
2769 o fixes a bug in processing the envelope test (when testing
2770 multiple envelope elements, the last element determinted the
2773 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2774 Electronic Mail") by including:
2776 Auto-submitted: auto-generated
2778 in the messages that it generates (bounce messages and others, such as
2779 warnings). In the case of bounce messages for non-SMTP mesages, there was
2780 also a typo: it was using "Auto_submitted" (underscore instead of
2781 hyphen). Since every message generated by Exim is necessarily in response
2782 to another message, thes have all been changed to:
2784 Auto-Submitted: auto-replied
2786 in accordance with these statements in the RFC:
2788 The auto-replied keyword:
2790 - SHOULD be used on messages sent in direct response to another
2791 message by an automatic process,
2793 - MUST NOT be used on manually-generated messages,
2795 - MAY be used on Delivery Status Notifications (DSNs) and Message
2796 Disposition Notifications (MDNs),
2798 - MUST NOT be used on messages generated by automatic or periodic
2799 processes, except for messages which are automatic responses to
2802 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2803 to the default Received: header definition.
2805 PH/12 Added log selector acl_warn_skipped (default on).
2807 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2808 variables because (a) they are in the wrong storage pool and (b) even if
2809 they were copied, it wouldn't work properly because of the caching.
2811 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2812 checking when decoding. Apparently there are clients that generate
2813 overlong encoded strings. Why am I not surprised?
2815 PH/15 If the first argument of "${if match_address" was not empty, but did not
2816 contain an "@" character, Exim crashed. Now it writes a panic log message
2817 and treats the condition as false.
2819 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2821 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2822 extension "envelope-auth". The code is finished and in agreement with
2823 other implementations, but there is no documentation so far and in fact,
2824 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2825 not changing the active code.
2827 Print executed "if" and "elsif" statements when debugging is used. This
2828 helps a great deal to understand what a filter does.
2830 Document more things not specified clearly in RFC3028. I had all this
2831 sorted out, when out of a sudden new issues came to my mind. Oops."
2833 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2836 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2837 canonical form (as documented). However, after a host name lookup from
2838 the IP address, check_host() was doing a simple string comparison with
2839 addresses acquired from the DNS when checking that the found name did
2840 have the original IP as one of its addresses. Since any found IPv6
2841 addresses are likely to be in abbreviated form, the comparison could
2842 fail. Luckily, there already exists a function for doing the comparison
2843 by converting both addresses to binary, so now that is used instead of
2844 the text comparison.
2846 PH/20 There was another similar case to PH/19, when a complete host name was
2847 given in a host list; looking up its IP address could give an abbreviated
2848 form, whereas the current host's name might or might not be abbreviated.
2849 The same fix has been applied.
2855 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2856 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2859 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2860 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2862 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2864 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2865 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2866 Also, for "Auto-Submitted", treat anything other than "no" as
2867 non-personal, in accordance with RFC 3834. (Previously it treated
2868 anything starting "auto-" as non-personal.)
2870 TF/01 The control=submission/name=... option had a problem with syntax
2871 errors if the name included a slash character. The /name= option
2872 now slurps the rest of the string, so it can include any characters
2873 but it must come last in the list of options (after /sender_retain
2876 PH/05 Some modifications to the interface to the fake nameserver for the new
2884 TK/01 Added the "success_on_redirect" address verification option. See
2885 NewStuff for rationale and an example.
2887 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2889 PH/02 Patch to exigrep to allow it to work on syslog lines.
2891 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2892 fread() to skip over the body file's header line, because in Cygwin the
2893 header line is locked and is inaccessible.
2895 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2896 co-exist for some time) to make it clear that it is the Exim ID that is
2897 referenced, not the Message-ID: header line.
2899 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2900 string_format() function, because snprintf() does not exist on all
2903 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2904 ${stat: expansion item.
2906 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2907 protocol synchronization error", to keep the pedants happy.
2909 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2910 well as for IRIX systems, when gcc is being used. See the host.c source
2913 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2915 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2918 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2919 transport and to the smtp transport in LMTP mode.
2921 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2923 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2924 run a filter in a subprocess. This could lead to confusion in subsequent
2925 lookups in the parent process. There should also be a search_tidyup() at
2926 the end of the subprocess.
2928 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2929 only if the host matched helo_try_verify_hosts, which caused the
2930 verification to occur when the EHLO/HELO command was issued. The ACL just
2931 tested the remembered result. Now, if a previous verification attempt has
2932 not happened, "verify = helo" does it there and then.
2934 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2936 TK/03 Fix log output including CR from clamd.
2938 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2939 exist provoked a memory error which could cause a segfault.
2941 PH/15 Installed PCRE 6.2
2943 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2945 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2946 of the problem. Specifically, suggested +O2 rather than +O1 for the
2949 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2951 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2952 record, the port was not correctly logged when the outgoing_port log
2953 selector was set (it logged the transort's default port).
2955 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2956 fallback_hosts, and "hosts" in the smtp transport.
2958 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2959 host errors such as "Connection refused".
2961 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2962 authentication with radiusclient 0.4.9:
2964 - Error returned from rc_read_config was caught wrongly
2965 - Username/password not passed on to radius server due to wrong length.
2967 The presumption is that some radiusclient API changes for 4.51/PH/17
2968 were not taken care of correctly. The code is still untested by me (my
2969 Linux distribution still has 0.3.2 of radiusclient), but it was
2970 contributed by a Radius user.
2972 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2973 expanding the "port" option of the smtp transport.
2975 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2976 while reading a MIME header. Thanks to Tom Hughes for a patch.
2978 PH/24 Include config.h inside local_scan.h so that configuration settings are
2981 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2982 that in an ACL for RCPT (for example), you can examine exactly what was
2985 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2986 commands, but it was not correctly comparing the address with the actual
2987 client host address. Thus, it would show the EHLO address in Received:
2988 header lines when this was not necessary.
2990 PH/27 Added the % operator to ${eval:}.
2992 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2993 it should be ignoring failures (because with -C, for example, it has lost
2994 privilege). It wasn't ignoring creation failures other than "already
2997 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3000 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3001 because an input line was too long, either on its own, or by virtue of
3002 too many continuations, the temporary file was not being removed, and the
3003 return code was incorrect.
3005 PH/31 Missing "BOOL" in function definition in filtertest.c.
3007 PH/32 Applied Sieve patches from the maintainer.
3009 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3011 PH/33 Added "verify = not_blind".
3013 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3014 Local/Makefile (with some defaults set). These are used in built scripts
3015 such as exicyclog, but they have never been used in the exim_install
3016 script (though there are many overriding facilities there). I have
3017 arranged that the exim_install script now takes note of these two
3020 PH/35 Installed configuration files for Dragonfly.
3022 PH/36 When a locally submitted message by a trusted user did not contain a
3023 From: header, and the sender address was obtained from -f or from an SMTP
3024 MAIL command, and the trusted user did not use -F to supply a sender
3025 name, $originator_name was incorrectly used when constructing a From:
3026 header. Furthermore, $originator_name was used for submission mode
3027 messages from external hosts without From: headers in a similar way,
3028 which is clearly wrong.
3030 PH/37 Added control=suppress_local_fixups.
3032 PH/38 When log_selector = +received_sender was set, and the addition of the
3033 sender made the log line's construction buffer exactly full, or one byte
3034 less than full, an overflow happened when the terminating "\n" was
3037 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3038 when the result of a list match is failure because a DNS lookup failed.
3040 PH/40 RM_COMMAND is now used in the building process.
3042 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3043 the "build-* directories that it finds.
3045 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3046 address in a domain literal was a prefix of an interface address.
3048 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3049 when verifying a sender address, unless rewrite_headers is false.
3051 PH/44 Wrote a long comment about why errors_to addresses are verified as
3052 recipients, not senders.
3054 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3055 the ratelimit ACL was added.
3057 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3059 PH/46 Added extra information about PostgreSQL errors to the error string.
3061 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3062 avoiding the need to install special zones in a real server. This is
3063 backwards compatible; if it can't find the fake resolver, it drops back.
3064 Thus, both old and new test suites can be run.
3066 TF/02 Added util/ratelimit.pl
3068 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3069 clock is set back in time.
3071 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3072 Candler <B.Candler@pobox.com>.
3074 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3075 OK for addresses that are the result of redirections.
3077 PH/48 A number of further additions for the benefit of the new test suite,
3078 including a fake gethostbyname() that interfaces to the fake DNS resolver
3081 TF/06 The fix for widen_domains has also been applied to qualify_single and
3082 search_parents which are the other dnslookup options that can cause
3085 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3088 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3090 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3091 service" if it encountered an MX record with an empty target hostname.
3092 The message is now "an MX or SRV record indicated no SMTP service".
3094 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3095 if the DNS of the sending site is misconfigured. This is quite a
3096 common situation. This change restores the behaviour of treating a
3097 helo verification defer as a failure.
3099 PH/51 If self=fail was set on a router, the bounce message did not include the
3100 actual error message.
3106 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3108 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3109 command itself ended in error, the underlying message about the transport
3110 filter timeout was being overwritten with the pipe command error. Now the
3111 underlying error message should be appended to the second error message.
3113 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3115 PH/02 Reduce the amount of output that "make" produces by default. Full output
3116 can still be requested.
3118 PH/03 The warning log line about a condition test deferring for a "warn" verb
3119 was being output only once per connection, rather than after each
3120 occurrence (because it was using the same function as for successful
3121 "warn" verbs). This seems wrong, so I have changed it.
3123 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3124 it should not have, which might have caused a crash in the right
3125 circumstances, but probably never did.
3127 PH/04 Installed a modified version of Tony Finch's patch to make submission
3128 mode fix the return path as well as the Sender: header line, and to
3129 add a /name= option so that you can make the user's friendly name appear
3132 TF/03 Added the control = fakedefer ACL modifier.
3134 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3135 Mark Lowes for thorough testing.
3137 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3139 TK/03 Merged latest SRS patch from Miles Wilton.
3141 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3142 with the definition in sysexits.h (which is #included earlier).
3143 Fortunately, Exim does not actually use EX_OK. The code used to try to
3144 preserve the sysexits.h value, by assumimg that macro definitions were
3145 scanned for macro replacements. I have been disabused of this notion,
3146 so now the code just undefines EX_OK before #including unistd.h.
3148 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3149 in the smtp transport. When a block could not be written in a single
3150 write() function, the timeout was being re-applied to each part-write.
3151 This seems wrong - if the receiver was accepting one byte at a time it
3152 would take for ever. The timeout is now adjusted when this happens. It
3153 doesn't have to be particularly precise.
3155 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3156 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3158 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3159 without a domain if the check to <postmaster@domain> fails.
3161 SC/01 Eximstats: added -xls and the ability to specify output files
3162 (patch written by Frank Heydlauf).
3164 SC/02 Eximstats: use FileHandles for outputing results.
3166 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3168 SC/04 Eximstats: fixed display of large numbers with -nvr option
3170 SC/05 Eximstats: fixed merging of reports with empty tables.
3172 SC/06 Eximstats: added the -include_original_destination flag
3174 SC/07 Eximstats: removed tabs and trailing whitespace.
3176 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3178 TK/06 MBOX spool code: Add real "From " MBOX separator line
3179 so the .eml file is really in mbox format (even though
3180 most programs do not really care). Patch from Alex Miller.
3182 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3183 The latter is generated from $received_to and is only set if the
3184 message has one envelope recipient. SA can use these headers,
3185 obviously out-of-the-box. Patch from Alex Miller.
3187 PH/08 The ${def test on a variable was returning false if the variable's
3188 value was "0", contrary to what the specification has always said!
3189 The result should be true unless the variable is empty.
3191 PH/09 The syntax error of a character other than { following "${if
3192 def:variable_name" (after optional whitespace) was not being diagnosed.
3193 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3194 accidental colon was present, for example, could give incorrect results.
3196 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3197 result is used (not including appendfile, where other changes are about
3200 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3201 This involved changing a lot of size variables from int to off_t. It
3202 should work with maildirs and everything.
3204 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3205 spamd dying while we are connected to it.
3207 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3210 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3211 function for BDB 4.3.
3213 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3215 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3216 the sources for PCRE have been re-arranged and more files are now
3219 PH/15 The code I had for printing potentially long long variables in PH/11
3220 above was not the best (it lost precision). The length of off_t variables
3221 is now inspected at build time, and an appropriate printing format (%ld
3222 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3223 to be "long long int" or "long int". This is needed for the internal
3224 formatting function string_vformat().
3226 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3227 the configuration file to be ":syslog", then the script "guesses" where
3228 the logs files are, rather than using the compiled in default. In our
3229 case the guess is not the same as the compiled default, so the script
3230 suddenly stopped working when I started to use syslog. The patch checks
3231 to see if log_file_path is "". If so, it attempts to read it from exim
3232 with no configuration file to get the compiled in version, before it
3233 falls back to the previous guessing code."
3235 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3236 implementing BATV in an Exim configuration. See NewStuff for the gory
3239 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3240 Makefile that are specific to HP-UX.
3242 PH/18 If the "use_postmaster" option was set for a recipient callout together
3243 with the "random" option, the postmaster address was used as the MAIL
3244 FROM address for the random test, but not for the subsequent recipient
3245 test. It is now used for both.
3247 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3248 patch removes a few documentation additions to RFC 3028, because the
3249 latest draft now contains them. It adds the new en;ascii-case comparator
3250 and a new error check for 8bit text in MIME parts. Comparator and
3251 require names are now matched exactly. I enabled the subaddress
3252 extension, but it is not well tested yet (read: it works for me)."
3254 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3255 rework some of the code of TK/09 above to avoid the hardwired use of
3256 "%lld" and "long long". Replaced the call to snprintf() with a call to
3259 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3260 records point to non-existent hosts", "retry timeout exceeded", and
3261 "retry time not reached for any host after a long failure period".
3263 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3264 experimental DomainKeys support:
3266 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3267 (2) On an error such as an illegally used "control", the wrong name for
3268 the control was given.
3270 These problems did NOT occur unless DomainKeys support was compiled.
3272 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3274 PH/24 Added ${if match_ip condition.
3276 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3277 databases so that it will be absolutely obvious if a crash occurs in the
3278 DB library. This is a regular occurrence (often caused by mis-matched
3281 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3282 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3283 picked up on a user's system that detects such things. There doesn't seem
3284 to be a gcc warning option for this - only an attribute that has to be
3285 put on the function's prototype. It seems that in Fedora Core 4 they have
3286 set this on a number of new functions. No doubt there will be more in due
3289 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3290 specify a transport. However, if an address that was verified by such a
3291 router was the subject of a callout, Exim crashed because it tried to
3292 read the rcpt_include_affixes from the non-existent transport. Now it
3293 just assumes that the setting of that option is false. This bug was
3294 introduced by 4.51/PH/31.
3296 PH/28 Changed -d+all to exclude +memory, because that information is very
3297 rarely of interest, but it makes the output a lot bigger. People tend to
3298 do -d+all out of habit.
3300 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3301 code in os-type was giving problems when libc.so lives in lib64, like on
3304 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3305 aren't the modern standard, and it seems that some systems' include files
3306 don't always have them. Exim was already checking for some of the newer
3307 ones like T_AAAA, and defining it itself. I've added checks for all the
3308 record types that Exim uses.
3310 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3311 not automatically generating a new one, as it is supposed to. This
3312 prevented TLS from working. If the file did exist, but contained invalid
3313 data, a new version was generated, as expected. It was only the case of a
3314 non-existent file that was broken.
3316 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3317 with a change in libdomainkeys > 0.64.
3319 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3320 from DNS. If the selector record carries the flag, it now has
3321 precedence over the domain-wide flag.
3323 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3325 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3326 the use of a port name that isn't defined in /etc/services) occurred, the
3327 message was deferred as in a normal delivery, and thus remained on the
3328 spool, instead of being failed because of the mua_wrapper setting. This
3329 is now fixed, and I tidied up some of the mua_wrapper messages at the
3332 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3333 the messages in a hash of arrays rather than using individual hashes.
3334 This is a bit cleaner and results in dramatic memory savings, albeit
3335 at a slight CPU cost.
3337 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3338 as requested by Marc Sherman.
3340 SC/10 Eximstats: added histograms for user specified patterns as requested
3343 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3345 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3346 fopen() in the content-scanning modules that did not already have it.
3352 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3353 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3355 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3357 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3359 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3360 to test Sieve filters that use "vacation".
3362 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3363 that changes the way the GnuTLS parameters are stored in the cache file.
3364 The new format can be generated externally. For backward compatibility,
3365 if the data in the cache doesn't make sense, Exim assumes it has read an
3366 old-format file, and it generates new data and writes a new file. This
3367 means that you can't go back to an older release without removing the
3370 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3371 work if there are any delivery delays because "one_time" forces the
3372 parent to be marked "delivered", so its unseen clone is never tried
3373 again. For this reason, Exim now forbids the simultaneous setting of
3376 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3377 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3378 entry if you want to know the details. The fix, however, neglected to
3379 consider the case when local delivery batching is involved. The test for
3380 "previously delivered" was not happening when checking to see if an
3381 address could be batched with a previous (undelivered) one; under
3382 certain circumstances this could lead to multiple deliveries to the same
3385 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3386 in its include files, and this causes problems building Exim.
3388 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3389 header_syntax) but Exim was just ignoring anything given after a slash.
3390 In particular, this caused confusion with an attempt to use "verify =
3391 reverse_host_lookup/defer_ok". An error is now given when options are
3392 supplied for verify items that do not have them. (Maybe reverse_host_
3393 lookup should have a defer_ok option, but that's a different point.)
3395 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3396 defined by RFC 821) to 2048, because there were problems with some AUTH
3397 commands, and RFC 1869 says the size should be increased for extended
3398 SMTP commands that take arguments.
3400 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3403 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3404 "unknown" error; now it says that the functionality isn't in the binary.
3406 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3407 an address' error message when a string expansion fails (syntax or
3408 whatever). Otherwise the password may appear in the log. Following change
3409 PH/42 below, there is no longer a chance of it appearing in a bounce
3412 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3414 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3415 of its addresses was ever tried. (Bugzilla bug #2.)
3417 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3418 the result incorrectly in the debug output. (It correctly added a newline
3419 to what was transported.)
3421 TF/01 Added $received_time.
3423 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3424 commented out examples of how to interface to a virus scanner and to
3425 SpamAssassin. Also added commented examples of av_scanner and
3426 spamd_address settings.
3428 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3429 and controls are allowed in which ACLs. There were a couple of minor
3430 errors. Some of the entries in the conditions table (which is a table of
3431 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3432 negation of where the condition IS allowed.
3434 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3436 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3437 header file does not have a version number, so I've had to invent a new
3438 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3439 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3440 radiusclient), but it was contributed by a Radius user.
3442 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3443 files or directories, for OpenSSL.
3445 PH/19 When an Exim process that is running as root has to create an Exim log
3446 file, it does so in a subprocess that runs as exim:exim so as to get the
3447 ownership right at creation (otherwise, other Exim processes might see
3448 the file with the wrong ownership). There was no test for failure of this
3449 fork() call, which would lead to the process getting stuck as it waited
3450 for a non-existent subprocess. Forks do occasionally fail when resources
3451 run out. I reviewed all the other calls to fork(); they all seem to check
3454 PH/20 When checking for unexpected SMTP input at connect time (before writing
3455 the banner), Exim was not dealing correctly with a non-positive return
3456 from the read() function. If the client had disconnected by this time,
3457 the result was a log entry for a synchronization error with an empty
3458 string after "input=" when read() returned zero. If read() returned -1
3459 (an event I could not check), uninitialized data bytes were printed.
3460 There were reports of junk text (parts of files, etc) appearing after
3463 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3465 PH/22 Added support for macro redefinition, and (re)definition in between
3466 driver and ACL definitions.
3468 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3469 forgetting to use the resulting value; it was using the unexpanded value.
3471 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3472 hadn't been configured. The fix is from Juergen Kreileder, who
3473 understands it better than I do:
3475 "Here's what I see happening with three configured cyrus_sasl
3476 authenticators configured (plain, login, cram-md5):
3478 On startup auth_cyrus_sasl_init() gets called for each of these.
3479 This means three calls to sasl_listmech() without a specified mech_list.
3480 => SASL tests which mechs of all available mechs actually work
3481 => three warnings about OTP not working
3482 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3484 With the patch, sasl_listmech() also gets called three times. But now
3485 SASL's mech_list option is set to the server_mech specified in the the
3486 authenticator. Or in other words, the answer from sasl_listmech()
3487 gets limited to just the mech you're testing for (which is different
3489 => the return list contains just 'plain' or 'login', 'cram-md5' or
3490 nothing depending on the value of ob->server_mech.
3492 I've just tested the patch: Authentication still works fine,
3493 unavailable mechs specified in the exim configuration are still
3494 caught, and the auth.log warnings about OTP are gone."
3496 PH/25 When debugging is enabled, the contents of the command line are added
3497 to the debugging output, even when log_selector=+arguments is not
3500 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3501 answer is "GNU", and only if the return is "GNU/something" is the answer
3504 PH/27 $acl_verify_message is now set immediately after the failure of a
3505 verification in an ACL, and so is available in subsequent modifiers. In
3506 particular, the message can be preserved by coding like this:
3508 warn !verify = sender
3509 set acl_m0 = $acl_verify_message
3511 Previously, $acl_verify_message was set only while expanding "message"
3512 and "log_message" when a very denied access.
3514 PH/28 Modified OS/os.c-Linux with
3516 -#ifndef OS_LOAD_AVERAGE
3517 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3519 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3520 nomenclature these days.)
3522 PH/29 Installed patch from the Sieve maintainer that adds the options
3523 sieve_useraddress and sieve_subaddress to the redirect router.
3525 PH/30 In these circumstances:
3526 . Two addresses routed to the same list of hosts;
3527 . First host does not offer TLS;
3528 . First host accepts first address;
3529 . First host gives temporary error to second address;
3530 . Second host offers TLS and a TLS session is established;
3531 . Second host accepts second address.
3532 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3533 and peerdn, if requested) that were in fact used only for the second
3536 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3537 attention to any local part prefix or suffix that was matched by the
3538 router that accepted the address. It now behaves in the same way as it
3539 does for delivery: the affixes are removed from the local part unless
3540 rcpt_include_affixes is set on the transport.
3542 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3543 timeout during the DATA phase of an incoming message.
3545 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3546 applied a patch sanctioned by the Sieve maintainer.
3548 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3549 the uid or gid is negative. A case of a negative gid caused this to be
3550 noticed. The fix allows for either to be negative.
3552 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3553 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3556 PH/36 The change PH/12 above was broken. Fixed it.
3558 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3559 the grounds that routing the same address twice would always produce the
3560 same answer. This might have been true once, but it is certainly no
3561 longer true now. Routing a child address may depend on the previous
3562 routing that produced that child. Some complicated redirection strategies
3563 went wrong when messages had multiple recipients, and made Exim's
3564 behaviour dependent on the order in which the addresses were given.
3566 I have moved the duplicate checking until after the routing is complete.
3567 Exim scans the addresses that are assigned to local and remote
3568 transports, and removes any duplicates. This means that more work will be
3569 done, as duplicates will always all be routed, but duplicates are
3570 presumably rare, so I don't expect this is of any significance.
3572 For deliveries to pipes, files, and autoreplies, the duplicate checking
3573 still happens during the routing process, since they are not going to be
3576 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3577 It corrects a timeout issue with spamd. This is Ian's comment: "The
3578 background is that sometimes spamd either never reads data from a
3579 connection it has accepted, or it never writes response data. The exiscan
3580 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3581 blindly assumes that writes won't block so it may never time out."
3583 PH/39 Allow G after quota size as well as K and M.
3585 PH/40 The value set for $authenticated_id in an authenticator may not contain
3586 binary zeroes or newlines because the value is written to log lines and
3587 to spool files. There was no check on this. Now the value is run through
3588 the string_printing() function so that such characters are converted to
3589 printable escape sequences.
3591 PH/41 $message_linecount is a new variable that contains the total number of
3592 lines in the message. Compare $body_linecount, which is the count for the
3595 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3596 bounce and delay warning messages, except in certain special cases, which
3599 (a) An SMTP error message from a remote host;
3600 (b) A message specified in a :fail: redirection;
3601 (c) A message specified in a "fail" command in a system filter;
3602 (d) A message specified in a FAIL return from the queryprogram router;
3603 (e) A message specified by the cannot_route_message router option.
3605 In these cases only, Exim does include the error details in bounce and
3606 warning messages. There are also a few cases where bland messages such
3607 as "unrouteable address" or "local delivery error" are given.
3609 PH/43 $value is now also set for the "else" part of a ${run expansion.
3611 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3612 being worked on, but at least Exim now implements the latest version to
3615 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3616 process to complete was treated as a delivery failure, a timeout while
3617 writing the message to the pipe was logged, but erroneously treated as a
3618 successful delivery. Such timeouts include transport filter timeouts. For
3619 consistency with the overall process timeout, these timeouts are now
3620 treated as errors, giving rise to delivery failures by default. However,
3621 there is now a new Boolean option for the pipe transport called
3622 timeout_defer, which, if set TRUE, converts the failures into defers for
3623 both kinds of timeout. A transport filter timeout is now identified in
3626 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3627 systems where "make" and "gmake" are different, calling "gmake" at top
3628 level broke things. I've arranged for the value of $(MAKE) to be passed
3629 from the Makefile to this script so that it can call the same version of
3633 A note about Exim versions 4.44 and 4.50
3634 ----------------------------------------
3636 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3637 changes of various kinds. As a consequence, a big documentation update was
3638 needed. This delayed the release for rather longer than seemed good, especially
3639 in the light of a couple of (minor) security issues. Therefore, the changes
3640 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3641 release. So 4.44 and 4.50 are in effect two different branches that both start
3644 I have left the 4.50 change log unchanged; it contains all the changes since
3645 4.43. The change log for 4.44 is below; many of its items are identical to
3646 those for 4.50. This seems to be the most sensible way to preserve the
3647 historical information.
3653 1. Minor wording change to the doc/README.SIEVE file.
3655 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3656 computation of the current number of files was incorrect.
3658 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3659 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3662 4. Give more explanation in the error message when the command for a transport
3663 filter fails to execute.
3665 5. There are several places where Exim runs a non-Exim command in a
3666 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3667 was being done only for the command run by the queryprogram router. It is
3668 now done for all such subprocesses. The other cases are: ${run, transport
3669 filters, and the commands run by the lmtp and pipe transports.
3671 6. Added CONFIGURE_GROUP build-time option.
3673 7. Some older OS have a limit of 256 on the maximum number of file
3674 descriptors. Exim was using setrlimit() to set 1000 as a large value
3675 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3676 systems. I've change it so that if it can't get 1000, it tries for 256.
3678 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3679 was an oversight, and furthermore, ever since the addition of extra
3680 controls (e.g. 4.43/32), the checks on when to allow different forms of
3681 "control" were broken. There should now be diagnostics for all cases when a
3682 control that does not make sense is encountered.
3684 9. Added the /retain_sender option to "control=submission".
3686 10. $recipients is now available in the predata ACL (oversight).
3688 11. Tidy the search cache before the fork to do a delivery from a message
3689 received from the command line. Otherwise the child will trigger a lookup
3690 failure and thereby defer the delivery if it tries to use (for example) a
3691 cached ldap connection that the parent has called unbind on.
3693 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3694 of $address_data from the recipient verification was clobbered by the
3695 sender verification.
3697 13. The value of address_data from a sender verification is now available in
3698 $sender_address_data in subsequent conditions in the ACL statement.
3700 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3702 15. Added a new option "connect=<time>" to callout options, to set a different
3705 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3706 was its contents. (It was OK if the option was not defined at all.)
3708 17. A "Completed" log line is now written for messages that are removed from
3709 the spool by the -Mrm option.
3711 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3712 information about exactly what failed.
3714 19. Added -dd to debug only the daemon process.
3716 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3717 handles timeouts, both on the server side and network timeouts. Renamed the
3718 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3720 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3721 It is now set to "smtps".
3723 22. $host_address is now set to the target address during the checking of
3724 ignore_target_hosts.
3726 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3727 being passed; this would have caused $sender_host_name to have been used if
3728 matching the list had actually called for a host name (not very likely,
3729 since this list is usually IP addresses). A host name is now passed as
3732 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3733 code that specifies a non-restarting handler (typically sigaction() in
3734 modern systems) in an attempt to fix a rare and obscure crash bug.
3736 25. Narrowed the window for a race in the daemon that could cause it to ignore
3737 SIGCHLD signals. This is not a major problem, because they are used only to
3738 wake it up if nothing else does.
3740 26. A malformed maildirsize file could cause Exim to calculate negative values
3741 for the mailbox size or file count. Odd effects could occur as a result.
3742 The maildirsize information is now recalculated if the size or filecount
3745 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3746 support for a long time. Removed HAVE_SYS_VFS_H.
3748 28. Installed the latest version of exipick from John Jetmore.
3750 29. In an address list, if the pattern was not a regular expression, an empty
3751 subject address (from a bounce message) matched only if the pattern was an
3752 empty string. Non-empty patterns were not even tested. This was the wrong
3753 because it is perfectly reasonable to use an empty address as part of a
3754 database query. An empty address is now tested by patterns that are
3755 lookups. However, all the other forms of pattern expect the subject to
3756 contain a local part and a domain, and therefore, for them, an empty
3757 address still always fails if the pattern is not itself empty.
3759 30. Exim went into a mad DNS loop when attempting to do a callout where the
3760 host was specified on an smtp transport, and looking it up yielded more
3761 than one IP address.
3763 31. Re-factored the code for checking spool and log partition space into a
3764 function that finds that data and another that does the check. The former
3765 is then used to implement four new variables: $spool_space, $log_space,
3766 $spool_inodes, and $log_inodes.
3768 32. The RFC2047 encoding function was originally intended for short strings
3769 such as real names; it was not keeping to the 75-character limit for
3770 encoded words that the RFC imposes. It now respects the limit, and
3771 generates multiple encoded words if necessary. To be on the safe side, I
3772 have increased the buffer size for the ${rfc2047: expansion operator from
3775 33. It is now permitted to omit both strings after an "if" condition; if the
3776 condition is true, the result is "true". As before, when the second string
3777 is omitted, a false condition yields an empty string. This makes it less
3778 cumbersome to write custom ACL and router conditions.
3780 34. Failure to deliver a bounce message always caused it to be frozen, even if
3781 there was an errors_to setting on the router. The errors_to setting is now
3784 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3785 canonical form (fully expanded) before being placed in
3786 $sender_host_address.
3788 36. The table in the code that translates DNS record types into text (T_A to
3789 "A" for instance) was missing entries for NS and CNAME. It is just possible
3790 that this could have caused confusion if both these types were looked up
3791 for the same domain, because the text type is used as part of Exim's
3792 per-process caching. But the chance of anyone hitting this buglet seems
3795 37. The dnsdb lookup has been extended in a number of ways.
3797 (1) There is a new type, "zns", which walks up the domain tree until it
3798 finds some nameserver records. It should be used with care.
3800 (2) There is a new type, "mxh", which is like "mx" except that it returns
3801 just the host names, not the priorities.
3803 (3) It is now possible to give a list of domains (or IP addresses) to be
3804 looked up. The behaviour when one of the lookups defers can be
3805 controlled by a keyword.
3807 (4) It is now possible to specify the separator character for use when
3808 multiple records are returned.
3810 38. The dnslists ACL condition has been extended: it is now possible to supply
3811 a list of IP addresses and/or domains to be looked up in a particular DNS
3814 39. Added log_selector=+queue_time_overall.
3816 40. When running the queue in the test harness, wait just a tad after forking a
3817 delivery process, to get repeatability of debugging output.
3819 41. Include certificate and key file names in error message when GnuTLS fails
3820 to set them up, because the GnuTLS error message doesn't include the name
3821 of the failing file when there is a problem reading it.
3823 42. Allow both -bf and -bF in the same test run.
3825 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3827 44. The "Exiscan patch" is now merged into the mainline Exim source.
3829 45. Sometimes the final signoff response after QUIT could fail to get
3830 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3831 < 0 before doing a fflush(). This bug looks as though it goes back to the
3832 introduction of TLS in release 3.20, but "sometimes" must have been rare
3833 because the tests only now provoked it.
3835 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3836 (this can affect the format of dates).
3838 47. exim_tidydb, when checking for the continued existence of a message for
3839 which it has found a message-specific retry record, was not finding
3840 messages that were in split spool directories. Consequently, it was
3841 deleting retry records that should have stayed in existence.
3843 48. Steve fixed some bugs in eximstats.
3845 49. The SPA authentication driver was not abandoning authentication and moving
3846 on to the next authenticator when an expansion was forced to fail,
3847 contradicting the general specification for all authenticators. Instead it
3848 was generating a temporary error. It now behaves as specified.
3850 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3851 (the order specifies the preference for clients). The order is now AES256,
3852 AES128, 3DES, ARCFOUR128.
3854 51. Small patch to Sieve code - explicitly set From: when generating an
3857 52. Exim crashed if a remote delivery caused a very long error message to be
3858 recorded - for instance if somebody sent an entire SpamAssassin report back
3859 as a large number of 550 error lines. This bug was coincidentally fixed by
3860 increasing the size of one of Exim's internal buffers (big_buffer) that
3861 happened as part of the Exiscan merge. However, to be on the safe side, I
3862 have made the code more robust (and fixed the comments that describe what
3865 53. Now that there can be additional text after "Completed" in log lines (if
3866 the queue_time_overall log selector is set), a one-byte patch to exigrep
3867 was needed to allow it to recognize "Completed" as not the last thing in
3870 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3871 patch that reportedly fixes this has been added. I am not expert enough to
3872 create a test for it. This is what the patch creator wrote:
3874 "I found a little strange behaviour of ldap code when working with
3875 Windows 2003 AD Domain, where users was placed in more than one
3876 Organization Units. When I tried to give exim partial DN, the exit code
3877 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3878 But simultaneously result of request was absolutely normal ldap result,
3879 so I produce this patch..."
3881 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3882 REFERENCE, so I have modified the code to exclude the patch when that macro
3885 55. Some experimental protocols are using DNS PTR records for new purposes. The
3886 keys for these records are domain names, not reversed IP addresses. The
3887 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3888 leaves it alone. Component reversal etc. now happens only for IP addresses.
3891 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3893 57. Double the size of the debug message buffer (to 2048) so that more of very
3894 long debug lines gets shown.
3896 58. The exicyclog utility now does better if the number of log files to keep
3897 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3899 59. Two changes related to the smtp_active_hostname option:
3901 (1) $smtp_active_hostname is now available as a variable.
3902 (2) The default for smtp_banner uses $smtp_active_hostname instead
3903 of $primary_hostname.
3905 60. The host_aton() function is supposed to be passed a string that is known
3906 to be a valid IP address. However, in the case of IPv6 addresses, it was
3907 not checking this. This is a hostage to fortune. Exim now panics and dies
3908 if the condition is not met. A case was found where this could be provoked
3909 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3910 components; fortuitously, this particular loophole had already been fixed
3911 by change 4.50/55 above.
3913 If there are any other similar loopholes, the new check in host_aton()
3914 itself should stop them being exploited. The report I received stated that
3915 data on the command line could provoke the exploit when Exim was running as
3916 exim, but did not say which command line option was involved. All I could
3917 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3918 running as the user.
3921 61. There was a buffer overflow vulnerability in the SPA authentication code
3922 (which came originally from the Samba project). I have added a test to the
3923 spa_base64_to_bits() function which I hope fixes it.
3926 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3927 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3929 63. The daemon start-up calls getloadavg() while still root for those OS that
3930 need the first call to be done as root, but it missed one case: when
3931 deliver_queue_load_max is set with deliver_drop_privilege. This is
3932 necessary for the benefit of the queue runner, because there is no re-exec
3933 when deliver_drop_privilege is set.
3935 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3936 This has been fixed.
3938 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3939 was in use, was not putting the data itself into the right store pool;
3940 consequently, it could be overwritten for a subsequent message in the same
3941 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3944 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3946 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3947 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3948 the function were treating the return as a boolean value, which happened to
3949 work because 0=false and not-0=true, but is not correct code.
3951 68. The host_aton() function was not handling scoped IPv6 addresses (those
3952 with, for example, "%eth0" on the end) correctly.
3954 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3955 negated items (that is, ~something) in unsigned ints. Some compilers
3956 apparently mutter when there is no cast.
3958 70. If an address verification called from an ACL failed, and did not produce a
3959 user-specific message (i.e. there was only a "system" message), nothing was
3960 put in $acl_verify_message. In this situation, it now puts the system
3963 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3964 session; change 4.31/43 added the unwanted input to the log line - except
3965 that it did not do this in the start of session case. It now does.
3967 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3968 This is wrong and can cause the other end to generate a synchronization
3969 error if it is another Exim or anything else that does the synchronization
3970 check. A QUIT command is no longer sent after a timeout.
3972 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3973 during host lookups.
3975 74. The defer_ok option of callout verification was not working if it was used
3976 when verifying addresses in header lines, that is, for this case:
3978 verify = header_sender/callout=defer_ok
3980 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3981 those file descriptors could be used for SMTP connections. If anything
3982 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3983 could be sent to the SMTP client, causing chaos. The daemon now opens
3984 stdin, stdout, and stderr to /dev/null when it puts itself into the
3987 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3988 log by default. The user can override this with suitable Perl magic.
3990 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3991 the log message when discard triggers, was not working for the DATA ACL or
3992 for the non-SMTP ACL.
3994 78. Error message wording change in sieve.c.
3996 79. If smtp_accept_max_per_host was set, the number of connections could be
3997 restricted to fewer than expected, because the daemon was trying to set up
3998 a new connection before checking whether the processes handling previous
3999 connections had finished. The check for completed processes is now done
4000 earlier. On busy systems, this bug wouldn't be noticed because something
4001 else would have woken the daemon, and it would have reaped the completed
4004 80. If a message was submitted locally by a user whose login name contained one
4005 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4006 It caused a spool format error. I have fixed the spool reading code. A
4007 related problem was that the "from" clause in the Received: line became
4008 illegal because of the space(s). It is now covered by ${quote_local_part.
4010 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4013 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4015 83. Patch from Sieve maintainer to fix unterminated string problem in
4016 "vacation" handling.
4018 84. Some minor changes to the Linux configuration files to help with other
4019 OS variants using glibc.
4021 85. One more patch for Sieve to update vacation handling to latest spec.
4024 ----------------------------------------------------
4025 See the note above about the 4.44 and 4.50 releases.
4026 ----------------------------------------------------
4032 1. Change 4.43/35 introduced a bug that caused file counts to be
4033 incorrectly computed when quota_filecount was set in an appendfile
4036 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4037 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4040 3. Give more explanation in the error message when the command for a transport
4041 filter fails to execute.
4043 4. There are several places where Exim runs a non-Exim command in a
4044 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4045 was being done only for the command run by the queryprogram router. It is
4046 now done for all such subprocesses. The other cases are: ${run, transport
4047 filters, and the commands run by the lmtp and pipe transports.
4049 5. Some older OS have a limit of 256 on the maximum number of file
4050 descriptors. Exim was using setrlimit() to set 1000 as a large value
4051 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4052 systems. I've change it so that if it can't get 1000, it tries for 256.
4054 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4055 was an oversight, and furthermore, ever since the addition of extra
4056 controls (e.g. 4.43/32), the checks on when to allow different forms of
4057 "control" were broken. There should now be diagnostics for all cases when a
4058 control that does not make sense is encountered.
4060 7. $recipients is now available in the predata ACL (oversight).
4062 8. Tidy the search cache before the fork to do a delivery from a message
4063 received from the command line. Otherwise the child will trigger a lookup
4064 failure and thereby defer the delivery if it tries to use (for example) a
4065 cached ldap connection that the parent has called unbind on.
4067 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4068 of $address_data from the recipient verification was clobbered by the
4069 sender verification.
4071 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4072 was its contents. (It was OK if the option was not defined at all.)
4074 11. A "Completed" log line is now written for messages that are removed from
4075 the spool by the -Mrm option.
4077 12. $host_address is now set to the target address during the checking of
4078 ignore_target_hosts.
4080 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4081 being passed; this would have caused $sender_host_name to have been used if
4082 matching the list had actually called for a host name (not very likely,
4083 since this list is usually IP addresses). A host name is now passed as
4086 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4087 code that specifies a non-restarting handler (typically sigaction() in
4088 modern systems) in an attempt to fix a rare and obscure crash bug.
4090 15. Narrowed the window for a race in the daemon that could cause it to ignore
4091 SIGCHLD signals. This is not a major problem, because they are used only to
4092 wake it up if nothing else does.
4094 16. A malformed maildirsize file could cause Exim to calculate negative values
4095 for the mailbox size or file count. Odd effects could occur as a result.
4096 The maildirsize information is now recalculated if the size or filecount
4099 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4100 support for a long time. Removed HAVE_SYS_VFS_H.
4102 18. Updated exipick to current release from John Jetmore.
4104 19. Allow an empty sender to be matched against a lookup in an address list.
4105 Previously the only cases considered were a regular expression, or an
4108 20. Exim went into a mad DNS lookup loop when doing a callout where the
4109 host was specified on the transport, if the DNS lookup yielded more than
4112 21. The RFC2047 encoding function was originally intended for short strings
4113 such as real names; it was not keeping to the 75-character limit for
4114 encoded words that the RFC imposes. It now respects the limit, and
4115 generates multiple encoded words if necessary. To be on the safe side, I
4116 have increased the buffer size for the ${rfc2047: expansion operator from
4119 22. Failure to deliver a bounce message always caused it to be frozen, even if
4120 there was an errors_to setting on the router. The errors_to setting is now
4123 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4124 canonical form (fully expanded) before being placed in
4125 $sender_host_address.
4127 24. Updated eximstats to version 1.33
4129 25. Include certificate and key file names in error message when GnuTLS fails
4130 to set them up, because the GnuTLS error message doesn't include the name
4131 of the failing file when there is a problem reading it.
4133 26. Expand error message when OpenSSL has problems setting up cert/key files.
4136 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4137 (this can affect the format of dates).
4139 28. exim_tidydb, when checking for the continued existence of a message for
4140 which it has found a message-specific retry record, was not finding
4141 messages that were in split spool directories. Consequently, it was
4142 deleting retry records that should have stayed in existence.
4144 29. eximstats updated to version 1.35
4145 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4146 1.35 - bugfix such that pie charts by volume are generated correctly
4148 30. The SPA authentication driver was not abandoning authentication and moving
4149 on to the next authenticator when an expansion was forced to fail,
4150 contradicting the general specification for all authenticators. Instead it
4151 was generating a temporary error. It now behaves as specified.
4153 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4154 (the order specifies the preference for clients). The order is now AES256,
4155 AES128, 3DES, ARCFOUR128.
4157 31. Small patch to Sieve code - explicitly set From: when generating an
4160 32. Exim crashed if a remote delivery caused a very long error message to be
4161 recorded - for instance if somebody sent an entire SpamAssassin report back
4162 as a large number of 550 error lines. This bug was coincidentally fixed by
4163 increasing the size of one of Exim's internal buffers (big_buffer) that
4164 happened as part of the Exiscan merge. However, to be on the safe side, I
4165 have made the code more robust (and fixed the comments that describe what
4168 33. Some experimental protocols are using DNS PTR records for new purposes. The
4169 keys for these records are domain names, not reversed IP addresses. The
4170 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4171 leaves it alone. Component reversal etc. now happens only for IP addresses.
4174 34. The host_aton() function is supposed to be passed a string that is known
4175 to be a valid IP address. However, in the case of IPv6 addresses, it was
4176 not checking this. This is a hostage to fortune. Exim now panics and dies
4177 if the condition is not met. A case was found where this could be provoked
4178 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4179 components; fortuitously, this particular loophole had already been fixed
4180 by change 4.50/55 or 4.44/33 above.
4182 If there are any other similar loopholes, the new check in host_aton()
4183 itself should stop them being exploited. The report I received stated that
4184 data on the command line could provoke the exploit when Exim was running as
4185 exim, but did not say which command line option was involved. All I could
4186 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4187 running as the user.
4190 35. There was a buffer overflow vulnerability in the SPA authentication code
4191 (which came originally from the Samba project). I have added a test to the
4192 spa_base64_to_bits() function which I hope fixes it.
4195 36. The daemon start-up calls getloadavg() while still root for those OS that
4196 need the first call to be done as root, but it missed one case: when
4197 deliver_queue_load_max is set with deliver_drop_privilege. This is
4198 necessary for the benefit of the queue runner, because there is no re-exec
4199 when deliver_drop_privilege is set.
4201 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4202 was in use, was not putting the data itself into the right store pool;
4203 consequently, it could be overwritten for a subsequent message in the same
4204 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4207 38. Sometimes the final signoff response after QUIT could fail to get
4208 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4209 < 0 before doing a fflush(). This bug looks as though it goes back to the
4210 introduction of TLS in release 3.20, but "sometimes" must have been rare
4211 because the tests only now provoked it.
4217 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4218 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4219 Now it returns an integer. A number of calls were still expecting a T/F
4220 return. Fortuitously, in all cases, the tests worked in OK situations,
4221 which is the norm. However, things would have gone wrong on any write
4222 failures on the smtp file descriptor. This function is used when sending
4223 messages over SMTP and also when doing verify callouts.
4225 2. When Exim is called to do synchronous delivery of a locally submitted
4226 message (the -odf or -odi options), it no longer closes stderr before doing
4229 3. Implemented the mua_wrapper option.
4231 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4233 5. Implemented the functions header_remove(), header_testname(),
4234 header_add_at_position(), and receive_remove_recipient(), and exported them
4237 6. If an ACL "warn" statement specified the addition of headers, Exim already
4238 inserted X-ACL-Warn: at the start if there was no header name. However, it
4239 was not making this test for the second and subsequent header lines if
4240 there were newlines in the string. This meant that an invalid header could
4241 be inserted if Exim was badly configured.
4243 7. Allow an ACL "warn" statement to add header lines at the start or after all
4244 the Received: headers, as well as at the end.
4246 8. Added the rcpt_4xx retry error code.
4248 9. Added postmaster_mailfrom=xxx to callout verification option.
4250 10. Added mailfrom=xxxx to the callout verification option, for verify=
4253 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4254 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4256 12. Inserted some casts to stop certain compilers warning when using pointer
4257 differences as field lengths or precisions in printf-type calls (mostly
4258 affecting debugging statements).
4260 13. Added optional readline() support for -be (dynamically loaded).
4262 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4263 same clock tick as a message's arrival, so that its received time was the
4264 same as the "first fail" time on the retry record, and that message
4265 remained on the queue past the ultimate address timeout, every queue runner
4266 would try a delivery (because it was past the ultimate address timeout) but
4267 after another failure, the ultimate address timeout, which should have then
4268 bounced the address, did not kick in. This was a "< instead of <=" error;
4269 in most cases the first failure would have been in the next clock tick
4270 after the received time, and all would be well.
4272 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4273 being recognized when the domain list was tested by the match_domain
4274 condition in an expansion string.
4276 16. Added the ${str2b64: operator.
4278 17. Exim was always calling setrlimit() to set a large limit for the number of
4279 processes, without checking whether the existing limit was already
4280 adequate. (It did check for the limit on file descriptors.) Furthermore,
4281 errors from getrlimit() and setrlimit() were being ignored. Now they are
4282 logged to the main and panic logs, but Exim does carry on, to try to do its
4283 job under whatever limits there are.
4285 18. Imported PCRE 5.0.
4287 19. Trivial typo in log message " temporarily refused connection" (the leading
4290 20. If the log selector return_path_on_delivery was set and an address was
4291 redirected to /dev/null, the delivery process crashed because it assumed
4292 that a return path would always be set for a "successful" delivery. In this
4293 case, the whole delivery is bypassed as an optimization, and therefore no
4296 21. Internal re-arrangement: the function for sending a challenge and reading
4297 a response while authentication was assuming a zero-terminated challenge
4298 string. It's now changed to take a pointer and a length, to allow for
4299 binary data in such strings.
4301 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4303 23. Exim was not respecting finduser_retries when seeking the login of the
4304 uid under which it was called; it was always trying 10 times. (The default
4305 setting of finduser_retries is zero.) Also, it was sleeping after the final
4306 failure, which is pointless.
4308 24. Implemented tls_on_connect_ports.
4310 25. Implemented acl_smtp_predata.
4312 26. If the domain in control=submission is set empty, Exim assumes that the
4313 authenticated id is a complete email address when it generates From: or
4314 Sender: header lines.
4316 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4317 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4318 chown and chgrp in /bin and hostname in /usr/bin.
4320 28. Exim was keeping the "process log" file open after each use, just as it
4321 does for the main log. This opens the possibility of it remaining open for
4322 long periods when the USR1 signal hits a daemon. Occasional processlog
4323 errors were reported, that could have been caused by this. Anyway, it seems
4324 much more sensible not to leave this file open at all, so that is what now
4327 29. The long-running daemon process does not normally write to the log once it
4328 has entered its main loop, and it closes the log before doing so. This is
4329 so that log files can straightforwardly be renamed and moved. However,
4330 there are a couple of unusual error situations where the daemon does write
4331 log entries, and I had neglected to close the log afterwards.
4333 30. The text of an SMTP error response that was received during a remote
4334 delivery was being truncated at 512 bytes. This is too short for some of
4335 the long messages that one sometimes sees. I've increased the limit to
4338 31. It is now possible to make retry rules that apply only when a message has a
4339 specific sender, in particular, an empty sender.
4341 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4342 it possible to be selective about when SMTP synchronization is enforced.
4344 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4346 32. Implemented hosts_connection_nolog.
4348 33. Added an ACL for QUIT.
4350 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4353 35. Added mailbox_size and mailbox_filecount to appendfile.
4355 36. Added control = no_multiline_responses to ACLs.
4357 37. There was a bug in the logic of the code that waits for the clock to tick
4358 in the case where the clock went backwards by a substantial amount such
4359 that the microsecond fraction of "now" was more than the microsecond
4360 fraction of "then" (but the whole seconds number was less).
4362 38. Added support for the libradius Radius client library this is found on
4363 FreeBSD (previously only the radiusclient library was supported).
4369 1. When certain lookups returned multiple values in the form name=value, the
4370 quoting of the values was not always being done properly. Specifically:
4371 (a) If the value started with a double quote, but contained no whitespace,
4373 (b) If the value contained whitespace other than a space character (i.e.
4374 tabs or newlines or carriage returns) it was not quoted.
4375 This fix has been applied to the mysql and pgsql lookups by writing a
4376 separate quoting function and calling it from the lookup code. The fix
4377 should probably also be applied to nisplus, ibase and oracle lookups, but
4378 since I cannot test any of those, I have not disturbed their existing code.
4380 2. A hit in the callout cache for a specific address caused a log line with no
4381 reason for rejecting RCPT. Now it says "Previous (cached) callout
4382 verification failure".
4384 3. There was an off-by-one bug in the queryprogram router. An over-long
4385 return line was truncated at 256 instead of 255 characters, thereby
4386 overflowing its buffer with the terminating zero. As well as fixing this, I
4387 have increased the buffer size to 1024 (and made a note to document this).
4389 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4390 when Exim is waiting for an SMTP response from a remote server, Exim
4391 restarts its select() call on the socket, thereby resetting its timeout.
4392 This is not a problem when such interrupts are rare. Somebody set up a cron
4393 job to run exiwhat every 2 minutes, which is less than the normal select()
4394 timeout (5 or 10 minutes). This meant that the select() timeout never
4395 kicked in because it was always reset. I have fixed this by comparing the
4396 time when an interrupt arrives with the time at the start of the first call
4397 to select(). If more time than the timeout has elapsed, the interrupt is
4398 treated as a timeout.
4400 5. Some internal re-factoring in preparation for the addition of Sieve
4401 extensions (by MH). In particular, the "personal" test is moved to a
4402 separate function, and given an option for scanning Cc: and Bcc: (which is
4403 not set for Exim filters).
4405 6. When Exim created an email address using the login of the caller as the
4406 local part (e.g. when creating a From: or Sender: header line), it was not
4407 quoting the local part when it contained special characters such as @.
4409 7. Installed new OpenBSD configuration files.
4411 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4412 try to make them clearer.
4414 9. Callout options, other than the timeout value, were being ignored when
4415 verifying sender addresses in header lines. For example, when using
4417 verify = header_sender/callout=no_cache
4419 the cache was (incorrectly) being used.
4421 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4422 only the Cygwin environment.
4424 11. When return_path_on_delivery was set as a log selector, if different remote
4425 addresses in the same message used different return paths and parallel
4426 remote delivery occurred, the wrong values would sometimes be logged.
4427 (Whenever a remote delivery process finished, the return path value from
4428 the most recently started remote delivery process was logged.)
4430 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4431 lines when AUTH and/or TLS are in use. This is the "received protocol"
4432 field. Exim used to use "asmtp" for authenticated SMTP, without any
4433 indication (in the protocol name) for TLS use. Now it follows the RFC and
4434 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4435 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4436 names appear in log lines as well as in Received: header lines.
4438 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4439 extensions, and comparison tests, and to fix some bugs.
4441 14. Changes to the "personal" filter test:
4443 (1) The test was buggy in that it was just doing the equivalent of
4444 "contains" tests on header lines. For example, if a user's address was
4445 anne@some.where, the "personal" test would incorrectly be true for
4447 To: susanne@some.where
4449 This test is now done by extracting each address from the header in turn,
4450 and checking the entire address. Other tests that are part of "personal"
4451 are now done using regular expressions (for example, to check local parts
4452 of addresses in From: header lines).
4454 (2) The list of non-personal local parts in From: addresses has been
4455 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4456 taken from the Sieve specification recommendations.
4458 (3) If the message contains any header line starting with "List-" it is
4459 treated as non-personal.
4461 (4) The test for "circular" in the Subject: header line has been removed
4462 because it now seems ill-conceived.
4464 15. Minor typos in src/EDITME comments corrected.
4466 16. Installed latest exipick from John Jetmore.
4468 17. If headers_add on a router specified a text string that was too long for
4469 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4470 of string_sprintf() is now avoided.
4472 18. $message_body_size was not set (it was always zero) when running the DATA
4473 ACL and the local_scan() function.
4475 19. For the "mail" command in an Exim filter, no default was being set for
4476 the once_repeat time, causing a random time value to be used if "once" was
4477 specified. (If the value happened to be <= 0, no repeat happened.) The
4478 default is now 0s, meaning "never repeat". The "vacation" command was OK
4479 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4480 (I found it when inspecting the code).
4482 20. There is now an overall timeout for performing a callout verification. It
4483 defaults to 4 times the callout timeout, which applies to individual SMTP
4484 commands during the callout. The overall timeout applies when there is more
4485 than one host that can be tried. The timeout is checked before trying the
4486 next host. This prevents very long delays if there are a large number of
4487 hosts and all are timing out (e.g. when the network connections are timing
4488 out). The value of the overall timeout can be changed by specifying an
4489 additional sub-option for "callout", called "maxwait". For example:
4491 verify = sender/callout=5s,maxwait=20s
4493 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4494 to the end before writing, but this should make it even safer).
4496 22. Exim was forgetting that it had advertised PIPELINING for the second and
4497 subsequent messages on an SMTP connection. It was also not resetting its
4498 memory on STARTTLS and an internal HELO.
4500 23. When Exim logs an SMTP synchronization error within a session, it now
4501 records whether PIPELINING has been advertised or not.
4503 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4504 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4505 rather than long int.
4507 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4509 26. Added the never_mail option to autoreply.
4515 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4516 crash if the getsockname() call failed; this can happen if a connection is
4517 closed very soon after it is established. The problem was simply in the
4518 order in which certain operations were done, causing Exim to try to write
4519 to the SMTP stream before it had set up the file descriptor. The bug has
4520 been fixed by making things happen in the correct order.
4526 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4527 before the connection was closed, thus losing the rejection response.
4529 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4530 some early Solaris releases, but causes trouble in current releases where
4531 socklen_t is defined.
4533 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4536 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4539 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4540 that releasing the top of it at the end releases what was used for sub-
4541 expansions (unless the block got too big). However, discard this block if
4542 the first thing is a variable or header, so that we can use its block when
4543 it is dynamic (useful for very large $message_headers, for example).
4545 6. Lookups now cache *every* query, not just the most recent. A new, separate
4546 store pool is used for this. It can be recovered when all lookup caches are
4547 flushed. Lookups now release memory at the end of their result strings.
4548 This has involved some general refactoring of the lookup sources.
4550 7. Some code has been added to the store_xxx() functions to reduce the amount
4551 of flapping under certain conditions.
4553 8. log_incoming_interface used to affect only the <= reception log lines. Now
4554 it causes the local interface and port to be added to several more SMTP log
4555 lines, for example "SMTP connection from", and rejection lines.
4557 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4559 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4561 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4562 could be overwritten at the end of the current message (or the start of a
4563 new message if it was set in a HELO ACL). The value is now preserved for
4564 the duration of the SMTP connection.
4566 12. If a transport had a headers_rewrite setting, and a matching header line
4567 contained an unqualified address, that address was qualified, even if it
4568 did not match any rewriting rules. The underlying bug was that the values
4569 of the flags that permit the existence of unqualified sender and recipient
4570 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4571 non-local messages, and by -bnq for local messages) were not being
4572 preserved with the message after it was received.
4574 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4575 "next input=" as part of the text comprising the host identity instead of
4576 the correct text. The code was using the same buffer for two different
4577 strings. However, depending on which order the printing function evaluated
4578 its arguments, the bug did not always show up. Under Linux, for example, my
4579 test suite worked just fine.
4581 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4582 doesn't work with some older versions of Perl. It has been changed to "my",
4583 which in any case is probably the better facility to use.
4585 15. A really picky compiler found some instances of statements for creating
4586 error messages that either had too many or two few arguments for the format
4589 16. The size of the buffer for calls to the DNS resolver has been increased
4590 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4591 for addresses that have a lot of PTR records. This alleviates a problem; it
4592 does not fully solve it.
4594 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4595 buffer now truncates the list and logs the incident, which is the same
4596 action as happens when Exim is looking up a host name and its aliases.
4597 Previously in this situation something unpredictable would happen;
4598 sometimes it was "internal error: store_reset failed".
4600 18. If a server dropped the connection unexpectedly when an Exim client was
4601 using GnuTLS and trying to read a response, the client delivery process
4602 crashed while trying to generate an error log message.
4604 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4605 string, for example:
4607 warn message = H1: something\nH2: something
4609 the text was added as a single header line from Exim's point of view
4610 though it ended up OK in the delivered message. However, searching for the
4611 second and subsequent header lines using $h_h2: did not work. This has been
4612 fixed. Similarly, if a system filter added multiple headers in this way,
4613 the routers could not see them.
4615 20. Expanded the error message when iplsearch is called with an invalid key to
4616 suggest using net-iplsearch in a host list.
4618 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4619 ACLs are no longer actually imposed (and a message to that effect is
4622 22. If a "gecos" field in a passwd entry contained escaped characters, in
4623 particular, if it contained a \" sequence, Exim got it wrong when building
4624 a From: or a Sender: header from that name. A second bug also caused
4625 incorrect handling when an unquoted " was present following a character
4626 that needed quoting.
4628 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4629 was not being matched caselessly.
4631 24. Arranged for all hyphens in the exim.8 source to be escaped with
4634 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4635 properly. Recipient callout cache records were still being keyed to include
4636 the sender, even when use_sender was set false. This led to far more
4637 callouts that were necessary. The sender is no longer included in the key
4638 when use_sender is false.
4640 26. Added "control = submission" modifier to ACLs.
4642 27. Added the ${base62d: operator to decode base 62 numbers.
4644 28. dnsdb lookups can now access SRV records.
4646 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4647 the configuration file.
4649 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4650 (-v) mode. This makes the output for a verbose queue run more intelligible.
4652 31. Added a use_postmaster feature to recipient callouts.
4654 32. Added the $body_zerocount variable, containing the number of binary zero
4655 bytes in the message body.
4657 33. The time of last modification of the "new" subdirectory is now used as the
4658 "mailbox time last read" when there is a quota error for a maildir
4661 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4663 35. Added +ignore_unknown as a special item in host lists.
4665 36. Code for decoding IPv6 addresses in host lists is now included, even if
4666 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4667 address was recognized as an IP address, but was then not correctly decoded
4668 into binary, causing unexpected and incorrect effects when compared with
4675 1. Very minor rewording of debugging text in manualroute to say "list of
4676 hosts" instead of "hostlist".
4678 2. If verify=header_syntax was set, and a header line with an unqualified
4679 address (no domain) and a large number of spaces between the end of the
4680 name and the colon was received, the reception process suffered a buffer
4681 overflow, and (when I tested it) crashed. This was caused by some obsolete
4682 code that should have been removed. The fix is to remove it!
4684 3. When running in the test harness, delay a bit after writing a bounce
4685 message to get a bit more predictability in the log output.
4687 4. Added a call to search_tidyup() just before forking a reception process. In
4688 theory, someone could use a lookup in the expansion of smtp_accept_max_
4689 per_host which, without the tidyup, could leave open a database connection.
4691 5. Added the variables $recipient_data and $sender_data which get set from a
4692 lookup success in an ACL "recipients" or "senders" condition, or a router
4693 "senders" option, similar to $domain_data and $local_part_data.
4695 6. Moved the writing of debug_print from before to after the "senders" test
4698 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4699 problems for message scanning, either using a data ACL, or using
4700 local_scan() because the Received: header was not generated till after they
4701 were called (in order to set the time as the time of reception completion).
4702 I have revised the way this works. The header is now generated after the
4703 body is received, but before the ACL or local_scan() are called. After they
4704 are run, the timestamp in the header is updated.
4710 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4711 before starting a queue runner without re-exec. This happened only when
4712 deliver_drop_privilege was set or when the Exim user was set to root. The
4713 effect of the bug was that timeouts during subsequent deliveries caused
4714 crashes instead of being properly handled. The handler is now left at its
4715 default (and expected) setting.
4717 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4718 message, again when deliver_drop_privilege is set or Exim is run as root.
4719 The bug described in (1) was not present in this case, but the tidying up
4720 of the other signals was missing. I have made the two cases consistent.
4722 3. The ignore_target_hosts setting on a manualroute router was being ignored
4723 for hosts that were looked up using the /MX notation.
4725 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4728 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4729 operated on the sender address. After changing the $sender_address to <>
4730 for the sender address verify, Exim was re-instated it as the original
4731 (before rewriting) address, but remembering that it had rewritten it, so it
4732 wasn't rewriting it again. This bug also had the effect of breaking the
4733 sender address verification caching when the sender address was rewritten.
4735 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4736 This has been changed so that if the ip literal address matches
4737 ignore_target_hosts, the router declines.
4739 7. Added expansion conditions match_domain, match_address, and match_local_
4740 part (NOT match_host).
4742 8. The placeholder for the Received: header didn't have a length field set.
4744 9. Added code to Exim itself and to exim_lock to test for a specific race
4745 condition that could lead to file corruption when using MBX delivery. The
4746 issue is with the lockfile that is created in /tmp. If this file is removed
4747 after a process has opened it but before that process has acquired a lock,
4748 there is the potential for a second process to recreate the file and also
4749 acquire a lock. This could lead to two Exim processes writing to the file
4750 at the same time. The added code performs the same test as UW imapd; it
4751 checks after acquiring the lock that its file descriptor still refers to
4752 the same named file.
4754 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4755 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4758 11. Added the smtp_active_hostname option. If used, this will typically be made
4759 to depend on the incoming interface address. Because $interface_address is
4760 not set up until the daemon has forked a reception process, error responses
4761 that can happen earlier (such as "too many connections") no longer contain
4764 12. If an expansion in a condition on a "warn" statement fails because a lookup
4765 defers, the "warn" statement is abandoned, and the next ACL statement is
4766 processed. Previously this caused the whole ACL to be aborted.
4768 13. Added the iplsearch lookup type.
4770 14. Added ident_timeout as a log selector.
4772 15. Added tls_certificate_verified as a log selector.
4774 16. Added a global option tls_require_ciphers (compare the smtp transport
4775 option of the same name). This controls incoming TLS connections.
4777 17. I finally figured out how to make tls_require_ciphers do a similar thing
4778 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4779 before starting the TLS session.
4781 18. Tabs are now shown as \t in -bP output.
4783 19. If the log selector return_path_on_delivery was set, Exim crashed when
4784 bouncing a message because it had too many Received: header lines.
4786 20. If two routers both had headers_remove settings, and the first one included
4787 a superfluous trailing colon, the final name in the first list and the
4788 first name in the second list were incorrectly joined into one item (with a
4789 colon in the middle).
4795 1. Added -C and -D options to the exinext utility, mainly to make it easier
4796 to include in the automated testing, but these could be helpful when
4797 multiple configurations are in use.
4799 2. The exinext utility was not formatting the output nicely when there was
4800 an alternate port involved in the retry record key, nor when there was a
4801 message id as well (for retries that were specific to a specific message
4802 and a specific host). It was also confused by IPv6 addresses, because of
4803 the additional colons they contain. I have fixed the IPv4 problem, and
4804 patched it up to do a reasonable job for IPv6.
4806 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4807 delivery, the log line now contains "pipelined" if PIPELINING was used.
4809 4. An SMTP transport process used to panic and die if the bind() call to set
4810 an explicit outgoing interface failed. This has been changed; it is now
4811 treated in the same way as a connect() failure.
4813 5. A reference to $sender_host_name in the part of a conditional expansion
4814 that was being skipped was still causing a DNS lookup. This no longer
4817 6. The def: expansion condition was not recognizing references to header lines
4818 that used bh_ and bheader_.
4820 7. Added the _cache feature to named lists.
4822 8. The code for checking quota_filecount in the appendfile transport was
4823 allowing one more file than it should have been.
4825 9. For compatibility with Sendmail, the command line option
4833 and sets the incoming protocol and host name (for trusted callers). The
4834 host name and its colon can be omitted when only the protocol is to be set.
4835 Note the Exim already has two private options, -pd and -ps, that refer to
4836 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4837 "s", but I don't think that's a major issue.
4839 10. A number of refactoring changes to the code, none of which should affect
4842 (a) The number of logging options was getting close to filling up the
4843 32-bit word that was used as a bit map. I have split them into two classes:
4844 those that are passed in the argument to log_write(), and those that are
4845 only ever tested independently outside of that function. These are now in
4846 separate 32-bit words, so there is plenty of room for expansion again.
4847 There is no change in the user interface or the logging behaviour.
4849 (b) When building, for example, log lines, the code previously used a
4850 macro that called string_cat() twice, in order to add two strings. This is
4851 not really sufficiently general. Furthermore, there was one instance where
4852 it was actually wrong because one of the argument was used twice, and in
4853 one call a function was used. (As it happened, calling the function twice
4854 did not affect the overall behaviour.) The macro has been replaced by a
4855 function that can join an arbitrary number of extra strings onto a growing
4858 (c) The code for expansion conditions now uses a table and a binary chop
4859 instead of a serial search (which was left over from when there were very
4860 few conditions). Also, it now recognizes conditions like "pam" even when
4861 the relevant support is not compiled in: a suitably worded error message is
4862 given if an attempt is made to use such a condition.
4864 11. Added ${time_interval:xxxxx}.
4866 12. A bug was causing one of the ddress fields not to be passed back correctly
4867 from remote delivery subprocesses. The field in question was not being
4868 subsequently used, so this caused to problems in practice.
4870 13. Added new log selectors queue_time and deliver_time.
4872 14. Might have fixed a bug in maildirsizefile handling that threw up
4873 "unexpected character" debug warnings, and recalculated the data
4874 unnecessarily. In any case, I expanded the warning message to give more
4877 15. Added the message "Restricted characters in address" to the statements in
4878 the default ACL that block characters like @ and % in local parts.
4880 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4881 Three changes have been made:
4883 (a) There was a serious bug; a negative response to MAIL caused the whole
4884 recipient domain to be cached as invalid, thereby blocking all messages
4885 to all local parts at the same domain, from all senders. This bug has
4886 been fixed. The domain is no longer cached after a negative response to
4887 MAIL if the sender used is not empty.
4889 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4892 (c) A new callout option, "use_sender" has been added for people who want
4893 the modified behaviour.
4899 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4902 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4903 indeed breaks things for older releases.
4905 3. Added additional logging to the case where there is a problem reading data
4906 from a filter that is running in a subprocess using a pipe, in order to
4907 try to track down a specific problem.
4909 4. Testing facility fudge: when running in the test harness and attempting
4910 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4911 getting "No route to host". Convert this to a timeout.
4913 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4916 6. Some OS don't have socklen_t but use size_t instead. This affects the
4917 fifth argument of getsockopt() amongst other things. This is now
4918 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4919 can be set for individual OS. I have set it for SunOS5, OSF1, and
4920 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4921 some earlier ones do not.
4923 7. Change 4.30/15 was not doing the test caselessly.
4925 8. The standard form for an IPv6 address literal was being rejected by address
4926 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4927 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4928 this, as well as the form without the "IPv6" on the front (but only when
4929 address literals are enabled, of course).
4931 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4933 10. Exim crashed if a message with an empty sender address specified by -f
4934 encountered a router with an errors_to setting. This could be provoked only
4935 by a command such as
4939 where an empty string was supplied; "<>" did not hit this bug.
4941 11. Installed PCRE release 4.5.
4943 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4944 remained set. It is now erased.
4946 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4947 times from message ids (which are base 36 rather than the normal 62).
4949 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4950 were being counted as actual protocol errors, and logged if the log
4951 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4952 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4953 and DATA following a set of rejected RCPTs do not count as protocol errors.
4954 In other words, Exim assumes they were pipelined, though this may not
4955 actually be the case. Of course, in all cases the client gets an
4956 appropriate error code.
4958 15. If a lookup fails in an ACL condition, a message about the failure may
4959 be available; it is used if testing the ACL cannot continue, because most
4960 such messages specify what the cause of the deferral is. However, some
4961 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4962 that caused an old message to be retained and used if a later statement
4963 caused a defer, replacing the real cause of the deferral.
4965 16. If an IP address had so many PTR records that the DNS lookup buffer
4966 was not large enough to hold them, Exim could crash while trying to process
4967 the truncated data. It now detects and logs this case.
4969 17. Further to 4.21/58, another change has been made: if (and only if) the
4970 first line of a message (the first header line) ends with CRLF, a bare LF
4971 in a subsequent header line has a space inserted after it, so as not to
4972 terminate the header.
4974 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4975 unnecessarily, used atoi() instead of strtol(), and didn't check the
4976 termination when getting file sizes from file names by regex.
4978 19. Completely re-implemented the support for maildirsize files, in the light
4979 of a number of problems with the previous contributed implementation
4980 (4.30/29). In particular:
4982 . If the quota is zero, the maildirsize file is maintained, but no quota is
4985 . If the maildir directory does not exist, it is created before any attempt
4986 to write a maildirsize file.
4988 . The quota value in the file is just a cache; if the quota is changed in
4989 the transport, the new value overrides.
4991 . A regular expression is available for excluding directories from the
4994 20. The autoreply transport checks the characters in options that define the
4995 message's headers; it allows continued headers, but it was checking with
4996 isspace() after an embedded newline instead of explicitly looking for a
4999 21. If all the "regular" hosts to which an address was routed had passed their
5000 expiry times, and had not reached their retry times, the address was
5001 bounced, even if fallback hosts were defined. Now Exim should go on to try
5004 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5005 equivalent code in the SMTP transport. Some hosts send humungous responses
5006 to HELO/EHLO, more than 1024 it seems.
5008 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5009 gives compiler warnings in some environments. I've now done it "properly",
5012 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5013 (because of problems with the built-in one) was declared to return uschar *
5014 instead of char *, causing compiler failure.
5016 25. Fixed a file descriptor leak when processing alias/forward files.
5018 26. Fixed a minor format string issue in dbfn.c.
5020 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5022 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5023 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5026 29. When a sender address is verified, it is cached, to save repeating the test
5027 when there is more than one recipient in a message. However, when the
5028 verification involves a callout, it is possible for different callout
5029 options to be set for different recipients. It is too complicated to keep
5030 track of this in the cache, so now Exim always runs a verification when a
5031 callout is required, relying on the callout cache for the optimization.
5032 The overhead is duplication of the address routing, but this should not be
5035 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5036 to be verified with callout=postmaster, and the main callout worked but the
5037 postmaster check failed, the verification correctly failed. However, if a
5038 subsequent RCPT command asked for sender verification *without* the
5039 postmaster check, incorrect caching caused this verification also to fail,
5042 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5043 it was not caching the DNS options (qualify_single, search_parents) that
5044 were used when the lookup failed. A subsequent lookup with different
5045 options therefore always gave the same answer, though there were cases
5046 where it should not have. (Example: a "domains = !$mx_any" option on a
5047 dnslookup router: the "domains" option is always processed without any
5048 widening, but the router might have qualify_single set.) Now Exim uses the
5049 cached value only when the same options are set.
5051 32. Added John Jetmore's "exipick" utility to the distribution.
5053 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5054 than a timeout (e.g. a certificate is required, and is not provided), an
5055 Exim server now closes the connection immediately. Previously it waited for
5056 the client to close - but if the client is SSL, it seems that they each
5057 wait for each other, leading to a delay before one of them times out.
5059 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5060 maintained 0.8.x compatibility because I don't think many are using it, and
5061 it is clearly obsolete.
5063 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5066 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5067 client certificate was expired. A simple patch fixes this, though I don't
5068 understand the full logic of why the verify callback is called multiple
5071 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5072 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5073 which causes problems with some clients (such as the Certicom SSL Plus
5074 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5075 disables the coutermeasure allowing Eudora to connect."
5077 38. Exim was not checking that a write() to a log file succeeded. This could
5078 lead to Bad Things if a log got too big, in particular if it hit a file
5079 size limit. Exim now panics and dies if it cannot write to a log file, just
5080 as it does if it cannot open a log file.
5082 39. Modified OS/Makefile-Linux so that it now contains
5084 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5086 The two -D definitions ensure that Exim is compiled with large file
5087 support, which makes it possible to handle log files that are bigger than
5090 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5091 instance) a domain was checked against a named list that involved a lookup,
5092 causing $domain_data to be set, then another domain was checked against the
5093 same list, then the first domain was re-checked, the value of $domain_data
5094 after the final check could be wrong. In particular, if the second check
5095 failed, it could be set empty. This bug probably also applied to
5098 41. The strip_trailing_dot option was not being applied to the address given
5099 with the -f command-line option.
5101 42. The code for reading a message's header from the spool was incrementing
5102 $received_count, but never initializing it. This meant that the value was
5103 incorrect (doubled) while delivering a message in the same process in which
5104 it was received. In the most common configuration of Exim, this never
5105 happens - a fresh exec is done - but it can happen when
5106 deliver_drop_privilege is set.
5108 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5109 it now includes up to 150 characters of the unexpected data in the log
5112 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5113 and building data strings. The size of both of these buffers was 10 000
5114 bytes - far larger than anybody would *ever* want, thought I. Needless to
5115 say, somebody hit the limit. I have increased the maximum line length to
5116 20 000 and the maximum data length of concatenated lines to 100 000. I have
5117 also fixed two bugs, because there was no checking on these buffers. Tsk,
5118 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5119 buffer is too small.
5121 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5122 lsearch lookups. Now it does.
5124 46. When parsing a route_list item in a manualroute router, a fixed-length
5125 buffer was used for the list of hosts. I made this 1024 bytes long,
5126 thinking that nobody would ever have a list of hosts that long. Wrong.
5127 Somebody had a whole pile of complicated expansion conditions, and the
5128 string was silently truncated, leading to an expansion error. It turns out
5129 that it is easier to change to an unlimited length (owing to other changes
5130 that have happened since this code was originally written) than to build
5131 structure for giving a limitation error. The length of the item that
5132 expands into the list of hosts is now unlimited.
5134 47. The lsearch lookup could not handle data where the length of text line was
5135 more than 4095 characters. Such lines were truncated, leading to shortened
5136 data being returned. It should now handle lines of any length.
5138 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5139 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5142 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5143 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5144 no longer happens in comment lines. A list of replacements is now placed
5145 at the head of all of the source files, except those whose only change is
5146 to replace PERL_COMMAND in the very first #! line.
5148 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5149 messages on the queue, with a bottom-up merge sort, using code contributed
5150 by Michael Haardt. This should make operations like -bp somewhat faster on
5151 large queues. It won't affect queue runners, except when queue_run_in_order
5154 51. Installed eximstats 1.31 in the distribution.
5156 52. Added support for SRV lookups to the dnslookup router.
5158 53. If an ACL referred to $message_body or $message_body_end, the value was not
5159 reset for any messages that followed in the same SMTP session.
5161 54. The store-handling optimization for building very long strings was not
5162 differentiating between the different store pools. I don't think this
5163 actually made any difference in practice, but I've tidied it.
5165 55. While running the routers to verify a sender address, $sender_address
5166 was still set to the sender address. This is wrong, because when routing to
5167 send a bounce to the sender, it would be empty. Therefore, I have changed
5168 it so that, while verifying a sender address, $sender_address is set to <>.
5169 (There is no change to what happens when verifying a recipient address.)
5171 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5172 target A or AAAA records (if not already returned) without resetting the
5173 qualify_single or search_parents options of the DNS resolver. These are
5174 inappropriate in this case because the targets of MX and SRV records must
5175 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5176 target that, when qualified, matched something in the local domain. These
5177 two options are now turned off when doing these lookups.
5179 57. It seems that at least some releases of Reiserfs (which does not have the
5180 concept of a fixed number of inodes) returns zero and not -1 for the
5181 number of available inodes. This interacted badly with check_spool_inodes,
5182 which assumed that -1 was the "no such thing" setting. What I have done is
5183 to check that the total number of inodes is greater than zero before doing
5184 the test of how many are available.
5186 58. When a "warn" ACL statement has a log_message modifier, the message is
5187 remembered, and not repeated. This is to avoid a lot of repetition when a
5188 message has many recipients that cause the same warning to be written.
5189 Howewer, Exim was preserving the list of already written lines for an
5190 entire SMTP session, which doesn't seem right. The memory is now reset if a
5191 new message is started.
5193 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5194 debugging output unless log_rewrite was also set.
5196 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5197 of (int)(handle) when we know that handle contains (void *)(-1).
5199 61. The Exim daemon panic-logs an error return when it closes the incoming
5200 connection. However "connection reset by peer" seems to be common, and
5201 isn't really an error worthy of noting specially, so that particular error
5204 62. When Exim is trying to find all the local interfaces, it used to panic and
5205 die if the ioctl to get the interface flags failed. However, it seems that
5206 on at least one OS (Solaris 9) it is possible to have an interface that is
5207 included in the list of interfaces, but for which you get a failure error
5208 for this call. This happens when the interface is not "plumbed" into a
5209 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5210 failure of the "get flags" call assumes that the interface is down.
5212 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5213 makes life easier for people who are doing arithmetic on fields extracted
5214 from dates, where you often get leading zeros that should not be
5215 interpreted as octal.
5217 64. Added qualify_domain to the redirect router, to override the global
5220 65. If a pathologically long header line contained very many addresses (the
5221 report of this problem mentioned 10 000) and each of them was rewritten,
5222 Exim could use up a very large amount of memory. (It kept on making new
5223 copies of the header line as it rewrote, and never released the old ones.)
5224 At the expense of a bit more processing, the header rewriting function has
5225 been changed so that it no longer eats memory in this way.
5227 66. The generation of the Received: header has been moved from the time that a
5228 message starts to be received, to the time that it finishes. The timestamp
5229 in the Received: header should now be very close to that of the <= log
5230 line. There are two side-effects of this change:
5232 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5233 logged header lines no longer include the local Received: line, because
5234 it has not yet been created. The same applies to a copy of the message
5235 that is returned to a non-SMTP sender when a message is rejected.
5237 (b) When a filter file is tested using -bf, no additional Received: header
5238 is added to the test message. After some thought, I decided that this
5241 This change does not affect the value of $received_for. It is still set
5242 after address rewriting, but before local_scan() is called.
5244 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5246 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5247 gave an unhelpful panic error message, and a defer error. I have managed to
5248 change this behaviour so that it now rejects any supplied certificate,
5249 which seems right, as the list of acceptable certificates is empty.
5251 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5252 gave an unhelpful defer error. I have not managed to make this reject any
5253 supplied certificates, but the error message it gives is "no certificate
5254 supplied", which is not helpful.
5256 70. exigrep's output now also includes lines that are not associated with any
5257 message, but which match the given pattern. Implemented by a patch from
5258 Martin Sluka, which also tidied up the Perl a bit.
5260 71. Recipient callout verification, like sender verification, was using <> in
5261 the MAIL FROM command. This isn't really the right thing, since the actual
5262 sender may affect whether the remote host accepts the recipient or not. I
5263 have changed it to use the actual sender in the callout; this means that
5264 the cache record is now keyed on a recipient/sender pair, not just the
5265 recipient address. There doesn't seem to be a real danger of callout loops,
5266 since a callout by the remote host to check the sender would use <>.
5267 [SEE ABOVE: changed after hitting problems.]
5269 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5270 temporary errors. However, in the case of such a code being given after
5271 the end of a data transmission (i.e. after ".") Exim was failing to write
5272 a retry record for the message. (Yes, there was some broken host that was
5273 actually sending 8xx at this point.)
5275 73. An unknown lookup type in a host list could cause Exim to panic-die when
5276 the list was checked. (An example that provoked this was putting <; in the
5277 middle of a list instead of at the start.) If this happened during a DATA
5278 ACL check, a -D file could be left lying around. This kind of configuration
5279 error no longer causes Exim to die; instead it causes a defer errror. The
5280 incident is still logged to the main and panic logs.
5282 74. Buglet left over from Exim 3 conversion. The message "too many messages
5283 in one connection" was written to the rejectlog but not the mainlog, except
5284 when address rewriting (yes!) was being logged.
5286 75. Added write_rejectlog option.
5288 76. When a system filter was run not as root (that is, when system_filter_user
5289 was set), the values of the $n variables were not being returned to the
5290 main process; thus, they were not subsequently available in the $sn
5293 77. Added +return_path_on_delivery log selector.
5295 78. A connection timeout was being treated differently from recipients deferred
5296 when testing hosts_max_try with a message that was older than the host's
5297 retry timeout. (The host should not be counted, thus allowing all hosts to
5298 be tried at least once before bouncing.) This may have been the cause of an
5299 occasionally reported bug whereby a message would remain on the queue
5300 longer than the retry timeout, but would be bounced if a delivery was
5301 forced. I say "may" because I never totally pinned down the problem;
5302 setting up timeout/retry tests is difficult. See also the next item.
5304 79. The ultimate address timeout was not being applied to errors that involved
5305 a combination of host plus message (for example, a timeout on a MAIL
5306 command). When an address resolved to a number of possible hosts, and they
5307 were not all tried for each delivery (e.g. because of hosts_max_try), a
5308 message could remain on the queue longer than the retry timeout.
5310 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5313 81. Fixed an obscure SMTP outgoing bug which required at least the following
5314 conditions: (a) there was another message waiting for the same server;
5315 (b) the server returned 5xx to all RCPT commands in the first message so
5316 that the message was not completed; (c) the server dropped the connection
5317 or gave a negative response to the RSET that Exim sends to abort the
5318 transaction. The observed case was a dropped connection after DATA that had
5319 been sent in pipelining mode. That is, the server had advertised PIPELINING
5320 but was not implementing it correctly. The effect of the bug was incorrect
5321 behaviour, such as trying another host, and this could lead to a crash.
5327 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5328 and daemon.c were passed as pointers to ints; they should have been
5329 pointers to socklen_t variables (which are typically unsigned ints).
5331 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5334 3. Fixed a really odd bug that affected only the testing scheme; patching a
5335 certain fixed string in the binary changed the value of another string that
5336 happened to be identical to the end of the original first string.
5338 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5339 name", it returns that address as the IP address. On some operating
5340 systems (e.g. Solaris), it also passes back the IP address string as the
5341 "host name". However, on others (e.g. Linux), it passes back an empty
5342 string. Exim wasn't checking for this, and was changing the host name to an
5343 empty string, assuming it had been canonicized.
5345 5. Although rare, it is permitted to have more than one PTR record for a given
5346 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5347 all the names associated with an address, because they do in Solaris.
5348 However, it seems that they do not in Linux for data that comes from the
5349 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5350 I found this out when I moved to a new Linux workstation and tried to run
5351 the Exim test suite.
5353 To get round this problem I have changed the code so that it now does its
5354 own call to the DNS to look up PTR records when searching for a host name.
5355 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5356 addresses that are only in /etc/hosts are still found.
5358 This behaviour is, however, controlled by an option called host_lookup_
5359 order, which defaults to "bydns:byaddr". If people want to use the other
5360 order, or indeed, just use one or the other means of lookup, they can
5361 specify it in this variable.
5363 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5364 some operating systems, this comes back from gethostbyaddr() as an empty
5365 string, and this is what Exim used to test for. However, it seems that in
5366 other systems, "." is yielded. Exim now tests for this case too.
5368 7. The values of check_spool_space and check_log_space are now held internally
5369 as a number of kilobytes instead of an absolute number of bytes. If a
5370 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5371 kilobyte. This means that much larger values can be stored.
5373 8. Exim monitor: an attempt to get the action menu when not actually pointing
5374 at a message produces an empty menu entitled "No message selected". This
5375 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5376 no entries in it ("Shell widget menu has zero width and/or height"). So I
5377 have added a single, blank menu entry in this case.
5379 9. Added ${quote_local_part.
5381 10. MIME decoding is now applied to the contents of Subject: header lines when
5384 11. Now that a reference to $sender_host_address automatically causes a reverse
5385 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5386 host lookup before query-style lookups in lists that might use this
5387 variable. This has therefore been abolished, and the "net-" prefix is no
5388 longer necessary for query-style lookups.
5390 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5391 have been a typo for LFLAGS, so it has been changed.
5393 13. The install script calls Exim with "-C /dev/null" in order to find the
5394 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5395 to be output. Howeve, since Exim outputs its version number before the
5396 error, it didn't break the script. It just looked ugly. I fixed this by
5397 always allowing "-C /dev/null" if the caller is root.
5399 14. Ignore overlarge ACL variable number when reading spool file - insurance
5400 against a later release with more variables having written the file.
5402 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5403 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5404 this, as well as the form without the "IPv6" on the front.
5406 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5407 OS/Makefile-Darwin file.
5409 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5410 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5412 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5413 variable. However, some OS use ints for inodes. I've added cast to long int
5414 to get rid of the compiler warning.
5416 19. I had forgotten to lock out "/../" in configuration file names when
5417 ALT_CONFIG_PREFIX was set.
5419 20. Routers used for verification do not need to specify transports. However,
5420 if such a router generated a host list, and callout was configured, Exim
5421 crashed, because it could not find a port number from the (non-existent)
5422 transport. It now assumes port 25 in this circumstance.
5424 21. Added the -t option to exigrep.
5426 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5427 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5428 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5429 them is set, LOOKUP_LSEARCH is forced.
5431 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5433 24. Added sender and host information to the "rejected by local_scan()" log
5434 line; previously there was no indication of these.
5436 25. Added .include_if_exists.
5438 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5439 Linux. It turns out that not all file systems support this. Apparently some
5440 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5441 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5442 not supported on the file descriptor, is now ignored when Exim is trying to
5443 sync a directory. This applies only to Linux.
5445 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5447 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5448 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5451 29. Added support for maildirsize files from supplied patch (modified a bit).
5453 30. The use of :fail: followed by an empty string could lead Exim to respond to
5454 sender verification failures with (e.g.):
5456 550 Verification failed for <xxx>
5457 550 Sender verify failed
5459 where the first response line was missing the '-' that indicates it is not
5460 the final line of the response.
5462 31. The loop for finding the name of the user that called Exim had a hardwired
5463 limit of 10; it now uses the value of finduser_retries, which is used for
5464 all other user lookups.
5466 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5469 33. Exim was neglecting to zero errno before one call of strtol() when
5470 expanding a string and expecting an integer value. On some systems this
5471 resulted in spurious "integer overflow" errors. Also, it was casting the
5472 result into an int without checking.
5474 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5475 did not work. The code looks as if it has *never* worked, though it appears
5476 to have been documented since at least releast 1.62. I have made it work.
5478 35. The "timeout_DNS" error in retry rules, also documented since at least
5479 1.62, also never worked. As it isn't clear exactly what this means, and
5480 clearly it isn't a major issue, I have abolished the feature by treating it
5481 as "timeout", and writing a warning to the main and panic logs.
5483 36. The display of retry rules for -brt wasn't always showing the error code
5486 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5487 timeout_connect_A, timeout_connect_MX.
5489 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5490 to the empty sender.
5492 39. The daemon was not analysing the content of -oX till after it had closed
5493 stderr and disconnected from the controlling terminal. This meant that any
5494 syntax errors were only noted on the panic log, and the return code from
5495 the command was 0. By re-arranging the code a little, I've made the
5496 decoding happen first, so such errors now appear on stderr, and the return
5497 code is 1. However, the actual setting up of the sockets still happens in
5498 the disconnected process, so errors there are still only recorded on the
5501 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5502 connections (as happens on some IP stacks) was logged at start up time as
5503 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5504 it from "IPv6 and IPv4", which means that two separate sockets are being
5507 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5508 says whether AF_INET or AF_INET6 was passed as an argument.
5510 42. Exiwhat output was messed up when time zones were included in log
5513 43. Exiwhat now gives more information about the daemon's listening ports,
5514 and whether -tls-on-connect was used.
5516 44. The "port" option of the smtp transport is now expanded.
5518 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5519 silently ignored. Now an error message is written to the main and panic
5522 46. There's a new ACL modifier called "logwrite" which writes to a log file
5523 as soon as it is encountered.
5525 47. Added $local_user_uid and $local_user_gid at routing time.
5527 48. Exim crashed when trying to verify a sender address that was being
5530 49. Exim was recognizing only a space character after ".include". It now also
5531 recognizes a tab character.
5533 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5534 extracting the relevant information from the specification. The man page no
5535 longer contains scrambled data for the -d option, and I've added a section
5536 at the front about calling Exim under different names.
5538 51. Added "extra_headers" argument to the "mail" command in filter files.
5540 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5543 53. Installed eximstats 1.29.
5545 54. Added transport_filter_timeout as a generic transport option.
5547 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5548 Cc: header lines. This was required by RFC 822, but it not required by RFC
5551 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5552 incoming messages that did not have them. Now it does so only if the
5553 message originates locally, that is, if there is no associated remote host
5554 address. When Resent- header lines are present, this applies to the Resent-
5555 lines rather than the non-Resent- lines.
5557 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5558 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5560 58. Messages for configuration errors now include the name of the main
5561 configuration file - useful now that there may be more than one file in a
5562 list (.included file names were always shown).
5564 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5565 for those rare installations that do not start the daemon as root or run it
5566 setuid root. I've cut out the call to initgroups() if the daemon is not
5569 60. The Exim user and group can now be bound into the binary as text strings
5570 that are looked up at the start of Exim's processing.
5572 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5574 62. Added $mailstore_basename variable.
5576 63. Installed patch to sieve.c from Michael Haardt.
5578 64. When Exim failed to open the panic log after failing to open the main log,
5579 the original message it was trying to log was written to stderr and debug
5580 output, but if they were not available (the usual case in production), it
5581 was lost. Now it is written to syslog before the two lines that record the
5582 failures to open the logs.
5584 65. Users' Exim filters run in subprocesses under the user's uid. It is
5585 possible for a "deliver" command or an alias in a "personal" command to
5586 provoke an address rewrite. If logging of address rewriting is configured,
5587 this fails because the process is not running as root or exim. There may be
5588 a better way of dealing with this, but for the moment (because 4.30 needs
5589 to be released), I have disabled address rewrite logging when running a
5590 filter in a non-root, non-exim process.
5596 1. The buildconfig auxiliary program wasn't quoting the value set for
5597 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5598 not defined. This bug was masked in 4.22 by the effect that was fixed in
5601 2. Some messages that were rejected after a message id was allocated were
5602 shown as "incomplete" by exigrep. It no longer does this for messages that
5603 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5605 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5606 have allow_domain_literals set, the ID did not get logged in the <= line.
5607 Domain literals are now always recognized in Message-ID: header lines.
5609 4. The first argument for a ${extract expansion item is the key name or field
5610 number. Leading and trailing spaces in this item were not being ignored,
5611 causing some misleading effects.
5613 5. When deliver_drop_privilege was set, single queue runner processes started
5614 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5615 same command in the process it spins off) were not dropping privilege.
5617 6. When the daemon running as "exim" started a queue runner, it always
5618 re-executed Exim in the spun-off process. This is a waste of effort when
5619 deliver_drop_privilege is set. The new process now just calls the
5620 queue-runner function directly.
5626 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5629 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5630 ignored. Though the use of -oP was forcing the writing of a pid file, it
5631 was always written to the default place.
5633 3. If the message "no IP address found for host xxxx" is generated during
5634 incoming verification, it is now followed by identification of the incoming
5635 connection (so you can more easily find what provoked it).
5637 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5639 5. Added some features to "harden" Exim a bit more against certain attacks:
5641 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5642 be put in Local/Makefile. This is like the never_users runtime option,
5643 but it cannot be overridden. The default setting is "root".
5645 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5646 prefix string with which any file named in a -C command line option
5649 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5650 is retained for -C and -D only if the caller of Exim is root. Without
5651 it, the exim user may also use -C and -D and retain privilege.
5653 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5654 command line option is disabled.
5656 6. Macro names set by the -D option must start with an upper case letter, just
5657 like macro names defined in the configuration file.
5659 7. Added "dereference=" facility to LDAP.
5661 8. Two instances of the typo "uknown" in the source files are fixed.
5663 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5664 the Configure-Makefile script screwed up while processing it.
5666 10. Incorporated PCRE 4.4.
5668 11. The SMTP synchronization check was not operating right at the start of an
5669 SMTP session. For example, it could not catch a HELO sent before the client
5670 waited for the greeting. There is now a check for outstanding input at the
5671 point when the greeting is written. Because of the duplex, asynchronous
5672 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5673 way, but not yet received, when the check is performed.
5675 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5676 on TCP/IP sockets, because this apparently causes some broken clients to
5679 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5680 unchanged) from the Cygwin maintainer.
5682 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5683 was supported instead of testing for mbx. Effectively a typo.
5685 15. The spa authenticator server code was not checking that the input it
5686 received was valid base64.
5688 16. The debug output line for the "set" modifier in ACLs was not showing the
5689 name of the variable that was being set.
5691 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5693 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5694 lookup on its own. Something else was needed to trigger the lookup. For
5695 example, a match in host_lookup or the need for a host name in a host list.
5696 Now, if $sender_host_name is referenced and the host name has not yet been
5697 looked up, a lookup is performed. If the lookup fails, the variable remains
5698 empty, and $host_lookup_failed is set to "1".
5700 19. Added "eqi" as a case-independent comparison operator.
5702 20. The saslauthd authentication condition could segfault if neither service
5703 nor realm was specified.
5705 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5706 error message that was logged was misleading, and incoming SMTP
5707 connections were dropped. The message is now more accurate, and temporary
5708 errors are given to SMTP connections.
5710 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5711 (except RSET) with a 421 error, until QUIT is received. However, it was
5712 failing to send a response to QUIT.
5714 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5715 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5716 ACL is now run after the helo_try_verify_hosts code.
5718 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5719 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5720 case-independent, so other case variants are also recognized). Apparently
5721 some systems use these upper case variants.
5723 25. If more than two messages were waiting for the same host, and a transport
5724 filter was specified for the transport, Exim sent two messages over the
5725 same TCP/IP connection, and then failed with "socket operation on non-
5726 socket" when it tried to send the third.
5728 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5730 27. The extern definition of crypt16() in expand.c was not being excluded when
5731 the OS had its own crypt16() function.
5733 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5734 as a preferred synonym for return_size_limit, both as an option and as an
5737 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5739 30. Changed the default configuration ACL to relax the local part checking rule
5740 for addresses that are not in any local domains. For these addresses,
5741 slashes and pipe symbols are allowed within local parts, but the sequence
5742 /../ is explicitly forbidden.
5744 31. SPA server authentication was not clearing the challenge buffer before
5747 32. log_message in a "warn" ACL statement was writing to the reject log as
5748 well as to the main log, which contradicts the documentation and doesn't
5749 seem right (because no rejection is happening). So I have stopped it.
5751 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5752 However, I am unable to do any testing of this.
5754 34. Fixed an infelicity in the appendfile transport. When checking directories
5755 for a mailbox, to see if any needed to be created, it was accidentally
5756 using path names with one or more superfluous leading slashes; tracing
5757 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5759 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5760 contents are added to the log line that is written for every discarded
5761 recipient. (Previously a log_message setting was ignored.)
5763 36. The ${quote: operator now quotes the string if it is empty.
5765 37. The install script runs exim in order to find its version number. If for
5766 some reason other than non-existence or emptiness, which it checks, it
5767 could not run './exim', it was installing it with an empty version number,
5768 i.e. as "exim-". This error state is now caught, and the installation is
5771 38. An argument was missing from the function that creates an error message
5772 when Exim fails to connect to the socket for saslauthd authentication.
5773 This could cause Exim to crash, or give a corrupted message.
5775 39. Added isip, isip4, and isip6 to ${if conditions.
5777 40. The ACL variables $acl_xx are now saved with the message, and can be
5778 accessed later in routers, transports, and filters.
5780 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5781 strings in the file are not string-expanded.
5783 42. If a MAIL command specified a SIZE value that was too large to fit into an
5784 int variable, the check against message_size_limit failed. Such values are
5785 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5786 day this will have to be increased, but I don't think I want to be around
5787 when emails are that large.
5794 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5795 iconv() is not standard in FreeBSD.
5797 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5798 IPv6 enabled. The observed symptom was a segmentation fault on return from
5799 the function os_common_find_running_interfaces() in src/os.c.
5801 3. In the check_special_case() function in daemon.c I had used "errno" as an
5802 argument name, which causes warnings on some systems. This was basically a
5803 typo, since it was named "eno" in the comments!
5805 4. The code that waits for the clock to tick (at a resolution of some fraction
5806 of a second) so as to ensure message-id uniqueness was always waiting for
5807 at least one whole tick, when it could have waited for less. [This is
5808 almost certainly not relevant at current processor speeds, where it is
5809 unlikely to ever wait at all. But we try to future-proof.]
5811 5. The function that sleeps for a time interval that includes fractions of a
5812 second contained a race. It did not block SIGALRM between setting the
5813 timer, and suspending (a couple of lines later). If the interval was short
5814 and the sigsuspend() was delayed until after it had expired, the suspension
5815 never ended. On busy systems this could lead to processes getting stuck for
5818 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5819 process, before it forks any delivery processes. The open lookup caching
5820 mechanism meant that the open file or database connection was passed into
5821 the delivery process. The problem was that delivery processes always tidy
5822 up cached lookup data. This could cause a problem for the next delivery
5823 process started by the queue runner, because the external queue runner
5824 process does not know about the closure. So the next delivery process
5825 still has data in the lookup cache. In the case of a file lookup, there was
5826 no problem because closing a file descriptor in a subprocess doesn't affect
5827 the parent. However, if the lookup was caching a connection to a database,
5828 the connection was closed, and the second delivery process was likely to
5829 see errors such as "PGSQL: query failed: server closed the connection
5830 unexpectedly". The problem has been fixed by closing all cached lookups
5831 in a queue runner before running a delivery process.
5833 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5834 seem to have the "const" qualifier which it has on other OS. I've
5837 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5838 *of the same type* (client or server) with the same public name that an
5839 error should be diagnosed.
5841 9. When Exim looked up a host name for an IP address, but failed to find the
5842 original IP address when looking up the host name (a safety check), it
5843 output the message "<ip address> does not match any IP for NULL", which was
5844 confusing, to say the least. The bug was that the host name should have
5845 appeared instead of "NULL".
5847 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5848 user that is built into the binary, and the -C or -D options is used, root
5849 privilege is dropped before the configuration file is read. In addition,
5850 logging is switched to stderr instead of the normal log files. If the
5851 configuration then re-defines the Exim user, the unprivileged environment
5852 is probably not what is expected, so Exim logs a panic warning message (but
5855 However, if deliver_drop_privilege is set, the unprivileged state may well
5856 be exactly what is intended, so the warning has been cut out in that case,
5857 and Exim is allowed to try to write to its normal log files.
5863 1. smtp_return_error_details was not giving details for temporary sender
5864 or receiver verification errors.
5866 2. Diagnose a configuration error if two authenticators have the same public
5869 3. Exim used not to create the message log file for a message until the first
5870 delivery attempt. This could be confusing when incoming messages were held
5871 for policy or load reasons. The message log file is now created at the time
5872 the message is received, and an initial "Received" line is written to it.
5874 4. The automatically generated man page for command line options had a minor
5875 bug that caused no ill effects; however, a more serious problem was that
5876 the procedure for building the man page automatically didn't always
5877 operate. Consequently, release 4.20 contains an out-of-date version. This
5878 shouldn't happen again.
5880 5. When building Exim with embedded Perl support, the script that builds the
5881 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5882 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5884 6. The freeze_tell option was not being used for messages that were frozen on
5885 arrival, either by an ACL or by local_scan().
5887 7. Added the smtp_incomplete_transaction log selector.
5889 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5890 was accepting AUTH without a new EHLO.
5892 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5893 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5896 10. Logging of TCP/IP connections (when configured) now happens in the main
5897 daemon process instead of the child process, so that the TCP/IP connection
5898 count is more accurate (but it can never be perfect).
5900 11. The use of "drop" in a nested ACL was not being handled correctly in the
5901 outer ACL. Now, if condition failure induced by the nested "drop" causes
5902 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5903 or "require"), the connection is dropped.
5905 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5906 that yield "discard" can now be used with an "accept" or a "discard" verb,
5907 but an error is generated for any others (because I can't see a useful way
5908 to define what should happen).
5910 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5911 are now processed in the same way as lines in the Exim configuration file.
5912 In particular, continuation lines are supported.
5914 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5916 15. Added -ti meaning -t -i.
5918 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5919 domains, and warn by logging if others are found.
5921 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5922 returned from GIFCONF when Exim is trying to find the list of interfaces on
5923 a host. The code in os.c has been modified to copy each ifreq to an aligned
5924 structure in all cases.
5926 Also, in some cases, the returned ifreq's were being copied to a 'struct
5927 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5928 means the last couple of bytes of an IPv6 address could be chopped if the
5929 ifreq contained only a normal sockaddr (14 bytes storage).
5931 18. Named domain lists were not supported in the hosts_treat_as_local option.
5932 An entry such as +xxxx was not recognized, and was treated as a literal
5935 19. Ensure that header lines added by a DATA ACL are included in the reject log
5936 if the ACL subsequently rejects the message.
5938 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5939 MD5 (which is deprecated).
5941 21. When testing a filter file using -bf, Exim was writing a message when it
5942 took the sender from a "From " line in the message, but it was not doing so
5943 when it took $return_path from a Return-Path: header line. It now does.
5945 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5946 with a valid header line field name (a series of printing characters
5947 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5949 23. Changed "disc" in the source to "disk" to conform to the documentation and
5950 the book and for uniformity.
5952 24. Ignore Sendmail's -Ooption=value command line item.
5954 25. When execve() failed while trying to run a command in a pipe transport,
5955 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5956 could be confused with a return value of 69 from the command itself. This
5957 has been changed to 127, the value the shell returns if it is asked to run
5958 a non-existent command. The wording for the related log line suggests a
5959 non-existent command as the problem.
5961 26. If received_header_text expands to an empty string, do not add a Received:
5962 header line to the message. (Well, it adds a token one on the spool, but
5963 marks it "old" so that it doesn't get used or transmitted.)
5965 27. Installed eximstats 1.28 (addition of -nt option).
5967 28. There was no check for failure on the call to getsockname() in the daemon
5968 code. This can fail if there is a shortage of resources on the system, with
5969 ENOMEM, for example. A temporary error is now given on failure.
5971 29. Contrary to the C standard, it seems that in some environments, the
5972 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5973 program. Exim now does this explicitly; it affects the formatting of
5974 timestamps using strftime().
5976 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5977 complaints. I've now initialized all the variables, to avoid this.
5979 32. Header lines added by a system filter were not being "seen" during
5980 transport-time rewrites.
5982 33. The info_callback() function passed to OpenSSL is set up with type void
5983 (*)(SSL *, int, int), as described somewhere. However, when calling the
5984 function (actually a macro) that sets it up, the type void(*)() is
5985 expected. I've put in a cast to prevent warnings from picky compilers.
5987 34. If a DNS black list lookup found a CNAME record, but there were no A
5988 records associated with the domain it pointed at, Exim crashed.
5990 35. If a DNS black list lookup returned more than one A record, Exim ignored
5991 all but the first. It now scans all returned addresses if a particular IP
5992 value is being sought. In this situation, the contents of the
5993 $dnslist_value variable are a list of all the addresses, separated by a
5996 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5997 to accept a host name and all its aliases if the forward lookup for any of
5998 them yielded the IP address of the incoming connection. Now it accepts only
5999 those names whose forward lookup yields the correct IP address. Any other
6000 names are discarded. This closes a loophole whereby a rogue DNS
6001 administrator could create reverse DNS records to break through a
6002 wildcarded host restriction in an ACL.
6004 37. If a user filter or a system filter that ran in a subprocess used any of
6005 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6006 the wrong values were passed to the pipe command ($thisaddress had the
6007 value of $0, $0 had the value of $1, etc). This bug was introduced by
6008 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6010 38. Improved the line breaking for long SMTP error messages from ACLs.
6011 Previously, if there was no break point between 40 and 75 characters, Exim
6012 left the rest of the message alone. Two changes have been made: (a) I've
6013 reduced the minimum length to 35 characters; (b) if it can't find a break
6014 point between 35 and 75 characters, it looks ahead and uses the first one
6015 that it finds. This may give the occasional overlong line, but at least the
6016 remaining text gets split now.
6018 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6019 file descriptors might be low, and that setting 1000 would always raise it.
6020 It turns out that in some environments, the limit is already over 1000 and
6021 that lowering it causes trouble. So now Exim takes care not to decrease it.
6023 40. When delivering a message, the value of $return_path is set to $sender_
6024 address at the start of routing (routers may change the value). By an
6025 oversight, this default was not being set up when an address was tested by
6026 -bt or -bv, which affected the outcome if any router or filter referred to
6029 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6030 log only when "message" or "log_message" are set. However, if one of the
6031 conditions was an address verification, or a call to a nested ACL, the
6032 messages generated by the underlying test were being passed through. This
6033 no longer happens. The underlying message is available in $acl_verify_
6034 message for both "message" and "log_message" expansions, so it can be
6035 passed through if needed.
6037 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6038 new expansion $bh_ to give the encoded byte string without charset
6039 translation. Translation happens only if iconv() is available; HAVE_ICONV
6040 indicates this at build time. HEADERS_CHARSET gives the charset to
6041 translate to; headers_charset can change it in the configuration, and
6042 "headers charset" can change it in an individual filter file.
6044 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6045 that creates RFC 2047 encoded "words" labels them as that charset instead
6046 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6047 expansion operator; (ii) when Exim creates a From: line for a local
6048 message; (iii) when a header line is rewritten to include a "phrase" part.
6050 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6051 buggy, causing it to skip the first lines of messages whose message ID
6052 ended in 'D'. This would not have bitten before Exim release 4.14, because
6053 message IDs were unlikely to end in 'D' before then. The effect was to have
6054 incorrect size information for certain domains.
6056 45. #include "config.h" was missing at the start of the crypt16.c module. This
6057 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6060 46. If there was a timeout during a "random" callout check, Exim treated it as
6061 a failure of the random address, and carried on sending RSET and the real
6062 address. If the delay was just some slowness somewhere, the response to the
6063 original RCPT would be taken as a response to RSET and so on, causing
6064 mayhem of various kinds.
6066 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6067 when I implemented it. It didn't allow for the fact that some option values
6068 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6069 the right test for positive values.
6071 48. Domain names in DNS records are case-independent. Exim always looks them up
6072 in lower case. Some resolvers return domain names in exactly the case they
6073 appear in the zone file, that is, they may contain uppercase letters. Not
6074 all resolvers do this - some return always lower case. Exim was treating a
6075 change of case by a resolver as a change of domain, similar to a widening
6076 of a domain abbreviation. This triggered its re-routing code and so it was
6077 trying to route what was effectively the same domain again. This normally
6078 caused routing to fail (because the router wouldn't handle the domain
6079 twice). Now Exim checks for this case specially, and just changes the
6080 casing of the domain that it ultimately uses when it transmits the message
6083 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6086 50. If a filter generated a file delivery with a non-absolute name (possible if
6087 no home directory exists for the router), the forbid_file option was not
6090 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6091 the existing equality matching.
6093 52. Exim was using ints instead of ino_t variables in some places where it was
6094 dealing with inode numbers.
6096 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6097 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6098 called TMPDIR, and if it finds it is different, it changes its value.
6100 54. The smtp_printf() function is now made available to local_scan() so
6101 additional output lines can be written before returning. There is also an
6102 smtp_fflush() function to enable the detection of a dropped connection.
6103 The variables smtp_input and smtp_batched_input are exported to
6106 55. Changed the default runtime configuration: the message "Unknown user"
6107 has been removed from the ACL, and instead placed on the localuser router,
6108 using the cannot_route_message feature. This means that any verification
6109 failures that generate their own messages won't get overridden. Similarly,
6110 the "Unrouteable address" message that was in the ACL for unverifiable
6111 relay addresses has also been removed.
6113 56. Added hosts_avoid_esmtp to the smtp transport.
6115 57. The exicyclog script was not checking for the esoteric option
6116 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6117 will work only if exicyclog is run under the appropriate euid.
6119 58. Following a discussion on the list, the rules by which Exim recognises line
6120 endings on incoming messages have been changed. The -dropcr and drop_cr
6121 options are now no-ops, retained only for backwards compatibility. The
6122 following line terminators are recognized: LF CRLF CR. However, special
6123 processing applies to CR:
6125 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6126 nor a local message in the state where . is a terminator.
6128 (ii) If a bare CR is encountered in a header line, an extra space is added
6129 after the line terminator so as not to end the header. The reasoning
6130 behind this is that bare CRs in header lines are most likely either
6131 to be mistakes, or people trying to play silly games.
6133 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6134 was being incorrectly given as 18 bytes larger than it should have been.
6135 This is a VOB (very old bug).
6137 60. This may never have affected anything current, but just in case it has:
6138 When the local host is found other than at the start of a list of hosts,
6139 the local host, those with the same MX, and any that follow, are discarded.
6140 When the list in question was part of a longer list of hosts, the following
6141 hosts (not currently being processed) were also being discarded. This no
6142 longer happens. I'm not sure if this situation could ever has previously
6145 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6148 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6149 header. This is something that is recommended in a new Internet Draft, and
6150 is something that is documented as being done by Sendmail. There are two
6151 possible values. For messages generated by the autoreply transport, Exim
6154 Auto-Submitted: auto-replied
6156 whereas for all other generated messages (e.g. bounces) it adds
6158 Auto-Submitted: auto-generated
6160 63. The "personal" condition in filters now includes a test for the
6161 Auto-Submitted: header. If it contains the string "auto-" the message it
6162 not considered personal.
6164 64. Added rcpt_include_affixes as a generic transport option.
6166 65. Added queue_only_override (default true).
6168 66. Added the syslog_duplication option.
6170 67. If what should have been the first header line of a message consisted of
6171 a space followed by a colon, Exim was mis-interpreting it as a header line.
6172 It isn't of course - it is syntactically invalid and should therefore be
6173 treated as the start of the message body. The misbehaviour could have
6174 caused a number of strange effects, including loss of data in subsequent
6175 header lines, and spool format errors.
6177 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6178 client host had authenticated. This control can now be exercised by an ACL
6179 for more flexibility.
6181 69. By default, callouts do not happen when testing with -bh. There is now a
6182 variant, -bhc, which does actually run the callout code, including
6183 consulting and updating the callout cache.
6185 70. Added support for saslauthd authentication, courtesy of Alexander
6188 71. If statvfs() failed on the spool or log directories while checking their
6189 size for availability, Exim confusingly gave the error "space shortage".
6190 Furthermore, in debugging mode it crashed with a floating point exception.
6191 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6192 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6193 serious problem, Exim now writes to the main and panic logs when this
6194 happens, with details of the failure. It then refuses to accept the
6195 incoming message, giving the message "spool directory problem" or "log
6196 directory problem" with a 421 code for SMTP messages.
6198 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6199 0, 1, and 2 exist, because some OS complain for execs without them (see
6200 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6201 descriptors. However, the code omitted to check that the open succeeded,
6202 causing mysterious errors if for some reason the permissions on /dev/null
6203 got screwed. Now Exim writes a message to the main and panic logs, and
6204 bombs out if it can't open /dev/null.
6206 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6207 interact so that it is all more flexible. It is supposed to remain
6208 backwards compatible. Also added extra_local_interfaces.
6210 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6211 to bomb out with an assertion failure - to the client this appears as a
6212 connection drop. This problem occurs in the part of the code that was taken
6213 from the Samba project. Fortunately, the assertion is in a very simple
6214 function, so I have fixed this by reproducing the function inline in the
6215 one place where it is called, and arranging for authentication to fail
6216 instead of killing the process with assert().
6218 75. The SPA client code was not working when the server requested OEM rather
6219 than Unicode encoding.
6221 76. Added code to make require_files with a specific uid setting more usable in
6222 the case where statting the file as root fails - usually a non-root-mounted
6223 NFS file system. When this happens and the failure is EACCES, Exim now
6224 forks a subprocess and does the per-uid checking as the relevant uid.
6226 77. Added process_log_path.
6228 78. If log_file_path was not explicitly set, a setting of check_log_space or
6229 check_log_inodes was ignored.
6231 79. If a space check for the spool or log partitions fails, the incident is now
6232 logged. Of course, in the latter case the data may get lost...
6234 80. Added the %p formatting code to string_format() so that it can be used to
6235 print addresses in debug_print(). Adjusted all the address printing in the
6236 debugging in store.c to use %p rather than %d.
6238 81. There was a concern that a line of code in smtp_in.c could overflow a
6239 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6240 initially expressed, the concern was not well-founded, because trailing
6241 spaces are removed early. However, if the trailing spaces were followed by
6242 a NULL, they did not get removed, so the overflow was possible. Two fixes
6245 (a) I re-wrote the offending code in a cleaner fashion.
6246 (b) If an incoming SMTP command contains a NULL character, it is rejected
6249 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6250 runs initgroups(), so that if the Exim user is in any additional groups,
6251 they will be used during message reception.
6257 The change log for 4.20 and earlier releases has been archived.