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18 EXPERIMENTAL_CERTNAMES is withdrawn.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 priority
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
76 JH/21 Support for the A6 type of dns record is withdrawn.
82 TL/01 When running the test suite, the README says that variables such as
83 no_msglog_check are global and can be placed anywhere in a specific
84 test's script, however it was observed that placement needed to be near
85 the beginning for it to behave that way. Changed the runtest perl
86 script to read through the entire script once to detect and set these
87 variables, reset to the beginning of the script, and then run through
88 the script parsing/test process like normal.
90 TL/02 The BSD's have an arc4random API. One of the functions to induce
91 adding randomness was arc4random_stir(), but it has been removed in
92 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
93 function when detected.
95 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
96 cause callback expansion.
98 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
99 syntax errors in an expansion can be treated as a string instead of
100 logging or causing an error, due to the internal use of bool_lax
101 instead of bool when processing it.
103 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
104 server certificates when making smtp deliveries.
106 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
108 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
110 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
112 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
113 Merged patch from Sebastian Wiedenroth.
115 JH/05 Fix results-pipe from transport process. Several recipients, combined
116 with certificate use, exposed issues where response data items split
117 over buffer boundaries were not parsed properly. This eventually
118 resulted in duplicates being sent. This issue only became common enough
119 to notice due to the introduction of conection certificate information,
120 the item size being so much larger. Found and fixed by Wolfgang Breyha.
122 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
123 size buffer was used, resulting in syntax errors when an expansion
126 JH/07 Add support for directories of certificates when compiled with a GnuTLS
127 version 3.3.6 or later.
129 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
130 is EXPERIMENTAL_EVENT, the main-configuration and transport options
131 both become "event_action", the variables become $event_name, $event_data
132 and $event_defer_errno. There is a new variable $verify_mode, usable in
133 routers, transports and related events. The tls:cert event is now also
134 raised for inbound connections, if the main configuration event_action
137 TL/06 In test suite, disable OCSP for old versions of openssl which contained
138 early OCSP support, but no stapling (appears to be less than 1.0.0).
140 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
141 server certificate names available under the smtp transport option
142 "tls_verify_cert_hostname" now do not permit multi-component wildcard
145 JH/10 Time-related extraction expansions from certificates now use the main
146 option "timezone" setting for output formatting, and are consistent
147 between OpenSSL and GnuTLS compilations. Bug 1541.
149 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
150 encoded parameter in the incoming message. Bug 1558.
152 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
153 include certificate info, eximon was claiming there were spoolfile
156 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
158 JH/14 Log delivery-related information more consistently, using the sequence
159 "H=<name> [<ip>]" wherever possible.
161 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
162 are problematic for Debian distribution, omit them from the release
165 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
167 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
170 JH/17 Fix a null-indirection in certextract expansions when a nondefault
171 output list separator was used.
176 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
177 checkers that were complaining about end of non-void function with no
180 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
181 This was a regression intruduced in 4.83 by another bugfix.
183 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
185 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
186 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
192 TF/01 Correctly close the server side of TLS when forking for delivery.
194 When a message was received over SMTP with TLS, Exim failed to clear up
195 the incoming connection properly after forking off the child process to
196 deliver the message. In some situations the subsequent outgoing
197 delivery connection happened to have the same fd number as the incoming
198 connection previously had. Exim would try to use TLS and fail, logging
199 a "Bad file descriptor" error.
201 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
202 utilities have not been installed.
204 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
205 temporary space as the ACL may create new global variables.
207 TL/01 LDAP support uses per connection or global context settings, depending
208 upon the detected version of the libraries at build time.
210 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
211 to extract and use the src ip:port in logging and expansions as if it
212 were a direct connection from the outside internet. PPv2 support was
213 updated based on HAProxy spec change in May 2014.
215 JH/02 Add ${listextract {number}{list}{success}{fail}}.
217 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
218 Properly escape header and check for NULL return.
220 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
223 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
225 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
226 characters in header names, implemented as a verify condition.
227 Contributed by Michael Fischer v. Mollard.
229 TL/05 Rename SPF condition results err_perm and err_temp to standardized
230 results permerror and temperror. Previous values are deprecated but
231 still accepted. In a future release, err_perm and err_temp will be
232 completely removed, which will be a backward incompatibility if the
233 ACL tests for either of these two old results. Patch contributed by
234 user bes-internal on the mailing list.
236 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
238 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
239 selectors, in both main and reject logs.
241 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
244 JH/07 Add malware type "sock" for talking to simple daemon.
246 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
248 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
249 routers/transports under cutthrough routing.
251 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
252 numbers. Touch up "bool" conditional to keep the same definition.
254 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
256 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
258 JH/12 Expand items in router/transport headers_add or headers_remove lists
259 individually rather than the list as a whole. Bug 1452.
261 Required for reasonable handling of multiple headers_ options when
262 they may be empty; requires that headers_remove items with embedded
263 colons must have them doubled (or the list-separator changed).
265 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
266 view the policy declared in the DMARC record. Currently, $dmarc_status
267 is a combined value of both the record presence and the result of the
270 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
272 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
273 dnslookup router and the smtp transport (applying to the forward
276 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
277 of ldap servers used for a specific lookup. Patch provided by Heiko
280 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
281 New variable $lookup_dnssec_authenticated for observability.
283 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
284 Patch submitted by Lars Timman.
286 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
288 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
289 Requires trusted mode and valid format message id, aborts otherwise.
290 Patch contributed by Heiko Schlichting.
292 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
293 certextract with support for various fields. Bug 1358.
295 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
296 is requested by default, modifiable by smtp transport option
299 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
300 operate on certificate variables to give certificate fingerprints
301 Also new ${sha256:cert_variable}.
303 JH/23 The PRDR feature is moved from being Experimental into the mainline.
305 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
306 Christian Aistleitner.
308 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
310 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
311 file. Patch from Wolfgang Breyha.
313 JH/25 Expand the coverage of the delivery $host and $host_address to
314 client authenticators run in verify callout. Bug 1476.
316 JH/26 Port service names are now accepted for tls_on_connect_ports, to
317 align with daemon_smtp_ports. Bug 72.
319 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
320 support and error reporting did not work properly.
322 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
323 and is readable. Patch from Andrew Colin Kissa.
325 TL/14 Enhance documentation of ${run expansion and how it parses the
326 commandline after expansion, particularly in the case when an
327 unquoted variable expansion results in an empty value.
329 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
331 PP/02 Fix internal collision of T_APL on systems which support RFC3123
332 by renaming away from it. Addresses GH issue 15, reported by
335 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
337 TL/15 SECURITY: prevent double expansion in math comparison functions
338 (can expand unsanitized data). Not remotely exploitable.
345 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
347 PP/02 Make -n do something, by making it not do something.
348 When combined with -bP, the name of an option is not output.
350 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
353 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
354 $sender_host_name and config options to manage this, and basic check
357 PP/05 DSCP support for outbound connections and control modifier for inbound.
359 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
360 (Only plugin which currently uses this is kerberos4, which nobody should
361 be using, but we should make it available and other future plugins might
362 conceivably use it, even though it would break NAT; stuff *should* be
363 using channel bindings instead).
365 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
366 name; added for Sendmail compatibility; requires admin caller.
367 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
368 just ignore it); requires trusted caller.
369 Also parse but ignore: -Ac -Am -X<logfile>
372 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
374 TL/02 Add +smtp_confirmation as a default logging option.
376 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
377 Patch by Magnus Holmgren from 2007-02-20.
379 TL/04 Bugzilla 1281 - Spec typo.
380 Bugzilla 1283 - Spec typo.
381 Bugzilla 1290 - Spec grammar fixes.
383 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
385 TL/06 Add Experimental DMARC support using libopendmarc libraries.
387 TL/07 Fix an out of order global option causing a segfault. Reported to dev
388 mailing list by by Dmitry Isaikin.
390 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
392 JH/02 Support "G" suffix to numbers in ${if comparisons.
394 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
396 NM/01 Bugzilla 1197 - Spec typo
397 Bugzilla 1196 - Spec examples corrections
399 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
401 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
402 gnutls_enable_pkcs11, but renamed to more accurately indicate its
405 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
406 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
408 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
409 "acl {{name}{arg}...}", and optional args on acl condition
412 JH/05 Permit multiple router/transport headers_add/remove lines.
414 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
416 JH/07 Avoid using a waiting database for a single-message-only transport.
417 Performance patch from Paul Fisher. Bugzilla 1262.
419 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
422 JH/09 Add $headers_added variable, with content from use of ACL modifier
423 add_header (but not yet added to the message). Bugzilla 199.
425 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
426 Pulled from Bugzilla 817 by Wolfgang Breyha.
428 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
430 (nb: this is the same fix as in Exim 4.80.1)
432 JH/11 Add A= logging on delivery lines, and a client_set_id option on
435 JH/12 Add optional authenticated_sender logging to A= and a log_selector
438 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
440 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
441 advertise SMTP AUTH mechanism to us, instead of a generic
442 protocol violation error. Also, make Exim more robust to bad
443 data from the Dovecot auth socket.
445 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
447 When a queue runner is handling a message, Exim first routes the
448 recipient addresses, during which it prunes them based on the retry
449 hints database. After that it attempts to deliver the message to
450 any remaining recipients. It then updates the hints database using
453 So if a recipient address works intermittently, it can get repeatedly
454 deferred at routing time. The retry hints record remains fresh so the
455 address never reaches the final cutoff time.
457 This is a fairly common occurrence when a user is bumping up against
458 their storage quota. Exim had some logic in its local delivery code
459 to deal with this. However it did not apply to per-recipient defers
460 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
462 This change adds a proper retry rule check during routing so that the
463 final cutoff time is checked against the message's age. We only do
464 this check if there is an address retry record and there is not a
465 domain retry record; this implies that previous attempts to handle
466 the address had the retry_use_local_parts option turned on. We use
467 this as an approximation for the destination being like a local
468 delivery, as in LMTP.
470 I suspect this new check makes the old local delivery cutoff check
471 redundant, but I have not verified this so I left the code in place.
473 TF/02 Correct gecos expansion when From: is a prefix of the username.
475 Test 0254 submits a message to Exim with the header
479 When I ran the test suite under the user fanf2, Exim expanded
480 the header to contain my full name, whereas it should have added
481 a Resent-Sender: header. It erroneously treats any prefix of the
482 username as equal to the username.
484 This change corrects that bug.
486 GF/01 DCC debug and logging tidyup
487 Error conditions log to paniclog rather than rejectlog.
488 Debug lines prefixed by "DCC: " to remove any ambiguity.
490 TF/03 Avoid unnecessary rebuilds of lookup-related code.
492 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
493 Bug spotted by Jeremy Harris; was flawed since initial commit.
494 Would have resulted in OCSP responses post-SNI triggering an Exim
495 NULL dereference and crash.
497 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
499 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
500 Bug detection, analysis and fix by Samuel Thibault.
501 Bugzilla 1331, Debian bug #698092.
503 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
505 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
506 Server implementation by Todd Lyons, client by JH.
507 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
508 config variable "prdr_enable" controls whether the server
509 advertises the facility. If the client requests PRDR a new
510 acl_data_smtp_prdr ACL is called once for each recipient, after
511 the body content is received and before the acl_smtp_data ACL.
512 The client is controlled by bolth of: a hosts_try_prdr option
513 on the smtp transport, and the server advertisement.
514 Default client logging of deliveries and rejections involving
515 PRDR are flagged with the string "PRDR".
517 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
518 fclose(). Diagnosis by Todd Lyons.
520 PP/17 Update configure.default to handle IPv6 localhost better.
521 Patch by Alain Williams (plus minor tweaks).
524 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
525 This is now consistent with GnuTLS, and is now documented: the
526 previous undocumented portable approach to treating the option as
527 unset was to force an expansion failure. That still works, and
528 an empty string is now equivalent.
530 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
531 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
532 not performing validation itself.
534 PP/20 Added force_command boolean option to pipe transport.
535 Patch from Nick Koston, of cPanel Inc.
537 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
540 TF/04 Added udpsend ACL modifer and hexquote expansion operator
542 PP/21 Fix eximon continuous updating with timestamped log-files.
543 Broken in a format-string cleanup in 4.80, missed when I repaired the
544 other false fix of the same issue.
545 Report and fix from Heiko Schlichting.
548 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
549 Report from Prashanth Katuri.
551 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
552 It's SecureTransport, so affects any MacOS clients which use the
553 system-integrated TLS libraries, including email clients.
555 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
556 using a MIME ACL for non-SMTP local injection.
557 Report and assistance in diagnosis by Warren Baker.
559 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
561 JH/16 Fix comparisons for 64b. Bugzilla 1385.
563 TL/09 Add expansion variable $authenticated_fail_id to keep track of
564 last id that failed so it may be referenced in subsequent ACL's.
566 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
569 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
570 ldap library initialization, allowing self-signed CA's to be
571 used. Also properly sets require_cert option later in code by
572 using NULL (global ldap config) instead of ldap handle (per
573 session). Bug diagnosis and testing by alxgomz.
575 TL/12 Enhanced documentation in the ratelimit.pl script provided in
576 the src/util/ subdirectory.
578 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
579 renamed to Transport Post Delivery Action by Jeremy Harris, as
582 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
583 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
584 redis_servers = needs to be configured which will be used by the redis
585 lookup. Patch from Warren Baker, of The Packet Hub.
587 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
589 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
590 hostname or reverse DNS when processing a host list. Used suggestions
591 from multiple comments on this bug.
593 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
595 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
596 Missed a few lines, added it to make the runtest require no keyboard
599 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
600 contains upper case chars. Make router use caseful_local_part.
602 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
603 support when GnuTLS has been built with p11-kit.
609 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
611 This, or similar/improved, will also be change PP/11 of 4.82.
617 PP/01 Handle short writes when writing local log-files.
618 In practice, only affects FreeBSD (8 onwards).
619 Bugzilla 1053, with thanks to Dmitry Isaikin.
621 NM/01 Bugzilla 949 - Documentation tweak
623 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
626 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
628 PP/02 Implemented gsasl authenticator.
630 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
632 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
633 `pkg-config foo` for cflags/libs.
635 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
636 with rest of GSASL and with heimdal_gssapi.
638 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
639 `pkg-config foo` for cflags/libs for the TLS implementation.
641 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
642 properties get this fed in as external SSF. A number of robustness
643 and debugging improvements to the cyrus_sasl authenticator.
645 PP/08 cyrus_sasl server now expands the server_realm option.
647 PP/09 Bugzilla 1214 - Log authentication information in reject log.
648 Patch by Jeremy Harris.
650 PP/10 Added dbmjz lookup type.
652 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
654 PP/12 MAIL args handles TAB as well as SP, for better interop with
655 non-compliant senders.
656 Analysis and variant patch by Todd Lyons.
658 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
659 Bug report from Lars Müller <lars@samba.org> (via SUSE),
660 Patch from Dirk Mueller <dmueller@suse.com>
662 PP/13 tls_peerdn now print-escaped for spool files.
663 Observed some $tls_peerdn in wild which contained \n, which resulted
664 in spool file corruption.
666 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
667 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
668 or write after TLS renegotiation, which otherwise led to messages
671 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
672 as a tracking header (ie: a signed header comes before the signature).
673 Patch from Wolfgang Breyha.
675 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
676 comma-sep list; embedded commas doubled.
678 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
680 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
682 Report and patch from Dmitry Banschikov.
684 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
685 Removed SSL_clear() after SSL_new() which led to protocol negotiation
686 failures. We appear to now support TLS1.1+ with Exim.
688 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
689 lets Exim select keys and certificates based upon TLS SNI from client.
690 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
691 before an outbound SMTP session. New log_selector, +tls_sni.
693 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
694 NULL dereference. Report and patch from Alun Jones.
696 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
697 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
698 Not seeing resolver debug output on NetBSD, but suspect this is a
699 resolver implementation change.
701 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
702 Left warnings. Added "eximon gdb" invocation mode.
704 PP/21 Defaulting "accept_8bitmime" to true, not false.
706 PP/22 Added -bw for inetd wait mode support.
708 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
709 locate the relevant includes and libraries. Made this the default.
711 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
712 Bugzilla 1246, report and most of solution from Tomasz Kusy.
714 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
715 This may cause build issues on older platforms.
717 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
718 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
719 gnutls_require_mac & gnutls_require_protocols (no longer supported).
720 Added SNI support via GnuTLS too.
721 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
723 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
725 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
726 Applied second patch from Janne, implementing suggestion to default
727 multiple-strings-in-record handling to match SPF spec.
729 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
731 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
732 read-only, out of scope).
733 Patch from Wolfgang Breyha, report from Stuart Northfield.
735 PP/29 Fix three issues highlighted by clang analyser static analysis.
736 Only crash-plausible issue would require the Cambridge-specific
737 iplookup router and a misconfiguration.
738 Report from Marcin Mirosław.
740 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
742 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
743 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
744 As part of this, removing so much warning spew let me fix some minor
745 real issues in debug logging.
747 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
748 assignment on my part. Fixed.
750 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
751 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
752 Janne Snabb (who went above and beyond: thank you).
754 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
755 string otherwise requires a connection and a bunch more work and it's
756 relatively easy to get wrong. Should also expose TLS library linkage
759 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
760 64-bit ${eval} (JH/03).
762 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
763 GNU libc to support some of the 64-bit stuff, should not lead to
764 conflicts. Defined before os.h is pulled in, so if a given platform
765 needs to override this, it can.
767 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
768 protection layer was required, which is not implemented.
769 Bugzilla 1254, patch from Wolfgang Breyha.
771 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
772 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
773 tls_dhparam take prime identifiers. Also unbreak combination of
774 OpenSSL+DH_params+TLSSNI.
776 PP/39 Disable SSLv2 by default in OpenSSL support.
782 PP/01 Solaris build fix for Oracle's LDAP libraries.
783 Bugzilla 1109, patch from Stephen Usher.
785 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
787 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
790 TF/02 Fix a couple more cases where we did not log the error message
791 when unlink() failed. See also change 4.74-TF/03.
793 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
794 lock up or crash if it happened to be inside a call to libc when it
795 got a SIGUSR1 from exiwhat.
797 The SIGUSR1 handler appends the current process status to the process
798 log which is later printed by exiwhat. It used to use the general
799 purpose logging code to do this, but several functions it calls are
800 not safe for signals.
802 The new output code in the SIGUSR1 handler is specific to the process
803 log, and simple enough that it's easy to inspect for signal safety.
804 Removing some special cases also simplifies the general logging code.
805 Removing the spurious timestamps from the process log simplifies
808 TF/04 Improved ratelimit ACL condition.
810 The /noupdate option has been deprecated in favour of /readonly which
811 has clearer semantics. The /leaky, /strict, and /readonly update modes
812 are mutually exclusive. The update mode is no longer included in the
813 database key; it just determines when the database is updated. (This
814 means that when you upgrde Exim will forget old rate measurements.)
816 Exim now checks that the per_* options are used with an update mode that
817 makes sense for the current ACL. For example, when Exim is processing a
818 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
819 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
820 must specify per_mail/readonly. If you omit the update mode it defaults to
821 /leaky where that makes sense (as before) or /readonly where required.
823 The /noupdate option is now undocumented but still supported for
824 backwards compatibility. It is equivalent to /readonly except that in
825 ACLs where /readonly is required you may specify /leaky/noupdate or
826 /strict/noupdate which are treated the same as /readonly.
828 A useful new feature is the /count= option. This is a generalization
829 of the per_byte option, so that you can measure the throughput of other
830 aggregate values. For example, the per_byte option is now equivalent
831 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
833 The per_rcpt option has been generalized using the /count= mechanism
834 (though it's more complicated than the per_byte equivalence). When it is
835 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
836 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
837 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
838 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
839 Note that using per_rcpt with a non-readonly update mode in more than
840 one ACL will cause the recipients to be double-counted. (The per_mail
841 and per_byte options don't have this problem.)
843 The handling of very low rates has changed slightly. If the computed rate
844 is less than the event's count (usually one) then this event is the first
845 after a long gap. In this case the rate is set to the same as this event's
846 count, so that the first message of a spam run is counted properly.
848 The major new feature is a mechanism for counting the rate of unique
849 events. The new per_addr option counts the number of different
850 recipients that someone has sent messages to in the last time period. It
851 behaves like per_rcpt if all the recipient addresses are different, but
852 duplicate recipient addresses do not increase the measured rate. Like
853 the /count= option this is a general mechanism, so the per_addr option
854 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
855 example, measure the rate that a client uses different sender addresses
856 with the options per_mail/unique=$sender_address. There are further
857 details in the main documentation.
859 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
861 TF/06 Removed a few PCRE remnants.
863 TF/07 Automatically extract Exim's version number from tags in the git
864 repository when doing development or release builds.
866 PP/02 Raise smtp_cmd_buffer_size to 16kB.
867 Bugzilla 879. Patch from Paul Fisher.
869 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
870 Heavily based on revision 40f9a89a from Simon Arlott's tree.
873 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
875 PP/05 Variable $av_failed, true if the AV scanner deferred.
876 Bugzilla 1078. Patch from John Horne.
878 PP/06 Stop make process more reliably on build failure.
879 Bugzilla 1087. Patch from Heiko Schlittermann.
881 PP/07 Make maildir_use_size_file an _expandable_ boolean.
882 Bugzilla 1089. Patch from Heiko Schlittermann.
884 PP/08 Handle ${run} returning more data than OS pipe buffer size.
885 Bugzilla 1131. Patch from Holger Weiß.
887 PP/09 Handle IPv6 addresses with SPF.
888 Bugzilla 860. Patch from Wolfgang Breyha.
890 PP/10 GnuTLS: support TLS 1.2 & 1.1.
892 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
895 PP/11 match_* no longer expand right-hand-side by default.
896 New compile-time build option, EXPAND_LISTMATCH_RHS.
897 New expansion conditions, "inlist", "inlisti".
899 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
901 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
903 PP/14 fix log_write() format string regression from TF/03.
904 Bugzilla 1152. Patch from Dmitry Isaikin.
910 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
912 PP/02 Harmonised TLS library version reporting; only show if debugging.
913 Layout now matches that introduced for other libraries in 4.74 PP/03.
915 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
917 PP/04 New "dns_use_edns0" global option.
919 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
922 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
923 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
925 TK/01 Updated PolarSSL code to 0.14.2.
926 Bugzilla 1097. Patch from Andreas Metzler.
928 PP/07 Catch divide-by-zero in ${eval:...}.
931 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
934 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
935 format-string attack -- SECURITY: remote arbitrary code execution.
937 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
938 time unintentionally subject to list matching rules, letting the header
939 cause arbitrary Exim lookups (of items which can occur in lists, *not*
940 arbitrary string expansion). This allowed for information disclosure.
942 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
943 INT_MIN/-1 -- value coerced to INT_MAX.
949 NM/01 Workround for PCRE version dependency in version reporting
952 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
953 This fixes portability to compilers other than gcc, notably
954 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
956 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
957 makefiles for portability to HP-UX and POSIX correctness.
959 PP/01 Permit LOOKUP_foo enabling on the make command-line.
960 Also via indented variable definition in the Makefile.
961 (Debugging by Oliver Heesakkers).
963 PP/02 Restore caching of spamd results with expanded spamd_address.
964 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
966 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
967 Improves build reliability. Fix from: Frank Elsner
969 NM/02 Fix wide character breakage in the rfc2047 coding
970 Fixes bug 1064. Patch from Andrey N. Oktyabrski
972 NM/03 Allow underscore in dnslist lookups
973 Fixes bug 1026. Patch from Graeme Fowler
975 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
976 Code patches from Adam Ciarcinski of NetBSD.
978 NM/04 Fixed exiqgrep to cope with mailq missing size issue
981 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
982 is logged, to avoid truncation. Patch from John Horne.
984 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
985 Patch from Jakob Hirsch.
987 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
988 SQL string expansion failure details.
989 Patch from Andrey Oktyabrski.
991 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
992 Patch from Simon Arlott.
994 PP/09 New lookups functionality failed to compile on old gcc which rejects
995 extern declarations in function scope.
996 Patch from Oliver Fleischmann
998 PP/10 Use sig_atomic_t for flags set from signal handlers.
999 Check getgroups() return and improve debugging.
1000 Fixed developed for diagnosis in bug 927 (which turned out to be
1003 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1004 Patch from Mark Zealey.
1006 PP/12 Bugzilla 1056: Improved spamd server selection.
1007 Patch from Mark Zealey.
1009 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1010 Based on patch from Heiko Schlittermann.
1012 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1013 Patch from Uwe Doering, sign-off by Michael Haardt.
1015 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1016 variable declaration deep within a block. Bug and patch from
1019 PP/15 lookups-Makefile IRIX compatibilty coercion.
1021 PP/16 Make DISABLE_DKIM build knob functional.
1023 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1024 Patch by Simon Arlott
1026 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1027 variable argument macros. Our copy now differs from upstream.
1033 TF/01 Failure to get a lock on a hints database can have serious
1034 consequences so log it to the panic log.
1036 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1037 controlled using the smtp_confirmation log selector.
1039 TF/03 Include the error message when we fail to unlink a spool file.
1041 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1042 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1043 for maintaining out-of-tree patches for some time.
1045 PP/01 Bugzilla 139: Documentation and portability issues.
1046 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1047 Handle per-OS dynamic-module compilation flags.
1049 PP/02 Let /dev/null have normal permissions.
1050 The 4.73 fixes were a little too stringent and complained about the
1051 permissions on /dev/null. Exempt it from some checks.
1052 Reported by Andreas M. Kirchwitz.
1054 PP/03 Report version information for many libraries, including
1055 Exim version information for dynamically loaded libraries. Created
1056 version.h, now support a version extension string for distributors
1057 who patch heavily. Dynamic module ABI change.
1059 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1060 privilege escalation vulnerability whereby the Exim run-time user
1061 can cause root to append content of the attacker's choosing to
1064 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1067 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1068 If dropping privileges for untrusted macros, we disabled normal logging
1069 on the basis that it would fail; for the Exim run-time user, this is not
1070 the case, and it resulted in successful deliveries going unlogged.
1071 Fixed. Reported by Andreas Metzler.
1077 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1078 only prepend for the Resent-* case. Fixes regression introduced in
1079 Exim 4.70 by NM/22 for Bugzilla 607.
1081 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1082 increasing numbers of administrators be bitten by this.
1084 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1086 PP/03 Bugzilla 994: added openssl_options main configuration option.
1088 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1090 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1092 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1094 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1095 without a peer certificate, leading to a segfault because of an
1096 assumption that peers always have certificates. Be a little more
1097 paranoid. Problem reported by Martin Tscholak.
1099 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1100 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1101 NB: ClamAV planning to remove STREAM in "middle of 2010".
1102 CL also introduces -bmalware, various -d+acl logging additions and
1103 more caution in buffer sizes.
1105 PP/09 Implemented reverse_ip expansion operator.
1107 PP/10 Bugzilla 937: provide a "debug" ACL control.
1109 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1111 PP/12 Bugzilla 973: Implement --version.
1113 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1115 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1117 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1119 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1120 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1121 ignore trailing whitespace.
1123 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1125 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1128 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1129 Notification from Dr Andrew Aitchison.
1131 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1132 ExtendedDetectionInfo response format.
1133 Notification from John Horne.
1135 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1138 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1139 XSL and documented dependency on system catalogs, with examples of how
1142 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1145 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1146 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1147 configuration file which is writeable by the Exim user or group.
1149 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1150 of configuration files to cover files specified with the -C option if
1151 they are going to be used with root privileges, not just the default
1154 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1155 option (effectively making it always true).
1157 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1158 files to be used while preserving root privileges.
1160 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1161 that rogue child processes cannot use them.
1163 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1164 run-time user, instead of root.
1166 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1167 Exim run-time user without dropping privileges.
1169 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1170 result string, instead of calling string_vformat() twice with the same
1173 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1174 for other users. Others should always drop root privileges if they use
1175 -C on the command line, even for a whitelisted configure file.
1177 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1179 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1185 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1186 $data_path, and $header_path variables; fixed documentation bugs and
1189 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1190 exipick to access non-standard spools, including the "frozen" queue
1193 NM/01 Bugzilla 965: Support mysql stored procedures.
1194 Patch from Alain Williams
1196 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1198 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1199 Patch from Andreas Metzler
1201 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1202 Patch from Kirill Miazine
1204 NM/05 Bugzilla 671: Added umask to procmail example.
1206 JJ/03 installed exipick 20100323.0, fixing doc bug
1208 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1209 directory. Notification and patch from Dan Rosenberg.
1211 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1213 TK/02 Improve log output when DKIM signing operation fails.
1215 MH/01 Treat the transport option dkim_domain as a colon separated
1216 list, not as a single string, and sign the message with each element,
1217 omitting multiple occurences of the same signer.
1219 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1220 Bugzilla 985, 986. Patch by Simon Arlott
1222 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1223 Patch by Simon Arlott
1225 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1226 MBX locking. Notification from Dan Rosenberg.
1232 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1234 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1236 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1238 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1240 NM/04 Bugzilla 915: Flag broken perl installation during build.
1246 TK/01 Added patch by Johannes Berg that expands the main option
1247 "spamd_address" if it starts with a dollar sign.
1249 TK/02 Write list of recipients to X-Envelope-Sender header when building
1250 the mbox-format spool file for content scanning (suggested by Jakob
1253 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1254 (http://www.dcc-servers.net/) support via dccifd. Activated by
1255 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1257 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1258 by Mark Daniel Reidel <mr@df.eu>.
1260 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1261 When building exim an external PCRE library is now needed -
1262 PCRE is a system library on the majority of modern systems.
1263 See entry on PCRE_LIBS in EDITME file.
1265 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1266 conversation. Added nologin parameter to request.
1267 Patch contributed by Kirill Miazine.
1269 TF/01 Do not log submission mode rewrites if they do not change the address.
1271 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1273 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1274 log files in place. Contributed by Roberto Lima.
1276 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1278 TF/03 Bugzilla 615: When checking the local_parts router precondition
1279 after a local_part_suffix or local_part_prefix option, Exim now
1280 does not use the address's named list lookup cache, since this
1281 contains cached lookups for the whole local part.
1283 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1284 Robert Millan. Documentation is in experimental-spec.txt.
1286 TF/04 Bugzilla 668: Fix parallel build (make -j).
1288 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1290 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1291 Patch provided by Jan Srzednicki.
1293 TF/05 Leading white space used to be stripped from $spam_report which
1294 wrecked the formatting. Now it is preserved.
1296 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1297 that they are available at delivery time.
1299 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1301 TF/08 TLS error reporting now respects the incoming_interface and
1302 incoming_port log selectors.
1304 TF/09 Produce a more useful error message if an SMTP transport's hosts
1305 setting expands to an empty string.
1307 NM/06 Bugzilla 744: EXPN did not work under TLS.
1308 Patch provided by Phil Pennock.
1310 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1311 Patch provided by Richard Godbee.
1313 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1314 acl_smtp_notquit, added index entry.
1316 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1317 Patch provided by Eugene Bujak.
1319 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1320 accept(). Patch provided by Maxim Dounin.
1322 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1323 Patch provided by Phil Pennock.
1325 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1327 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1328 Patch provided by Brad "anomie" Jorsch.
1330 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1331 Patch provided by Dean Brooks.
1333 TK/05 Add native DKIM support (does not depend on external libraries).
1335 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1336 Patch provided by Graeme Fowler.
1338 NM/16 Bugzilla 851: Documentation example syntax fix.
1340 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1342 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1345 NM/19 Bugzilla 745: TLS version reporting.
1346 Patch provided by Phil Pennock.
1348 NM/20 Bugzilla 167: bool: condition support.
1349 Patch provided by Phil Pennock.
1351 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1352 clients. Patch provided by Phil Pennock.
1354 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1355 Patch provided by Brad "anomie" Jorsch.
1357 NM/23 Bugzilla 687: Fix misparses in eximstats.
1358 Patch provided by Heiko Schlittermann.
1360 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1361 Patch provided by Heiko Schlittermann.
1363 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1364 plus update to original patch.
1366 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1368 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1369 Patch provided by David Brownlee.
1371 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1373 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1375 NM/30 Bugzilla 888: TLS documentation bugfixes.
1377 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1379 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1380 Unlike the original bugzilla I have changed all shell scripts in src tree.
1382 NM/33 Bugzilla 898: Transport filter timeout fix.
1383 Patch by Todd Rinaldo.
1385 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1386 Patch by Serge Demonchaux.
1388 NM/35 Bugzilla 39: Base64 decode bug fixes.
1389 Patch by Jakob Hirsch.
1391 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1393 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1395 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1397 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1403 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1404 ALT-N's libdkim that I have put here:
1405 http://duncanthrax.net/exim-experimental/
1407 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1408 were called 'true' and it seems that C99 defines that as a reserved
1409 keyword to be used with 'bool' variable types. That means you could
1410 not include C99-style headers which use bools without triggering
1411 build errors in sieve.c.
1413 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1414 as mailq or other aliases. Changed the --help handling significantly
1415 to do whats expected. exim_usage() emits usage/help information.
1417 SC/01 Added the -bylocaldomain option to eximstats.
1419 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1421 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1423 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1429 PH/01 Another patch from the Sieve maintainer.
1431 PH/02 When an IPv6 address is converted to a string for single-key lookup
1432 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1433 dots are used instead of colons so that keys in lsearch files need not
1434 contain colons. This was done some time before quoting was made available
1435 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1436 (notated using the quote facility) so as to distinguish them from IPv4
1437 keys. This meant that lookups for IP addresses in host lists did not work
1438 for iplsearch lookups.
1440 This has been fixed by arranging for IPv6 addresses to be expressed with
1441 colons if the lookup type is iplsearch. This is not incompatible, because
1442 previously such lookups could never work.
1444 The situation is now rather anomolous, since one *can* have colons in
1445 ordinary lsearch keys. However, making the change in all cases is
1446 incompatible and would probably break a number of configurations.
1448 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1451 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1452 conversion specification without a maximum field width, thereby enabling
1453 a rogue spamd server to cause a buffer overflow. While nobody in their
1454 right mind would setup Exim to query an untrusted spamd server, an
1455 attacker that gains access to a server running spamd could potentially
1456 exploit this vulnerability to run arbitrary code as the Exim user.
1458 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1459 $primary_hostname instead of what libspf2 thinks the hosts name is.
1461 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1462 a directory entry by the name of the lookup key. Previously, if a
1463 symlink pointed to a non-existing file or a file in a directory that
1464 Exim lacked permissions to read, a lookup for a key matching that
1465 symlink would fail. Now it is enough that a matching directory entry
1466 exists, symlink or not. (Bugzilla 503.)
1468 PH/03 The body_linecount and body_zerocount variables are now exported in the
1471 PH/04 Added the $dnslist_matched variable.
1473 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1474 This means they are set thereafter only if the connection becomes
1477 PH/06 Added the client_condition to authenticators so that some can be skipped
1478 by clients under certain conditions.
1480 PH/07 The error message for a badly-placed control=no_multiline_responses left
1481 "_responses" off the end of the name.
1483 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1485 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1486 (without spaces) instead of just copying the configuration text.
1488 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1490 PH/11 Added $max_received_linelength.
1492 PH/12 Added +ignore_defer and +include_defer to host lists.
1494 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1495 way in which PCRE > 7.0 is built.
1497 PH/14 Implemented queue_only_load_latch.
1499 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1500 MAIL command. The effect was to mangle the value on 64-bit systems.
1502 PH/16 Another patch from the Sieve maintainer.
1504 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1506 PH/18 If a system quota error occurred while trying to create the file for
1507 a maildir delivery, the message "Mailbox is full" was not appended to the
1508 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1509 only to a quota excession during the actual writing of the file.
1511 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1512 characters?) which causes problems in log lines. The DN values are now
1513 passed through string_printing() before being added to log lines.
1515 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1516 and InterBase are left for another time.)
1518 PH/21 Added message_body_newlines option.
1520 PH/22 Guard against possible overflow in moan_check_errorcopy().
1522 PH/23 POSIX allows open() to be a macro; guard against that.
1524 PH/24 If the recipient of an error message contained an @ in the local part
1525 (suitably quoted, of course), incorrect values were put in $domain and
1526 $local_part during the evaluation of errors_copy.
1532 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1533 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1536 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1537 issue a MAIL command.
1539 PH/02 In an ACL statement such as
1541 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1543 if a client was not listed at all, or was listed with a value other than
1544 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1545 the condition was not true (as it should be), so access was not denied.
1546 The bug was that the ! inversion was incorrectly passed on to the second
1547 item. This has been fixed.
1549 PH/03 Added additional dnslists conditions == and =& which are different from
1550 = and & when the dns lookup returns more than one IP address.
1552 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1553 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1555 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1556 FSYNC, which compiles an option called disable_fsync that allows for
1557 bypassing fsync(). The documentation is heavily laced with warnings.
1559 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1561 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1562 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1563 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1564 including adding "make clean"; (3) Added -fPIC when compiling the test
1565 dynamically loaded module, to get rid of a warning.
1567 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1568 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1569 = 0s. The bug is otherwise harmless.
1571 PH/07 There was a bug in the dovecot authenticator such that the value of
1572 $auth1 could be overwritten, and so not correctly preserved, after a
1573 successful authentication. This usually meant that the value preserved by
1574 the server_setid option was incorrect.
1576 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1578 PH/09 Installed PCRE release 7.0.
1580 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1581 run for batched SMTP input. It is now run at the start of every message
1582 in the batch. While fixing this I discovered that the process information
1583 (output by running exiwhat) was not always getting set for -bs and -bS
1584 input. This is fixed, and it now also says "batched" for BSMTP.
1586 PH/11 Added control=no_pipelining.
1588 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1589 patch, slightly modified), and move the expansion of helo_data till after
1590 the connection is made in the smtp transport (so it can use these
1593 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1595 PH/14 Added log_selector = +pid.
1597 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1599 PH/16 Add ${if forany and ${if forall.
1601 PH/17 Added dsn_from option to vary the From: line in DSNs.
1603 PH/18 Flush SMTP output before performing a callout, unless control =
1604 no_callout_flush is set.
1606 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1607 was true (the default) a successful delivery failed to delete the retry
1608 item, thus causing premature timeout of the address. The bug is now
1611 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1613 PH/21 Long custom messages for fakedefer and fakereject are now split up
1614 into multiline reponses in the same way that messages for "deny" and
1615 other ACL rejections are.
1617 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1618 with slight modification.
1620 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1621 draft, changing the syntax and factoring some duplicate code".
1623 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1624 for deliveries of the second and subsequent messages over the same SMTP
1627 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1628 ${reduce, with only minor "tidies".
1630 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1632 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1633 expansion side effects.
1635 PH/27 When a message times out after an over-quota error from an Exim-imposed
1636 quota, the bounce message says "mailbox is full". This message was not
1637 being given when it was a system quota that was exceeded. It now should
1640 MH/03 Made $recipients available in local_scan(). local_scan() already has
1641 better access to the recipient list through recipients_list[], but
1642 $recipients can be useful in postmaster-provided expansion strings.
1644 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1645 in the case of a MAIL command with additional options following the
1646 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1647 were accidentally chopped off.
1649 PH/29 SMTP synchronization checks are implemented when a command is read -
1650 there is a check that no more input is waiting when there shouldn't be
1651 any. However, for some commands, a delay in an ACL can mean that it is
1652 some time before the response is written. In this time, more input might
1653 arrive, invalidly. So now there are extra checks after an ACL has run for
1654 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1655 pipelining has not been advertised.
1657 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1659 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1660 returned in the SMTP response when smtp_return_error_details was false.
1661 This has been fixed.
1663 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1664 instead of the C I/O that was originally supplied, because problems were
1665 reported on Solaris.
1667 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1668 Exim which did not show up earlier: it was assuming that a call to
1669 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1670 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1671 was a macro that became an assignment, so it seemed to work. This has
1672 changed to a proper function call with a void return, hence the compile
1673 error. Exim's code has been fixed.
1675 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1678 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1680 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1682 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1683 because it turns out to be too big or there is a timeout) there is no
1684 "Completed" line in the log. When some messages of this type were
1685 selected by exigrep, they were listed as "not completed". Others were
1686 picked up by some special patterns. I have improved the selection
1687 criteria to be more general.
1689 PH/38 The host_find_failed option in the manualroute router can now be set
1690 to "ignore", to completely ignore a host whose IP address cannot be
1691 found. If all hosts are ignored, the behaviour is controlled by the new
1692 host_all_ignored option.
1694 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1695 homing or because of multiple MX records with /mx) generated more than
1696 one IP address, and the following item turned out to be the local host,
1697 all the secondary addresses of the first item were incorrectly removed
1698 from the list, along with the local host and any following hosts (which
1699 is what is supposed to happen).
1701 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1702 whoever called Exim into the -H file. In the case of the daemon it was
1703 behaving confusingly. When first started, it used values for whoever
1704 started the daemon, but after a SIGHUP it used the Exim user (because it
1705 calls itself on a restart). I have changed the code so that it now always
1708 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1709 message are rejected with the same error (e.g. no authentication or bad
1710 sender address), and a DATA command is nevertheless sent (as can happen
1711 with PIPELINING or a stupid MUA), the error message that was given to the
1712 RCPT commands is included in the rejection of the DATA command. This is
1713 intended to be helpful for MUAs that show only the final error to their
1716 PH/42 Another patch from the Sieve maintainer.
1718 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1719 Eximstats - Fixed some broken HTML links and added missing column headers
1721 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1722 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1724 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1726 PH/43 Yet another patch from the Sieve maintainer.
1728 PH/44 I found a way to check for a TCP/IP connection going away before sending
1729 the response to the final '.' that terminates a message, but only in the
1730 case where the client has not sent further data following the '.'
1731 (unfortunately, this is allowed). However, in many cases there won't be
1732 any further data because there won't be any more messages to send. A call
1733 to select() can be used: if it shows that the input is "ready", there is
1734 either input waiting, or the socket has been closed. An attempt to read
1735 the next input character can distinguish the two cases. Previously, Exim
1736 would have sent an OK response which the client would never have see.
1737 This could lead to message repetition. This fix should cure that, at
1738 least in a lot of common cases.
1740 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1741 advertised in response to EHLO.
1747 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1748 fixed by 4.65/MH/01 (is this a record?) are fixed:
1750 (i) An empty string was always treated as zero by the numeric comparison
1751 operators. This behaviour has been restored.
1753 (ii) It is documented that the numeric comparison operators always treat
1754 their arguments as decimal numbers. This was broken in that numbers
1755 starting with 0 were being interpreted as octal.
1757 While fixing these problems I realized that there was another issue that
1758 hadn't been noticed. Values of message_size_limit (both the global option
1759 and the transport option) were treated as octal if they started with 0.
1760 The documentation was vague. These values are now always treated as
1761 decimal, and I will make that clear in the documentation.
1767 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1768 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1771 MH/01 Don't check that the operands of numeric comparison operators are
1772 integers when their expansion is in "skipping" mode (fixes bug
1773 introduced by 4.64-PH/07).
1775 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1776 child addresses, Exim now panics and dies. Previously, because the count
1777 is held in a short int, deliveries were likely to be lost. As such a
1778 large number of recipients for a single message is ridiculous
1779 (performance will be very, very poor), I have chosen to impose a limit
1780 rather than extend the field.
1786 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1787 leftover -K file (the existence of which was triggered by #402).
1788 While we were at it, introduced process PID as part of the -K
1789 filename. This should rule out race conditions when creating
1792 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1793 processing considerably. Previous code took too long for large mails,
1794 triggering a timeout which in turn triggers #401.
1796 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1797 in the DK code in transports.c. sendfile() is not really portable,
1798 hence the _LINUX specificness.
1800 TF/01 In the add_headers option to the mail command in an Exim filter,
1801 there was a bug that Exim would claim a syntax error in any
1802 header after the first one which had an odd number of characters
1805 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1806 callout verification, Exim cached a "reject" for the entire domain. This
1807 is correct for most verifications, but it is not correct for a recipient
1808 verification with use_sender or use_postmaster set, because in that case
1809 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1810 case of MAIL FROM:<> rejection from other early rejections (e.g.
1811 rejection of HELO). When verifying a recipient using a non-null MAIL
1812 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1813 Whatever the result of the callout, the value of the domain cache is
1814 left unchanged (for any other kind of callout, getting as far as trying
1815 RCPT means that the domain itself is ok).
1817 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1820 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1821 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1822 session, the daemon ignores ECONNECT errors and logs others; it now
1823 ignores EPIPE as well.
1825 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1826 (quoted-printable decoding).
1828 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1829 later the small subsequent patch to fix an introduced bug.
1831 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1833 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1835 PH/08 An error is now given if message_size_limit is specified negative.
1837 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1838 to be given (somewhat) arbitrary names.
1840 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1843 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1844 miscellaneous code fixes
1846 PH/10 Added the log_reject_target ACL modifier to specify where to log
1849 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1850 hostname. This is wrong, because it relates to the incoming message (and
1851 probably the interface on which it is arriving) and not to the outgoing
1852 callout (which could be using a different interface). This has been
1853 changed to use the value of the helo_data option from the smtp transport
1854 instead - this is what is used when a message is actually being sent. If
1855 there is no remote transport (possible with a router that sets up host
1856 addresses), $smtp_active_hostname is used.
1858 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1859 tweaks were necessary in order to get it to work (see also 21 below):
1860 (a) The code assumed that strncpy() returns a negative number on buffer
1861 overflow, which isn't the case. Replaced with Exim's string_format()
1863 (b) There were several signed/unsigned issues. I just did the minimum
1864 hacking in of casts. There is scope for a larger refactoring.
1865 (c) The code used strcasecmp() which is not a standard C function.
1866 Replaced with Exim's strcmpic() function.
1867 (d) The code set only $1; it now sets $auth1 as well.
1868 (e) A simple test gave the error "authentication client didn't specify
1869 service in request". It would seem that Dovecot has changed its
1870 interface. Fortunately there's a specification; I followed it and
1871 changed what the client sends and it appears to be working now.
1873 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1876 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1877 address A is aliased to B and C, where B exists and C does not. Without
1878 -v the output is "A verified" because verification stops after a
1879 successful redirection if more than one address is generated. However,
1880 with -v the child addresses are also verified. Exim was outputting "A
1881 failed to verify" and then showing the successful verification for C,
1882 with its parentage. It now outputs "B failed to verify", showing B's
1883 parentage before showing the successful verification of C.
1885 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1886 look up a TXT record in a specific list after matching in a combined
1889 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1890 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1891 they consult the DNS. I had assumed they would set it the way they
1892 wanted; and indeed my experiments on Linux seem to show that in some
1893 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1894 To be on the safe side, however, I have now made the interface to
1895 host_find_byname() similar to host_find_bydns(), with an argument
1896 containing the DNS resolver options. The host_find_byname() function now
1897 sets these options at its start, just as host_find_bydns() does. The smtp
1898 transport options dns_qualify_single and dns_search_parents are passed to
1899 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1900 of host_find_byname() use the default settings of RES_DEFNAMES
1901 (qualify_single) but not RES_DNSRCH (search_parents).
1903 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1904 spool_read_header() do less string testing, by means of a preliminary
1905 switch on the second character of optional "-foo" lines. (This is
1906 overdue, caused by the large number of possibilities that now exist.
1907 Originally there were few.) While I was there, I also converted the
1908 str(n)cmp tests so they don't re-test the leading "-" and the first
1909 character, in the hope this might squeeze out yet more improvement.
1911 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1912 flag allowing group syntax was set by the header_syntax check but not
1913 turned off, possible causing trouble later; (2) The flag was not being
1914 set at all for the header_verify test, causing "group"-style headers to
1915 be rejected. I have now set it in this case, and also caused header_
1916 verify to ignore an empty address taken from a group. While doing this, I
1917 came across some other cases where the code for allowing group syntax
1918 while scanning a header line wasn't quite right (mostly, not resetting
1919 the flag correctly in the right place). These bugs could have caused
1920 trouble for malformed header lines. I hope it is now all correct.
1922 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1923 with the "reply" argument non-NULL. The code, however (which originally
1924 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1925 but it didn't always do it. This confused somebody who was copying the
1926 code for some other use. I have removed all the tests.
1928 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1929 feature that was used to support insecure browsers during the U.S. crypto
1930 embargo. It requires special client support, and Exim is probably the
1931 only MTA that supported it -- and would never use it because real RSA is
1932 always available. This code has been removed, because it had the bad
1933 effect of slowing Exim down by computing (never used) parameters for the
1934 RSA_EXPORT functionality.
1936 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1937 authenticator to fail if there's a tab character in the incoming data
1938 (there should never be unless someone is messing about, as it's supposed
1939 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1940 if the connection is using TLS or if the remote IP is the same as the
1941 local IP, and the "valid-client-cert option" if a client certificate has
1944 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1945 authenticators. This can be used for authorization after authentication
1946 succeeds. (In the case of plaintext, it servers for both authentication
1949 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1950 if any retry times were supplied.
1952 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1953 connection, where there is no client IP address to check. In this
1954 situation, the verify now always succeeds.
1956 PH/25 Applied John Jetmore's -Mset patch.
1958 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1960 PH/27 In a string expansion for a processed (not raw) header when multiple
1961 headers of the same name were present, leading whitespace was being
1962 removed from all of them, but trailing whitespace was being removed only
1963 from the last one. Now trailing whitespace is removed from each header
1964 before concatenation. Completely empty headers in a concatenation (as
1965 before) are ignored.
1967 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1968 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1970 PH/29 [Removed. This was a change that I later backed out, and forgot to
1971 correct the ChangeLog entry (that I had efficiently created) before
1972 committing the later change.]
1974 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1975 address errors (4xx response to RCPT) over the same connection as other
1976 messages routed to the same hosts. Such deliveries are always "forced",
1977 so retry times are not inspected. This resulted in far too many retries
1978 for the affected addresses. The effect occurred only when there were more
1979 hosts than the hosts_max_try setting in the smtp transport when it had
1980 the 4xx errors. Those hosts that it had tried were not added to the list
1981 of hosts for which the message was waiting, so if all were tried, there
1982 was no problem. Two fixes have been applied:
1984 (i) If there are any address or message errors in an SMTP delivery, none
1985 of the hosts (tried or untried) are now added to the list of hosts
1986 for which the message is waiting, so the message should not be a
1987 candidate for sending over the same connection that was used for a
1988 successful delivery of some other message. This seems entirely
1989 reasonable: after all the message is NOT "waiting for some host".
1990 This is so "obvious" that I'm not sure why it wasn't done
1991 previously. Hope I haven't missed anything, but it can't do any
1992 harm, as the worst effect is to miss an optimization.
1994 (ii) If, despite (i), such a delivery is accidentally attempted, the
1995 routing retry time is respected, so at least it doesn't keep
1996 hammering the server.
1998 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1999 in ${readsocket because some servers need this prod.
2001 PH/32 Added some extra debug output when updating a wait-xxx database.
2003 PH/33 The hint "could be header name not terminated by colon", which has been
2004 given for certain expansion errors for a long time, was not being given
2005 for the ${if def:h_colon_omitted{... case.
2007 PH/34 The spec says: "With one important exception, whenever a domain list is
2008 being scanned, $domain contains the subject domain." There was at least
2009 one case where this was not true.
2011 PH/35 The error "getsockname() failed: connection reset by peer" was being
2012 written to the panic log as well as the main log, but it isn't really
2013 panic-worthy as it just means the connection died rather early on. I have
2014 removed the panic log writing for the ECONNRESET error when getsockname()
2017 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2018 runs only) independently of the message's sender address. This meant
2019 that, if the 4xx error was in fact related to the sender, a different
2020 message to the same recipient with a different sender could confuse
2021 things. In particualar, this can happen when sending to a greylisting
2022 server, but other circumstances could also provoke similar problems.
2023 I have changed the default so that the retry time for these errors is now
2024 based a combination of the sender and recipient addresses. This change
2025 can be overridden by setting address_retry_include_sender=false in the
2028 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2029 remote server are returned as part of bounce messages. This was not
2030 happening for LMTP over a pipe (the lmtp transport), but now it is the
2031 same for both kinds of LMTP.
2033 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2034 in header lines that were in fact CNAMEs. This is no longer the case.
2036 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2037 and queue runs started by the daemon processed all messages. This has
2038 been fixed so that -R and -S can now usefully be given with -q<time>.
2040 PH/40 Import PCRE release 6.7 (fixes some bugs).
2042 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2044 PH/42 Give an error if -q is specified more than once.
2046 PH/43 Renamed the variables $interface_address and $interface_port as
2047 $received_ip_address and $received_port, to make it clear that these
2048 values apply to message reception, and not to the outgoing interface when
2049 a message is delivered. (The old names remain recognized, of course.)
2051 PH/44 There was no timeout on the connect() call when using a Unix domain
2052 socket in the ${readsocket expansion. There now is.
2054 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2055 be meaningful with "accept".
2057 SC/01 Eximstats V1.43
2058 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2060 SC/02 Eximstats V1.44
2061 Use a glob alias rather than an array ref in the generated
2062 parser. This improves both readability and performance.
2064 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2065 Collect SpamAssassin and rejection statistics.
2066 Don't display local sender or destination tables unless
2067 there is data to show.
2068 Added average volumes into the top table text output.
2070 SC/04 Eximstats V1.46
2071 Collect data on the number of addresses (recipients)
2072 as well as the number of messages.
2074 SC/05 Eximstats V1.47
2075 Added 'Message too big' to the list of mail rejection
2076 reasons (thanks to Marco Gaiarin).
2078 SC/06 Eximstats V1.48
2079 Mainlog lines which have GMT offsets and are too short to
2080 have a flag are now skipped.
2082 SC/07 Eximstats V1.49 (Alain Williams)
2083 Added the -emptyok flag.
2085 SC/08 Eximstats V1.50
2086 Fixes for obtaining the IP address from reject messages.
2088 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2089 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2090 whitesspace changes from 4.64-PH/27
2092 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2095 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2096 are found, allow negative numbers in numeric criteria)
2098 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2100 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2101 to match changes made in 4.64-PH/43
2103 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2105 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2106 "domain" is actually an IP address, and force a failure. This locks out
2107 those revolvers/nameservers that support "A-for-A" lookups, in
2108 contravention of the specifications.
2110 PH/48 When a host name was looked up from an IP address, and the subsequent
2111 forward lookup of the name timed out, the host name was left in
2112 $sender_host_name, contrary to the specification.
2114 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2115 restricted to single-key lookups, Exim was not diagnosing an error if
2116 * or *@ was used with a query-style lookup.
2118 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2120 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2121 long ago, but noone interested enough thought of it. Let's just say that
2122 the "1.1" means that there are some new functions that weren't there at
2123 some point in the past.
2125 PH/51 Error processing for expansion failure of helo_data from an smtp
2126 transport during callout processing was broken.
2128 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2129 tested/used via the -bh/-bhc/-bs options.
2131 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2132 bug, fixed in subsequent PCRE releases).
2134 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2135 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2137 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2143 SC/01 Use a glob alias rather than an array ref in eximstats generated
2144 parser. This improves both readability and performance.
2146 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2147 Don't display local sender or destination tables in eximstats unless
2148 there is data to show.
2149 Added average volumes into the eximstats top table text output.
2151 SC/03 Collect data on the number of addresses (recipients) as well
2152 as the number of messages in eximstats.
2154 TF/01 Correct an error in the documentation for the redirect router. Exim
2155 does (usually) call initgroups() when daemonizing.
2157 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2158 with consistent privilege compared to when running as a daemon.
2160 TF/03 Note in the spec that $authenticated_id is not set for local
2161 submissions from trusted users.
2163 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2164 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2166 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2167 by adding some example configuration directives to the default
2168 configuration file. A little bit of work is required to uncomment the
2169 directives and define how usernames and passwords are checked, but
2170 there is now a framework to start from.
2172 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2173 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2174 without this. I don't know how relevant this is to other LDAP libraries.
2176 PH/02 Add the verb name to the "unknown ACL verb" error.
2178 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2180 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2182 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2183 directory not expanded when it should be if an expanded home directory
2184 was set for the address (which is overridden by the transport).
2186 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2189 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2190 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2191 because it is too late at that time, and has no effect.
2193 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2194 security issue with \' (bugzilla #107). I could not use the
2195 PQescapeStringConn() function, because it needs a PGconn value as one of
2198 PH/08 When testing addresses using -bt, indicate those final addresses that
2199 are duplicates that would not cause an additional delivery. At least one
2200 person was confused, thinking that -bt output corresponded to deliveries.
2201 (Suppressing duplicates isn't a good idea as you lose the information
2202 about possibly different redirections that led to the duplicates.)
2204 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2205 systems where poll() doesn't work, in particular OS X.
2207 PH/10 Added more information to debugging output for retry time not reached.
2209 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2210 operations in malware.c.
2212 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2215 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2216 syslog_duplication set false, log lines that would normally be written
2217 both the the main log and to the reject log were not written to syslog at
2220 PH/14 In the default configuration, change the use of "message" in ACL warn
2221 statements to "add_header".
2223 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2224 not followed by a command (e.g. "seen endif").
2226 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2227 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2230 PH/17 Added extra conditions to the default value of delay_warning_condition
2234 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2235 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2236 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2239 The Auto-Submitted: and various List- headers are standardised, whereas I
2240 don't think Precedence: ever was.
2242 PH/18 Refactored debugging code in route_finduser() to show more information,
2243 in particular, the error code if getpwnam() issues one.
2245 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2246 This is apparently needed in addition to the PH/07 change above to avoid
2247 any possible encoding problems.
2249 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2250 but not after initializing Perl.
2252 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2253 output them only if debugging. By default they are written stderr,
2254 apparently, which is not desirable.
2256 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2259 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2262 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2264 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2265 authenticated or an ident call has been made. Suppress the default
2266 values for $authenticated_id and $authenticated_sender (but permit -oMai
2267 and -oMas) when testing with -bh.
2269 PH/24 Re-jigged the order of the tests in the default configuration so that the
2270 tests for valid domains and recipients precede the DNS black list and CSA
2271 tests, on the grounds that those ones are more expensive.
2273 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2274 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2275 This bug exists in every version of Exim that I still have, right back to
2278 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2279 However, an attempt to turn on case-sensitivity in a regex key by
2280 including (?-i) didn't work because the subject string was already
2281 lowercased, and the effects were non-intuitive. It turns out that a
2282 one-line patch can be used to allow (?-i) to work as expected.
2288 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2289 other effects) broke the use of negated acl sub-conditions.
2291 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2294 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2295 "Deny" causes Exim to reject the incoming connection with a 554 error.
2296 Unfortunately, if there is a major crisis, such as a disk failure,
2297 tcp-wrappers gives "deny", whereas what one would like would be some
2298 kind of temporary error. A kludge has been added to help with this.
2299 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2300 554 error is used if errno is still zero or contains ENOENT (which occurs
2301 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2304 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2306 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2307 errors. Otherwise a message that provokes a temporary error (when other
2308 messages do not) can cause a whole host to time out.
2310 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2311 addresses were routed directly to files or pipes from a redirect router.
2312 File deliveries just didn't batch; pipe deliveries might have suffered
2315 PH/06 A failure to get a lock for a hints database would erroneously always say
2316 "Failed to get write lock", even when it was really a read lock.
2318 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2319 of 0600. This has been changed to use the value of the lockfile_mode
2320 option (which defaults to 0600).
2322 PH/08 Applied small patch from the Sieve maintainer.
2324 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2325 folder from quota calculations, a direct delivery into this folder messed
2326 up the contents of the maildirsize file. This was because the regex was
2327 used only to exclude .Trash (or whatever) when the size of the mailbox
2328 was calculated. There was no check that a delivery was happening into an
2329 excluded directory. This bug has been fixed by ignoring all quota
2330 processing for deliveries into excluded directories.
2332 PH/10 Added the maildirfolder_create_regex option to appendfile.
2338 PH/01 The code for finding all the local interface addresses on a FreeBSD
2339 system running IPv6 was broken. This may well have applied to all BSD
2340 systems, as well as to others that have similar system calls. The broken
2341 code found IPv4 interfaces correctly, but gave incorrect values for the
2342 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2343 that it would not match correctly against @[] and not recognize the IPv6
2346 PH/02 The ipliteral router was not recognizing addresses of the form user@
2347 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2349 PH/03 Added disable_ipv6.
2351 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2352 decoded form, because it is most often used to construct To: headers
2353 lines in autoreplies, and the decoded form may well be syntactically
2354 invalid. However, $reply_address has leading white space removed, and all
2355 newlines turned into spaces so that the autoreply transport does not
2358 PH/05 If group was specified without a user on a router, and no group or user
2359 was specified on a transport, the group from the router was ignored.
2361 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2362 for visible compile-time settings that can be used to change these
2363 numbers, for those that want even more. Backwards compatibility with old
2364 spool files has been maintained. However, going back to a previous Exim
2365 release will lost any variables that are in spool files.
2367 PH/07 Two small changes when running in the test harness: increase delay when
2368 passing a TCP/IP connection to a new process, in case the original
2369 process has to generate a bounce, and remove special handling of
2370 127.0.0.2 (sic), which is no longer necessary.
2372 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2373 be the same on different OS.
2375 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2378 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2379 whether --show-vars was specified or not
2381 JJ/02 exipick: Added support for new ACL variable spool format introduced
2384 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2385 syntactically invalid From: or Reply-to: line, and a filter used this to
2386 generate an autoreply, and therefore failed to obtain an address for the
2387 autoreply, Exim could try to deliver to a non-existent relative file
2388 name, causing unrelated and misleading errors. What now happens is that
2389 it logs this as a hard delivery error, but does not attempt to create a
2392 PH/11 The exinext utility has a -C option for testing purposes, but although
2393 the given file was scanned by exinext itself; it wasn't being passed on
2394 when Exim was called.
2396 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2397 an end-of-file indication when reading a command response.
2399 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2400 compiled. In many other places in Exim, IPv6 addresses are always
2401 recognized, so I have changed this. It also means that IPv4 domain
2402 literals of the form [IPV4:n.n.n.n] are now always recognized.
2404 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2405 used if the router is not running as root, for example, when verifying at
2406 ACL time, or when using -bh. The debugging output from this situation was
2407 non-existent - all you got was a failure to exec. I have made two
2410 (a) Failures to set uid/gid, the current directory, or a process leader
2411 in a subprocess such as that created by queryprogram now generate
2412 suitable debugging ouput when -d is set.
2414 (b) The queryprogram router detects when it is not running as root,
2415 outputs suitable debugging information if -d is set, and then runs
2416 the subprocess without attempting to change uid/gid.
2418 PH/15 Minor change to Makefile for building test_host (undocumented testing
2421 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2422 additional section of a DNS packet that returns MX or SRV records.
2423 Instead, it always explicitly searches for A/AAAA records. This avoids
2424 major problems that occur when a DNS server includes only records of one
2425 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2426 fixed another bug: if SRV records were looked up and the corresponding
2427 address records were *not* found in the additional section, the port
2428 values from the SRV records were lost.
2430 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2431 using the correct key (the original address) when searching the retry
2432 rules in order to find which one to use for generating the retry hint.
2434 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2435 adding the default one. Similarly, if it contains a Reply-To: header, the
2436 errors_reply_to option, if set, is not used.
2438 PH/19 When calculating a retry time, Exim used to measure the "time since
2439 failure" by looking at the "first failed" field in the retry record. Now
2440 it does not use this if it is later than than the arrival time of the
2441 message. Instead it uses the arrival time. This makes for better
2442 behaviour in cases where some deliveries succeed, thus re-setting the
2443 "first failed" field. An example is a quota failure for a huge message
2444 when small messages continue to be delivered. Without this change, the
2445 "time since failure" will always be short, possible causing more frequent
2446 delivery attempts for the huge message than are intended.
2447 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2449 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2450 $1, $2, $3) because the numerical variables can be reset during some
2451 expansion items (e.g. "match"), thereby losing the authentication data.
2453 PH/21 Make -bV show the size of off_t variables so that the test suite can
2454 decide whether to run tests for quotas > 2G.
2456 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2457 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2458 filecount value is greater than 2G or if a quota value is greater than 2G
2459 on a system where the size of off_t is not greater than 4, a panic error
2462 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2463 never match. The debug and -bh output now contains an explicit error
2464 message indicating a malformed IPv4 address or mask.
2466 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2467 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2468 PH/23 above applies.
2470 PH/25 Do not write to syslog when running in the test harness. The only
2471 occasion when this arises is a failure to open the main or panic logs
2472 (for which there is an explicit test).
2474 PH/26 Added the /no_tell option to "control=freeze".
2476 PH/27 If a host name lookup failed very early in a connection, for example, if
2477 the IP address matched host_lookup and the reverse lookup yielded a name
2478 that did not have a forward lookup, an error message of the form "no IP
2479 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2480 could be logged. Now it outputs the IP address instead of "NULL".
2482 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2483 allows the sender and the authenticated sender to be set when
2484 submitting a message from within Exim. Since child_open_exim() is
2485 documented for local_scan(), the new function should be too.
2487 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2488 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2489 results in an empty string is now treated as unset.
2491 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2493 PH/31 Added sender_verify_fail logging option.
2495 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2496 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2497 tidied the source and removed it altogether.
2499 PH/33 When a queue run was abandoned because the load average was too high, a
2500 log line was always written; now it is written only if the queue_run log
2501 selector is set. In addition, the log line for abandonment now contains
2502 information about the queue run such as the pid. This is always present
2503 in "start" and "stop" lines but was omitted from the "abandon" line.
2505 PH/34 Omit spaces between a header name and the colon in the error message that
2506 is given when verify = headers_syntax fails (if there are lots of them,
2507 the message gets confusing).
2509 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2510 names, as there are now some PTR records that contain slashes. This check
2511 is only to protect against broken name servers that fall over on strange
2512 characters, so the fact that it applies to all lookups doesn't matter.
2514 PH/36 Now that the new test suite is complete, we can remove some of the
2515 special code in Exim that was needed for the old test suite. For example,
2516 sorting DNS records because real resolvers return them in an arbitrary
2517 order. The new test suite's fake resolver always returns records in the
2520 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2521 bounces) except when queue_only is set, to avoid logging races between
2522 the different processes.
2524 PH/38 Panic-die if .include specifies a non-absolute path.
2526 PH/39 A tweak to the "H" retry rule from its user.
2528 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2529 a label. They prevented compilation on older perls.
2531 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2532 a warning to be raised on newish perls.
2534 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2535 on queue. Changes to match documented behaviour of showing count of
2536 messages matching specified criteria.
2538 PH/40 Changed the default ident timeout from 30s to 5s.
2540 PH/41 Added support for the use of login_cap features, on those BSD systems
2541 that have them, for controlling the resources used by pipe deliveries.
2543 PH/42 The content-scanning code uses fopen() to create files in which to put
2544 message data. Previously it was not paying any attention to the mode of
2545 the files. Exim runs with umask(0) because the rest of the code creates
2546 files with open(), and sets the required mode explicitly. Thus, these
2547 files were ending up world-writeable. This was not a big issue, because,
2548 being within the spool directory, they were not world-accessible. I have
2549 created a function called modefopen, which takes an additional mode
2550 argument. It sets umask(777), creates the file, chmods it to the required
2551 mode, then resets the umask. All the relevant calls to fopen() in the
2552 content scanning code have been changed to use this function.
2554 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2555 to 24 hours. This avoids potential overflow problems when processing G
2556 and H retry rules. I suspect nobody ever tinkers with this value.
2558 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2560 PH/45 When the plaintext authenticator is running as a client, the server's
2561 challenges are checked to ensure they are valid base64 strings. By
2562 default, the authentication attempt is cancelled if an invalid string is
2563 received. Setting client_ignore_invalid_base64 true ignores these errors.
2564 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2565 they are received. Thus, the responses can be made to depend on the
2566 challenges. If an invalid string is ignored, an empty string is placed in
2569 PH/46 Messages that are created by the autoreply transport now contains a
2570 References: header, in accordance with RFCs 2822 and 3834.
2572 PH/47 Added authenticated_sender_force to the smtp transport.
2574 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2576 PH/49 Installed latest patch from the Sieve maintainer.
2578 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2579 was also set, the appendfile transport was unnecessarily scanning a
2580 directory of message files (e.g. for maildir delivery) to find the count
2581 of files (along with the size), even though it did not need this
2582 information. It now does the scan only if it needs to find either the
2583 size of the count of files.
2585 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2587 PH/52 Two bugs concerned with error handling when the smtp transport is
2590 (i) Exim was not creating retry information for temporary errors given
2591 for individual recipients after the DATA command when the smtp transport
2592 was used in LMTP mode. This meant that they could be retried too
2593 frequently, and not timed out correctly.
2595 (ii) Exim was setting the flag that allows error details to be returned
2596 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2597 recipients that were returned after the DATA command.
2599 PH/53 This is related to PH/52, but is more general: for any failing address,
2600 when detailed error information was permitted to be returned to the
2601 sender, but the error was temporary, then after the final timeout, only
2602 "retry timeout exceeded" was returned. Now it returns the full error as
2603 well as "retry timeout exceeded".
2605 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2606 do this, and (what is worse) MTAs that accept it.
2608 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2609 will now be deprecated.
2611 PH/56 New os.c-cygwin from the Cygwin maintainer.
2613 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2614 formats (previously only available in exim formats via -bpr, -bpru,
2615 and -bpra. Now also available in native and exiqgrep formats)
2617 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2618 with very large, slow to parse queues
2620 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2622 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2624 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2625 responses to authentication challenges, though it was showing the
2626 challenges; (ii) I've removed the CR characters from the debug output for
2629 PH/58 Allow for the insertion of a newline as well as a space when a string
2630 is turned into more than one encoded-word during RFC 2047 encoding. The
2631 Sieve code now uses this.
2633 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2634 data_4xx, lost_connection, tls_required.
2636 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2637 message was being sent as an SMTP response.
2639 PH/61 Add -l and -k options to exicyclog.
2641 PH/62 When verifying, if an address was redirected to one new address, so that
2642 verification continued, and the new address failed or deferred after
2643 having set something in $address_data, the value of $address_data was not
2644 passed back to the ACL. This was different to the case when no
2645 redirection occurred. The value is now passed back in both cases.
2647 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2648 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2649 use login_cap.h, so on its own it isn't the distinguishing feature. The
2650 new name refers directly to the setclassresources() function.
2652 PH/65 Added configuration files for NetBSD3.
2654 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2656 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2657 is preferred over IPv4.
2659 PH/68 The bounce_return_message and bounce_return_body options were not being
2660 honoured for bounces generated during the reception of non-SMTP messages.
2661 In particular, this applied to messages rejected by the ACL. This bug has
2662 been fixed. However, if bounce_return_message is true and bounce_return_
2663 body is false, the headers that are returned for a non-SMTP message
2664 include only those that have been read before the error was detected.
2665 (In the case of an ACL rejection, they have all been read.)
2667 PH/69 The HTML version of the specification is now built in a directory called
2668 spec_html instead of spec.html, because the latter looks like a path with
2669 a MIME-type, and this confuses some software.
2671 PH/70 Catch two compiler warnings in sieve.c.
2673 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2674 function verify_get_ident() calls ip_connect() to connect a socket, but
2675 if the "connect()" function timed out, ip_connect() used to close the
2676 socket. However, verify_get_ident() also closes the socket later, and in
2677 between Exim writes to the log, which may get opened at this point. When
2678 the socket was closed in ip_connect(), the log could get the same file
2679 descriptor number as the socket. This naturally causes chaos. The fix is
2680 not to close the socket in ip_connect(); the socket should be closed by
2681 the function that creates it. There was only one place in the code where
2682 this was missing, in the iplookup router, which I don't think anybody now
2683 uses, but I've fixed it anyway.
2685 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2686 well as to direct DNS lookups. Otherwise the handling of names in host
2687 lists is inconsistent and therefore confusing.
2693 PH/01 Two changes to the default runtime configuration:
2695 (1) Move the checks for relay_from_hosts and authenticated clients from
2696 after to before the (commented out) DNS black list checks.
2698 (2) Add control=submission to the relay_from_hosts and authenticated
2699 clients checks, on the grounds that messages accepted by these
2700 statements are most likely to be submissions.
2702 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2704 (1) Generate an error if the third argument for the ${prvs expansion is
2707 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2710 (3) Reset the variables that are obtained from the first argument of
2711 ${prvscheck and used in the second argument before leaving the code,
2712 because their memory is reclaimed, so using them afterwards may do
2715 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2716 one (it's much easier than Tom thought :-).
2718 (5) Because of (4), we can now allow for the use of $prvscheck_result
2719 inside the third argument.
2721 PH/03 For some reason, the default setting of PATH when running a command from
2722 a pipe transport was just "/usr/bin". I have changed it to
2725 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2726 anything to be listed in the output from -bV.
2728 PH/05 When a filter generated an autoreply, the entire To: header line was
2729 quoted in the delivery log line, like this:
2731 => >A.N.Other <ano@some.domain> <original@ddress> ...
2733 This has been changed so that it extracts the operative address. There
2734 may be more than one such address. If so, they are comma-separated, like
2737 => >ano@some.domain,ona@other.domain <original@ddress> ...
2739 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2740 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2741 not being looked up in the rDNS to get a host name, Exim was showing the
2742 IP address twice in Received: lines, even though the IP addresses were
2743 identical. For example:
2745 Received: from [1.2.3.4] (helo=[1.2.3.4])
2747 However, if the real host name was known, it was omitting the HELO data
2748 if it matched the actual IP address. This has been tidied up so that it
2749 doesn't show the same IP address twice.
2751 PH/07 When both +timestamp and +memory debugging was on, the value given by
2752 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2753 called by the debug printing, thereby overwriting the timestamp buffer.
2754 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2756 PH/08 When the original message was included in an autoreply transport, it
2757 always said "this is a copy of the message, including all the headers",
2758 even if body_only or headers_only was set. It now gives an appropriate
2761 PH/09 Applied a patch from the Sieve maintainer which:
2763 o fixes some comments
2764 o adds the (disabled) notify extension core
2765 o adds some debug output for the result of if/elsif tests
2766 o points to the current vacation draft in the documentation
2767 and documents the missing references header update
2771 o fixes a bug in processing the envelope test (when testing
2772 multiple envelope elements, the last element determinted the
2775 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2776 Electronic Mail") by including:
2778 Auto-submitted: auto-generated
2780 in the messages that it generates (bounce messages and others, such as
2781 warnings). In the case of bounce messages for non-SMTP mesages, there was
2782 also a typo: it was using "Auto_submitted" (underscore instead of
2783 hyphen). Since every message generated by Exim is necessarily in response
2784 to another message, thes have all been changed to:
2786 Auto-Submitted: auto-replied
2788 in accordance with these statements in the RFC:
2790 The auto-replied keyword:
2792 - SHOULD be used on messages sent in direct response to another
2793 message by an automatic process,
2795 - MUST NOT be used on manually-generated messages,
2797 - MAY be used on Delivery Status Notifications (DSNs) and Message
2798 Disposition Notifications (MDNs),
2800 - MUST NOT be used on messages generated by automatic or periodic
2801 processes, except for messages which are automatic responses to
2804 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2805 to the default Received: header definition.
2807 PH/12 Added log selector acl_warn_skipped (default on).
2809 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2810 variables because (a) they are in the wrong storage pool and (b) even if
2811 they were copied, it wouldn't work properly because of the caching.
2813 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2814 checking when decoding. Apparently there are clients that generate
2815 overlong encoded strings. Why am I not surprised?
2817 PH/15 If the first argument of "${if match_address" was not empty, but did not
2818 contain an "@" character, Exim crashed. Now it writes a panic log message
2819 and treats the condition as false.
2821 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2823 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2824 extension "envelope-auth". The code is finished and in agreement with
2825 other implementations, but there is no documentation so far and in fact,
2826 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2827 not changing the active code.
2829 Print executed "if" and "elsif" statements when debugging is used. This
2830 helps a great deal to understand what a filter does.
2832 Document more things not specified clearly in RFC3028. I had all this
2833 sorted out, when out of a sudden new issues came to my mind. Oops."
2835 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2838 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2839 canonical form (as documented). However, after a host name lookup from
2840 the IP address, check_host() was doing a simple string comparison with
2841 addresses acquired from the DNS when checking that the found name did
2842 have the original IP as one of its addresses. Since any found IPv6
2843 addresses are likely to be in abbreviated form, the comparison could
2844 fail. Luckily, there already exists a function for doing the comparison
2845 by converting both addresses to binary, so now that is used instead of
2846 the text comparison.
2848 PH/20 There was another similar case to PH/19, when a complete host name was
2849 given in a host list; looking up its IP address could give an abbreviated
2850 form, whereas the current host's name might or might not be abbreviated.
2851 The same fix has been applied.
2857 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2858 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2861 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2862 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2864 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2866 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2867 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2868 Also, for "Auto-Submitted", treat anything other than "no" as
2869 non-personal, in accordance with RFC 3834. (Previously it treated
2870 anything starting "auto-" as non-personal.)
2872 TF/01 The control=submission/name=... option had a problem with syntax
2873 errors if the name included a slash character. The /name= option
2874 now slurps the rest of the string, so it can include any characters
2875 but it must come last in the list of options (after /sender_retain
2878 PH/05 Some modifications to the interface to the fake nameserver for the new
2886 TK/01 Added the "success_on_redirect" address verification option. See
2887 NewStuff for rationale and an example.
2889 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2891 PH/02 Patch to exigrep to allow it to work on syslog lines.
2893 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2894 fread() to skip over the body file's header line, because in Cygwin the
2895 header line is locked and is inaccessible.
2897 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2898 co-exist for some time) to make it clear that it is the Exim ID that is
2899 referenced, not the Message-ID: header line.
2901 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2902 string_format() function, because snprintf() does not exist on all
2905 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2906 ${stat: expansion item.
2908 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2909 protocol synchronization error", to keep the pedants happy.
2911 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2912 well as for IRIX systems, when gcc is being used. See the host.c source
2915 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2917 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2920 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2921 transport and to the smtp transport in LMTP mode.
2923 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2925 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2926 run a filter in a subprocess. This could lead to confusion in subsequent
2927 lookups in the parent process. There should also be a search_tidyup() at
2928 the end of the subprocess.
2930 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2931 only if the host matched helo_try_verify_hosts, which caused the
2932 verification to occur when the EHLO/HELO command was issued. The ACL just
2933 tested the remembered result. Now, if a previous verification attempt has
2934 not happened, "verify = helo" does it there and then.
2936 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2938 TK/03 Fix log output including CR from clamd.
2940 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2941 exist provoked a memory error which could cause a segfault.
2943 PH/15 Installed PCRE 6.2
2945 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2947 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2948 of the problem. Specifically, suggested +O2 rather than +O1 for the
2951 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2953 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2954 record, the port was not correctly logged when the outgoing_port log
2955 selector was set (it logged the transort's default port).
2957 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2958 fallback_hosts, and "hosts" in the smtp transport.
2960 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2961 host errors such as "Connection refused".
2963 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2964 authentication with radiusclient 0.4.9:
2966 - Error returned from rc_read_config was caught wrongly
2967 - Username/password not passed on to radius server due to wrong length.
2969 The presumption is that some radiusclient API changes for 4.51/PH/17
2970 were not taken care of correctly. The code is still untested by me (my
2971 Linux distribution still has 0.3.2 of radiusclient), but it was
2972 contributed by a Radius user.
2974 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2975 expanding the "port" option of the smtp transport.
2977 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2978 while reading a MIME header. Thanks to Tom Hughes for a patch.
2980 PH/24 Include config.h inside local_scan.h so that configuration settings are
2983 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2984 that in an ACL for RCPT (for example), you can examine exactly what was
2987 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2988 commands, but it was not correctly comparing the address with the actual
2989 client host address. Thus, it would show the EHLO address in Received:
2990 header lines when this was not necessary.
2992 PH/27 Added the % operator to ${eval:}.
2994 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2995 it should be ignoring failures (because with -C, for example, it has lost
2996 privilege). It wasn't ignoring creation failures other than "already
2999 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3002 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3003 because an input line was too long, either on its own, or by virtue of
3004 too many continuations, the temporary file was not being removed, and the
3005 return code was incorrect.
3007 PH/31 Missing "BOOL" in function definition in filtertest.c.
3009 PH/32 Applied Sieve patches from the maintainer.
3011 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3013 PH/33 Added "verify = not_blind".
3015 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3016 Local/Makefile (with some defaults set). These are used in built scripts
3017 such as exicyclog, but they have never been used in the exim_install
3018 script (though there are many overriding facilities there). I have
3019 arranged that the exim_install script now takes note of these two
3022 PH/35 Installed configuration files for Dragonfly.
3024 PH/36 When a locally submitted message by a trusted user did not contain a
3025 From: header, and the sender address was obtained from -f or from an SMTP
3026 MAIL command, and the trusted user did not use -F to supply a sender
3027 name, $originator_name was incorrectly used when constructing a From:
3028 header. Furthermore, $originator_name was used for submission mode
3029 messages from external hosts without From: headers in a similar way,
3030 which is clearly wrong.
3032 PH/37 Added control=suppress_local_fixups.
3034 PH/38 When log_selector = +received_sender was set, and the addition of the
3035 sender made the log line's construction buffer exactly full, or one byte
3036 less than full, an overflow happened when the terminating "\n" was
3039 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3040 when the result of a list match is failure because a DNS lookup failed.
3042 PH/40 RM_COMMAND is now used in the building process.
3044 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3045 the "build-* directories that it finds.
3047 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3048 address in a domain literal was a prefix of an interface address.
3050 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3051 when verifying a sender address, unless rewrite_headers is false.
3053 PH/44 Wrote a long comment about why errors_to addresses are verified as
3054 recipients, not senders.
3056 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3057 the ratelimit ACL was added.
3059 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3061 PH/46 Added extra information about PostgreSQL errors to the error string.
3063 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3064 avoiding the need to install special zones in a real server. This is
3065 backwards compatible; if it can't find the fake resolver, it drops back.
3066 Thus, both old and new test suites can be run.
3068 TF/02 Added util/ratelimit.pl
3070 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3071 clock is set back in time.
3073 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3074 Candler <B.Candler@pobox.com>.
3076 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3077 OK for addresses that are the result of redirections.
3079 PH/48 A number of further additions for the benefit of the new test suite,
3080 including a fake gethostbyname() that interfaces to the fake DNS resolver
3083 TF/06 The fix for widen_domains has also been applied to qualify_single and
3084 search_parents which are the other dnslookup options that can cause
3087 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3090 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3092 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3093 service" if it encountered an MX record with an empty target hostname.
3094 The message is now "an MX or SRV record indicated no SMTP service".
3096 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3097 if the DNS of the sending site is misconfigured. This is quite a
3098 common situation. This change restores the behaviour of treating a
3099 helo verification defer as a failure.
3101 PH/51 If self=fail was set on a router, the bounce message did not include the
3102 actual error message.
3108 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3110 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3111 command itself ended in error, the underlying message about the transport
3112 filter timeout was being overwritten with the pipe command error. Now the
3113 underlying error message should be appended to the second error message.
3115 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3117 PH/02 Reduce the amount of output that "make" produces by default. Full output
3118 can still be requested.
3120 PH/03 The warning log line about a condition test deferring for a "warn" verb
3121 was being output only once per connection, rather than after each
3122 occurrence (because it was using the same function as for successful
3123 "warn" verbs). This seems wrong, so I have changed it.
3125 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3126 it should not have, which might have caused a crash in the right
3127 circumstances, but probably never did.
3129 PH/04 Installed a modified version of Tony Finch's patch to make submission
3130 mode fix the return path as well as the Sender: header line, and to
3131 add a /name= option so that you can make the user's friendly name appear
3134 TF/03 Added the control = fakedefer ACL modifier.
3136 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3137 Mark Lowes for thorough testing.
3139 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3141 TK/03 Merged latest SRS patch from Miles Wilton.
3143 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3144 with the definition in sysexits.h (which is #included earlier).
3145 Fortunately, Exim does not actually use EX_OK. The code used to try to
3146 preserve the sysexits.h value, by assumimg that macro definitions were
3147 scanned for macro replacements. I have been disabused of this notion,
3148 so now the code just undefines EX_OK before #including unistd.h.
3150 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3151 in the smtp transport. When a block could not be written in a single
3152 write() function, the timeout was being re-applied to each part-write.
3153 This seems wrong - if the receiver was accepting one byte at a time it
3154 would take for ever. The timeout is now adjusted when this happens. It
3155 doesn't have to be particularly precise.
3157 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3158 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3160 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3161 without a domain if the check to <postmaster@domain> fails.
3163 SC/01 Eximstats: added -xls and the ability to specify output files
3164 (patch written by Frank Heydlauf).
3166 SC/02 Eximstats: use FileHandles for outputing results.
3168 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3170 SC/04 Eximstats: fixed display of large numbers with -nvr option
3172 SC/05 Eximstats: fixed merging of reports with empty tables.
3174 SC/06 Eximstats: added the -include_original_destination flag
3176 SC/07 Eximstats: removed tabs and trailing whitespace.
3178 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3180 TK/06 MBOX spool code: Add real "From " MBOX separator line
3181 so the .eml file is really in mbox format (even though
3182 most programs do not really care). Patch from Alex Miller.
3184 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3185 The latter is generated from $received_to and is only set if the
3186 message has one envelope recipient. SA can use these headers,
3187 obviously out-of-the-box. Patch from Alex Miller.
3189 PH/08 The ${def test on a variable was returning false if the variable's
3190 value was "0", contrary to what the specification has always said!
3191 The result should be true unless the variable is empty.
3193 PH/09 The syntax error of a character other than { following "${if
3194 def:variable_name" (after optional whitespace) was not being diagnosed.
3195 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3196 accidental colon was present, for example, could give incorrect results.
3198 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3199 result is used (not including appendfile, where other changes are about
3202 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3203 This involved changing a lot of size variables from int to off_t. It
3204 should work with maildirs and everything.
3206 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3207 spamd dying while we are connected to it.
3209 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3212 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3213 function for BDB 4.3.
3215 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3217 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3218 the sources for PCRE have been re-arranged and more files are now
3221 PH/15 The code I had for printing potentially long long variables in PH/11
3222 above was not the best (it lost precision). The length of off_t variables
3223 is now inspected at build time, and an appropriate printing format (%ld
3224 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3225 to be "long long int" or "long int". This is needed for the internal
3226 formatting function string_vformat().
3228 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3229 the configuration file to be ":syslog", then the script "guesses" where
3230 the logs files are, rather than using the compiled in default. In our
3231 case the guess is not the same as the compiled default, so the script
3232 suddenly stopped working when I started to use syslog. The patch checks
3233 to see if log_file_path is "". If so, it attempts to read it from exim
3234 with no configuration file to get the compiled in version, before it
3235 falls back to the previous guessing code."
3237 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3238 implementing BATV in an Exim configuration. See NewStuff for the gory
3241 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3242 Makefile that are specific to HP-UX.
3244 PH/18 If the "use_postmaster" option was set for a recipient callout together
3245 with the "random" option, the postmaster address was used as the MAIL
3246 FROM address for the random test, but not for the subsequent recipient
3247 test. It is now used for both.
3249 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3250 patch removes a few documentation additions to RFC 3028, because the
3251 latest draft now contains them. It adds the new en;ascii-case comparator
3252 and a new error check for 8bit text in MIME parts. Comparator and
3253 require names are now matched exactly. I enabled the subaddress
3254 extension, but it is not well tested yet (read: it works for me)."
3256 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3257 rework some of the code of TK/09 above to avoid the hardwired use of
3258 "%lld" and "long long". Replaced the call to snprintf() with a call to
3261 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3262 records point to non-existent hosts", "retry timeout exceeded", and
3263 "retry time not reached for any host after a long failure period".
3265 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3266 experimental DomainKeys support:
3268 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3269 (2) On an error such as an illegally used "control", the wrong name for
3270 the control was given.
3272 These problems did NOT occur unless DomainKeys support was compiled.
3274 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3276 PH/24 Added ${if match_ip condition.
3278 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3279 databases so that it will be absolutely obvious if a crash occurs in the
3280 DB library. This is a regular occurrence (often caused by mis-matched
3283 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3284 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3285 picked up on a user's system that detects such things. There doesn't seem
3286 to be a gcc warning option for this - only an attribute that has to be
3287 put on the function's prototype. It seems that in Fedora Core 4 they have
3288 set this on a number of new functions. No doubt there will be more in due
3291 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3292 specify a transport. However, if an address that was verified by such a
3293 router was the subject of a callout, Exim crashed because it tried to
3294 read the rcpt_include_affixes from the non-existent transport. Now it
3295 just assumes that the setting of that option is false. This bug was
3296 introduced by 4.51/PH/31.
3298 PH/28 Changed -d+all to exclude +memory, because that information is very
3299 rarely of interest, but it makes the output a lot bigger. People tend to
3300 do -d+all out of habit.
3302 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3303 code in os-type was giving problems when libc.so lives in lib64, like on
3306 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3307 aren't the modern standard, and it seems that some systems' include files
3308 don't always have them. Exim was already checking for some of the newer
3309 ones like T_AAAA, and defining it itself. I've added checks for all the
3310 record types that Exim uses.
3312 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3313 not automatically generating a new one, as it is supposed to. This
3314 prevented TLS from working. If the file did exist, but contained invalid
3315 data, a new version was generated, as expected. It was only the case of a
3316 non-existent file that was broken.
3318 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3319 with a change in libdomainkeys > 0.64.
3321 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3322 from DNS. If the selector record carries the flag, it now has
3323 precedence over the domain-wide flag.
3325 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3327 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3328 the use of a port name that isn't defined in /etc/services) occurred, the
3329 message was deferred as in a normal delivery, and thus remained on the
3330 spool, instead of being failed because of the mua_wrapper setting. This
3331 is now fixed, and I tidied up some of the mua_wrapper messages at the
3334 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3335 the messages in a hash of arrays rather than using individual hashes.
3336 This is a bit cleaner and results in dramatic memory savings, albeit
3337 at a slight CPU cost.
3339 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3340 as requested by Marc Sherman.
3342 SC/10 Eximstats: added histograms for user specified patterns as requested
3345 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3347 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3348 fopen() in the content-scanning modules that did not already have it.
3354 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3355 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3357 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3359 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3361 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3362 to test Sieve filters that use "vacation".
3364 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3365 that changes the way the GnuTLS parameters are stored in the cache file.
3366 The new format can be generated externally. For backward compatibility,
3367 if the data in the cache doesn't make sense, Exim assumes it has read an
3368 old-format file, and it generates new data and writes a new file. This
3369 means that you can't go back to an older release without removing the
3372 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3373 work if there are any delivery delays because "one_time" forces the
3374 parent to be marked "delivered", so its unseen clone is never tried
3375 again. For this reason, Exim now forbids the simultaneous setting of
3378 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3379 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3380 entry if you want to know the details. The fix, however, neglected to
3381 consider the case when local delivery batching is involved. The test for
3382 "previously delivered" was not happening when checking to see if an
3383 address could be batched with a previous (undelivered) one; under
3384 certain circumstances this could lead to multiple deliveries to the same
3387 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3388 in its include files, and this causes problems building Exim.
3390 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3391 header_syntax) but Exim was just ignoring anything given after a slash.
3392 In particular, this caused confusion with an attempt to use "verify =
3393 reverse_host_lookup/defer_ok". An error is now given when options are
3394 supplied for verify items that do not have them. (Maybe reverse_host_
3395 lookup should have a defer_ok option, but that's a different point.)
3397 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3398 defined by RFC 821) to 2048, because there were problems with some AUTH
3399 commands, and RFC 1869 says the size should be increased for extended
3400 SMTP commands that take arguments.
3402 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3405 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3406 "unknown" error; now it says that the functionality isn't in the binary.
3408 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3409 an address' error message when a string expansion fails (syntax or
3410 whatever). Otherwise the password may appear in the log. Following change
3411 PH/42 below, there is no longer a chance of it appearing in a bounce
3414 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3416 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3417 of its addresses was ever tried. (Bugzilla bug #2.)
3419 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3420 the result incorrectly in the debug output. (It correctly added a newline
3421 to what was transported.)
3423 TF/01 Added $received_time.
3425 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3426 commented out examples of how to interface to a virus scanner and to
3427 SpamAssassin. Also added commented examples of av_scanner and
3428 spamd_address settings.
3430 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3431 and controls are allowed in which ACLs. There were a couple of minor
3432 errors. Some of the entries in the conditions table (which is a table of
3433 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3434 negation of where the condition IS allowed.
3436 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3438 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3439 header file does not have a version number, so I've had to invent a new
3440 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3441 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3442 radiusclient), but it was contributed by a Radius user.
3444 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3445 files or directories, for OpenSSL.
3447 PH/19 When an Exim process that is running as root has to create an Exim log
3448 file, it does so in a subprocess that runs as exim:exim so as to get the
3449 ownership right at creation (otherwise, other Exim processes might see
3450 the file with the wrong ownership). There was no test for failure of this
3451 fork() call, which would lead to the process getting stuck as it waited
3452 for a non-existent subprocess. Forks do occasionally fail when resources
3453 run out. I reviewed all the other calls to fork(); they all seem to check
3456 PH/20 When checking for unexpected SMTP input at connect time (before writing
3457 the banner), Exim was not dealing correctly with a non-positive return
3458 from the read() function. If the client had disconnected by this time,
3459 the result was a log entry for a synchronization error with an empty
3460 string after "input=" when read() returned zero. If read() returned -1
3461 (an event I could not check), uninitialized data bytes were printed.
3462 There were reports of junk text (parts of files, etc) appearing after
3465 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3467 PH/22 Added support for macro redefinition, and (re)definition in between
3468 driver and ACL definitions.
3470 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3471 forgetting to use the resulting value; it was using the unexpanded value.
3473 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3474 hadn't been configured. The fix is from Juergen Kreileder, who
3475 understands it better than I do:
3477 "Here's what I see happening with three configured cyrus_sasl
3478 authenticators configured (plain, login, cram-md5):
3480 On startup auth_cyrus_sasl_init() gets called for each of these.
3481 This means three calls to sasl_listmech() without a specified mech_list.
3482 => SASL tests which mechs of all available mechs actually work
3483 => three warnings about OTP not working
3484 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3486 With the patch, sasl_listmech() also gets called three times. But now
3487 SASL's mech_list option is set to the server_mech specified in the the
3488 authenticator. Or in other words, the answer from sasl_listmech()
3489 gets limited to just the mech you're testing for (which is different
3491 => the return list contains just 'plain' or 'login', 'cram-md5' or
3492 nothing depending on the value of ob->server_mech.
3494 I've just tested the patch: Authentication still works fine,
3495 unavailable mechs specified in the exim configuration are still
3496 caught, and the auth.log warnings about OTP are gone."
3498 PH/25 When debugging is enabled, the contents of the command line are added
3499 to the debugging output, even when log_selector=+arguments is not
3502 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3503 answer is "GNU", and only if the return is "GNU/something" is the answer
3506 PH/27 $acl_verify_message is now set immediately after the failure of a
3507 verification in an ACL, and so is available in subsequent modifiers. In
3508 particular, the message can be preserved by coding like this:
3510 warn !verify = sender
3511 set acl_m0 = $acl_verify_message
3513 Previously, $acl_verify_message was set only while expanding "message"
3514 and "log_message" when a very denied access.
3516 PH/28 Modified OS/os.c-Linux with
3518 -#ifndef OS_LOAD_AVERAGE
3519 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3521 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3522 nomenclature these days.)
3524 PH/29 Installed patch from the Sieve maintainer that adds the options
3525 sieve_useraddress and sieve_subaddress to the redirect router.
3527 PH/30 In these circumstances:
3528 . Two addresses routed to the same list of hosts;
3529 . First host does not offer TLS;
3530 . First host accepts first address;
3531 . First host gives temporary error to second address;
3532 . Second host offers TLS and a TLS session is established;
3533 . Second host accepts second address.
3534 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3535 and peerdn, if requested) that were in fact used only for the second
3538 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3539 attention to any local part prefix or suffix that was matched by the
3540 router that accepted the address. It now behaves in the same way as it
3541 does for delivery: the affixes are removed from the local part unless
3542 rcpt_include_affixes is set on the transport.
3544 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3545 timeout during the DATA phase of an incoming message.
3547 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3548 applied a patch sanctioned by the Sieve maintainer.
3550 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3551 the uid or gid is negative. A case of a negative gid caused this to be
3552 noticed. The fix allows for either to be negative.
3554 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3555 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3558 PH/36 The change PH/12 above was broken. Fixed it.
3560 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3561 the grounds that routing the same address twice would always produce the
3562 same answer. This might have been true once, but it is certainly no
3563 longer true now. Routing a child address may depend on the previous
3564 routing that produced that child. Some complicated redirection strategies
3565 went wrong when messages had multiple recipients, and made Exim's
3566 behaviour dependent on the order in which the addresses were given.
3568 I have moved the duplicate checking until after the routing is complete.
3569 Exim scans the addresses that are assigned to local and remote
3570 transports, and removes any duplicates. This means that more work will be
3571 done, as duplicates will always all be routed, but duplicates are
3572 presumably rare, so I don't expect this is of any significance.
3574 For deliveries to pipes, files, and autoreplies, the duplicate checking
3575 still happens during the routing process, since they are not going to be
3578 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3579 It corrects a timeout issue with spamd. This is Ian's comment: "The
3580 background is that sometimes spamd either never reads data from a
3581 connection it has accepted, or it never writes response data. The exiscan
3582 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3583 blindly assumes that writes won't block so it may never time out."
3585 PH/39 Allow G after quota size as well as K and M.
3587 PH/40 The value set for $authenticated_id in an authenticator may not contain
3588 binary zeroes or newlines because the value is written to log lines and
3589 to spool files. There was no check on this. Now the value is run through
3590 the string_printing() function so that such characters are converted to
3591 printable escape sequences.
3593 PH/41 $message_linecount is a new variable that contains the total number of
3594 lines in the message. Compare $body_linecount, which is the count for the
3597 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3598 bounce and delay warning messages, except in certain special cases, which
3601 (a) An SMTP error message from a remote host;
3602 (b) A message specified in a :fail: redirection;
3603 (c) A message specified in a "fail" command in a system filter;
3604 (d) A message specified in a FAIL return from the queryprogram router;
3605 (e) A message specified by the cannot_route_message router option.
3607 In these cases only, Exim does include the error details in bounce and
3608 warning messages. There are also a few cases where bland messages such
3609 as "unrouteable address" or "local delivery error" are given.
3611 PH/43 $value is now also set for the "else" part of a ${run expansion.
3613 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3614 being worked on, but at least Exim now implements the latest version to
3617 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3618 process to complete was treated as a delivery failure, a timeout while
3619 writing the message to the pipe was logged, but erroneously treated as a
3620 successful delivery. Such timeouts include transport filter timeouts. For
3621 consistency with the overall process timeout, these timeouts are now
3622 treated as errors, giving rise to delivery failures by default. However,
3623 there is now a new Boolean option for the pipe transport called
3624 timeout_defer, which, if set TRUE, converts the failures into defers for
3625 both kinds of timeout. A transport filter timeout is now identified in
3628 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3629 systems where "make" and "gmake" are different, calling "gmake" at top
3630 level broke things. I've arranged for the value of $(MAKE) to be passed
3631 from the Makefile to this script so that it can call the same version of
3635 A note about Exim versions 4.44 and 4.50
3636 ----------------------------------------
3638 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3639 changes of various kinds. As a consequence, a big documentation update was
3640 needed. This delayed the release for rather longer than seemed good, especially
3641 in the light of a couple of (minor) security issues. Therefore, the changes
3642 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3643 release. So 4.44 and 4.50 are in effect two different branches that both start
3646 I have left the 4.50 change log unchanged; it contains all the changes since
3647 4.43. The change log for 4.44 is below; many of its items are identical to
3648 those for 4.50. This seems to be the most sensible way to preserve the
3649 historical information.
3655 1. Minor wording change to the doc/README.SIEVE file.
3657 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3658 computation of the current number of files was incorrect.
3660 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3661 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3664 4. Give more explanation in the error message when the command for a transport
3665 filter fails to execute.
3667 5. There are several places where Exim runs a non-Exim command in a
3668 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3669 was being done only for the command run by the queryprogram router. It is
3670 now done for all such subprocesses. The other cases are: ${run, transport
3671 filters, and the commands run by the lmtp and pipe transports.
3673 6. Added CONFIGURE_GROUP build-time option.
3675 7. Some older OS have a limit of 256 on the maximum number of file
3676 descriptors. Exim was using setrlimit() to set 1000 as a large value
3677 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3678 systems. I've change it so that if it can't get 1000, it tries for 256.
3680 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3681 was an oversight, and furthermore, ever since the addition of extra
3682 controls (e.g. 4.43/32), the checks on when to allow different forms of
3683 "control" were broken. There should now be diagnostics for all cases when a
3684 control that does not make sense is encountered.
3686 9. Added the /retain_sender option to "control=submission".
3688 10. $recipients is now available in the predata ACL (oversight).
3690 11. Tidy the search cache before the fork to do a delivery from a message
3691 received from the command line. Otherwise the child will trigger a lookup
3692 failure and thereby defer the delivery if it tries to use (for example) a
3693 cached ldap connection that the parent has called unbind on.
3695 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3696 of $address_data from the recipient verification was clobbered by the
3697 sender verification.
3699 13. The value of address_data from a sender verification is now available in
3700 $sender_address_data in subsequent conditions in the ACL statement.
3702 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3704 15. Added a new option "connect=<time>" to callout options, to set a different
3707 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3708 was its contents. (It was OK if the option was not defined at all.)
3710 17. A "Completed" log line is now written for messages that are removed from
3711 the spool by the -Mrm option.
3713 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3714 information about exactly what failed.
3716 19. Added -dd to debug only the daemon process.
3718 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3719 handles timeouts, both on the server side and network timeouts. Renamed the
3720 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3722 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3723 It is now set to "smtps".
3725 22. $host_address is now set to the target address during the checking of
3726 ignore_target_hosts.
3728 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3729 being passed; this would have caused $sender_host_name to have been used if
3730 matching the list had actually called for a host name (not very likely,
3731 since this list is usually IP addresses). A host name is now passed as
3734 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3735 code that specifies a non-restarting handler (typically sigaction() in
3736 modern systems) in an attempt to fix a rare and obscure crash bug.
3738 25. Narrowed the window for a race in the daemon that could cause it to ignore
3739 SIGCHLD signals. This is not a major problem, because they are used only to
3740 wake it up if nothing else does.
3742 26. A malformed maildirsize file could cause Exim to calculate negative values
3743 for the mailbox size or file count. Odd effects could occur as a result.
3744 The maildirsize information is now recalculated if the size or filecount
3747 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3748 support for a long time. Removed HAVE_SYS_VFS_H.
3750 28. Installed the latest version of exipick from John Jetmore.
3752 29. In an address list, if the pattern was not a regular expression, an empty
3753 subject address (from a bounce message) matched only if the pattern was an
3754 empty string. Non-empty patterns were not even tested. This was the wrong
3755 because it is perfectly reasonable to use an empty address as part of a
3756 database query. An empty address is now tested by patterns that are
3757 lookups. However, all the other forms of pattern expect the subject to
3758 contain a local part and a domain, and therefore, for them, an empty
3759 address still always fails if the pattern is not itself empty.
3761 30. Exim went into a mad DNS loop when attempting to do a callout where the
3762 host was specified on an smtp transport, and looking it up yielded more
3763 than one IP address.
3765 31. Re-factored the code for checking spool and log partition space into a
3766 function that finds that data and another that does the check. The former
3767 is then used to implement four new variables: $spool_space, $log_space,
3768 $spool_inodes, and $log_inodes.
3770 32. The RFC2047 encoding function was originally intended for short strings
3771 such as real names; it was not keeping to the 75-character limit for
3772 encoded words that the RFC imposes. It now respects the limit, and
3773 generates multiple encoded words if necessary. To be on the safe side, I
3774 have increased the buffer size for the ${rfc2047: expansion operator from
3777 33. It is now permitted to omit both strings after an "if" condition; if the
3778 condition is true, the result is "true". As before, when the second string
3779 is omitted, a false condition yields an empty string. This makes it less
3780 cumbersome to write custom ACL and router conditions.
3782 34. Failure to deliver a bounce message always caused it to be frozen, even if
3783 there was an errors_to setting on the router. The errors_to setting is now
3786 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3787 canonical form (fully expanded) before being placed in
3788 $sender_host_address.
3790 36. The table in the code that translates DNS record types into text (T_A to
3791 "A" for instance) was missing entries for NS and CNAME. It is just possible
3792 that this could have caused confusion if both these types were looked up
3793 for the same domain, because the text type is used as part of Exim's
3794 per-process caching. But the chance of anyone hitting this buglet seems
3797 37. The dnsdb lookup has been extended in a number of ways.
3799 (1) There is a new type, "zns", which walks up the domain tree until it
3800 finds some nameserver records. It should be used with care.
3802 (2) There is a new type, "mxh", which is like "mx" except that it returns
3803 just the host names, not the priorities.
3805 (3) It is now possible to give a list of domains (or IP addresses) to be
3806 looked up. The behaviour when one of the lookups defers can be
3807 controlled by a keyword.
3809 (4) It is now possible to specify the separator character for use when
3810 multiple records are returned.
3812 38. The dnslists ACL condition has been extended: it is now possible to supply
3813 a list of IP addresses and/or domains to be looked up in a particular DNS
3816 39. Added log_selector=+queue_time_overall.
3818 40. When running the queue in the test harness, wait just a tad after forking a
3819 delivery process, to get repeatability of debugging output.
3821 41. Include certificate and key file names in error message when GnuTLS fails
3822 to set them up, because the GnuTLS error message doesn't include the name
3823 of the failing file when there is a problem reading it.
3825 42. Allow both -bf and -bF in the same test run.
3827 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3829 44. The "Exiscan patch" is now merged into the mainline Exim source.
3831 45. Sometimes the final signoff response after QUIT could fail to get
3832 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3833 < 0 before doing a fflush(). This bug looks as though it goes back to the
3834 introduction of TLS in release 3.20, but "sometimes" must have been rare
3835 because the tests only now provoked it.
3837 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3838 (this can affect the format of dates).
3840 47. exim_tidydb, when checking for the continued existence of a message for
3841 which it has found a message-specific retry record, was not finding
3842 messages that were in split spool directories. Consequently, it was
3843 deleting retry records that should have stayed in existence.
3845 48. Steve fixed some bugs in eximstats.
3847 49. The SPA authentication driver was not abandoning authentication and moving
3848 on to the next authenticator when an expansion was forced to fail,
3849 contradicting the general specification for all authenticators. Instead it
3850 was generating a temporary error. It now behaves as specified.
3852 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3853 (the order specifies the preference for clients). The order is now AES256,
3854 AES128, 3DES, ARCFOUR128.
3856 51. Small patch to Sieve code - explicitly set From: when generating an
3859 52. Exim crashed if a remote delivery caused a very long error message to be
3860 recorded - for instance if somebody sent an entire SpamAssassin report back
3861 as a large number of 550 error lines. This bug was coincidentally fixed by
3862 increasing the size of one of Exim's internal buffers (big_buffer) that
3863 happened as part of the Exiscan merge. However, to be on the safe side, I
3864 have made the code more robust (and fixed the comments that describe what
3867 53. Now that there can be additional text after "Completed" in log lines (if
3868 the queue_time_overall log selector is set), a one-byte patch to exigrep
3869 was needed to allow it to recognize "Completed" as not the last thing in
3872 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3873 patch that reportedly fixes this has been added. I am not expert enough to
3874 create a test for it. This is what the patch creator wrote:
3876 "I found a little strange behaviour of ldap code when working with
3877 Windows 2003 AD Domain, where users was placed in more than one
3878 Organization Units. When I tried to give exim partial DN, the exit code
3879 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3880 But simultaneously result of request was absolutely normal ldap result,
3881 so I produce this patch..."
3883 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3884 REFERENCE, so I have modified the code to exclude the patch when that macro
3887 55. Some experimental protocols are using DNS PTR records for new purposes. The
3888 keys for these records are domain names, not reversed IP addresses. The
3889 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3890 leaves it alone. Component reversal etc. now happens only for IP addresses.
3893 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3895 57. Double the size of the debug message buffer (to 2048) so that more of very
3896 long debug lines gets shown.
3898 58. The exicyclog utility now does better if the number of log files to keep
3899 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3901 59. Two changes related to the smtp_active_hostname option:
3903 (1) $smtp_active_hostname is now available as a variable.
3904 (2) The default for smtp_banner uses $smtp_active_hostname instead
3905 of $primary_hostname.
3907 60. The host_aton() function is supposed to be passed a string that is known
3908 to be a valid IP address. However, in the case of IPv6 addresses, it was
3909 not checking this. This is a hostage to fortune. Exim now panics and dies
3910 if the condition is not met. A case was found where this could be provoked
3911 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3912 components; fortuitously, this particular loophole had already been fixed
3913 by change 4.50/55 above.
3915 If there are any other similar loopholes, the new check in host_aton()
3916 itself should stop them being exploited. The report I received stated that
3917 data on the command line could provoke the exploit when Exim was running as
3918 exim, but did not say which command line option was involved. All I could
3919 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3920 running as the user.
3923 61. There was a buffer overflow vulnerability in the SPA authentication code
3924 (which came originally from the Samba project). I have added a test to the
3925 spa_base64_to_bits() function which I hope fixes it.
3928 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3929 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3931 63. The daemon start-up calls getloadavg() while still root for those OS that
3932 need the first call to be done as root, but it missed one case: when
3933 deliver_queue_load_max is set with deliver_drop_privilege. This is
3934 necessary for the benefit of the queue runner, because there is no re-exec
3935 when deliver_drop_privilege is set.
3937 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3938 This has been fixed.
3940 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3941 was in use, was not putting the data itself into the right store pool;
3942 consequently, it could be overwritten for a subsequent message in the same
3943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3946 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3948 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3949 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3950 the function were treating the return as a boolean value, which happened to
3951 work because 0=false and not-0=true, but is not correct code.
3953 68. The host_aton() function was not handling scoped IPv6 addresses (those
3954 with, for example, "%eth0" on the end) correctly.
3956 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3957 negated items (that is, ~something) in unsigned ints. Some compilers
3958 apparently mutter when there is no cast.
3960 70. If an address verification called from an ACL failed, and did not produce a
3961 user-specific message (i.e. there was only a "system" message), nothing was
3962 put in $acl_verify_message. In this situation, it now puts the system
3965 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3966 session; change 4.31/43 added the unwanted input to the log line - except
3967 that it did not do this in the start of session case. It now does.
3969 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3970 This is wrong and can cause the other end to generate a synchronization
3971 error if it is another Exim or anything else that does the synchronization
3972 check. A QUIT command is no longer sent after a timeout.
3974 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3975 during host lookups.
3977 74. The defer_ok option of callout verification was not working if it was used
3978 when verifying addresses in header lines, that is, for this case:
3980 verify = header_sender/callout=defer_ok
3982 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3983 those file descriptors could be used for SMTP connections. If anything
3984 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3985 could be sent to the SMTP client, causing chaos. The daemon now opens
3986 stdin, stdout, and stderr to /dev/null when it puts itself into the
3989 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3990 log by default. The user can override this with suitable Perl magic.
3992 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3993 the log message when discard triggers, was not working for the DATA ACL or
3994 for the non-SMTP ACL.
3996 78. Error message wording change in sieve.c.
3998 79. If smtp_accept_max_per_host was set, the number of connections could be
3999 restricted to fewer than expected, because the daemon was trying to set up
4000 a new connection before checking whether the processes handling previous
4001 connections had finished. The check for completed processes is now done
4002 earlier. On busy systems, this bug wouldn't be noticed because something
4003 else would have woken the daemon, and it would have reaped the completed
4006 80. If a message was submitted locally by a user whose login name contained one
4007 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4008 It caused a spool format error. I have fixed the spool reading code. A
4009 related problem was that the "from" clause in the Received: line became
4010 illegal because of the space(s). It is now covered by ${quote_local_part.
4012 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4015 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4017 83. Patch from Sieve maintainer to fix unterminated string problem in
4018 "vacation" handling.
4020 84. Some minor changes to the Linux configuration files to help with other
4021 OS variants using glibc.
4023 85. One more patch for Sieve to update vacation handling to latest spec.
4026 ----------------------------------------------------
4027 See the note above about the 4.44 and 4.50 releases.
4028 ----------------------------------------------------
4034 1. Change 4.43/35 introduced a bug that caused file counts to be
4035 incorrectly computed when quota_filecount was set in an appendfile
4038 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4039 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4042 3. Give more explanation in the error message when the command for a transport
4043 filter fails to execute.
4045 4. There are several places where Exim runs a non-Exim command in a
4046 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4047 was being done only for the command run by the queryprogram router. It is
4048 now done for all such subprocesses. The other cases are: ${run, transport
4049 filters, and the commands run by the lmtp and pipe transports.
4051 5. Some older OS have a limit of 256 on the maximum number of file
4052 descriptors. Exim was using setrlimit() to set 1000 as a large value
4053 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4054 systems. I've change it so that if it can't get 1000, it tries for 256.
4056 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4057 was an oversight, and furthermore, ever since the addition of extra
4058 controls (e.g. 4.43/32), the checks on when to allow different forms of
4059 "control" were broken. There should now be diagnostics for all cases when a
4060 control that does not make sense is encountered.
4062 7. $recipients is now available in the predata ACL (oversight).
4064 8. Tidy the search cache before the fork to do a delivery from a message
4065 received from the command line. Otherwise the child will trigger a lookup
4066 failure and thereby defer the delivery if it tries to use (for example) a
4067 cached ldap connection that the parent has called unbind on.
4069 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4070 of $address_data from the recipient verification was clobbered by the
4071 sender verification.
4073 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4074 was its contents. (It was OK if the option was not defined at all.)
4076 11. A "Completed" log line is now written for messages that are removed from
4077 the spool by the -Mrm option.
4079 12. $host_address is now set to the target address during the checking of
4080 ignore_target_hosts.
4082 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4083 being passed; this would have caused $sender_host_name to have been used if
4084 matching the list had actually called for a host name (not very likely,
4085 since this list is usually IP addresses). A host name is now passed as
4088 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4089 code that specifies a non-restarting handler (typically sigaction() in
4090 modern systems) in an attempt to fix a rare and obscure crash bug.
4092 15. Narrowed the window for a race in the daemon that could cause it to ignore
4093 SIGCHLD signals. This is not a major problem, because they are used only to
4094 wake it up if nothing else does.
4096 16. A malformed maildirsize file could cause Exim to calculate negative values
4097 for the mailbox size or file count. Odd effects could occur as a result.
4098 The maildirsize information is now recalculated if the size or filecount
4101 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4102 support for a long time. Removed HAVE_SYS_VFS_H.
4104 18. Updated exipick to current release from John Jetmore.
4106 19. Allow an empty sender to be matched against a lookup in an address list.
4107 Previously the only cases considered were a regular expression, or an
4110 20. Exim went into a mad DNS lookup loop when doing a callout where the
4111 host was specified on the transport, if the DNS lookup yielded more than
4114 21. The RFC2047 encoding function was originally intended for short strings
4115 such as real names; it was not keeping to the 75-character limit for
4116 encoded words that the RFC imposes. It now respects the limit, and
4117 generates multiple encoded words if necessary. To be on the safe side, I
4118 have increased the buffer size for the ${rfc2047: expansion operator from
4121 22. Failure to deliver a bounce message always caused it to be frozen, even if
4122 there was an errors_to setting on the router. The errors_to setting is now
4125 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4126 canonical form (fully expanded) before being placed in
4127 $sender_host_address.
4129 24. Updated eximstats to version 1.33
4131 25. Include certificate and key file names in error message when GnuTLS fails
4132 to set them up, because the GnuTLS error message doesn't include the name
4133 of the failing file when there is a problem reading it.
4135 26. Expand error message when OpenSSL has problems setting up cert/key files.
4138 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4139 (this can affect the format of dates).
4141 28. exim_tidydb, when checking for the continued existence of a message for
4142 which it has found a message-specific retry record, was not finding
4143 messages that were in split spool directories. Consequently, it was
4144 deleting retry records that should have stayed in existence.
4146 29. eximstats updated to version 1.35
4147 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4148 1.35 - bugfix such that pie charts by volume are generated correctly
4150 30. The SPA authentication driver was not abandoning authentication and moving
4151 on to the next authenticator when an expansion was forced to fail,
4152 contradicting the general specification for all authenticators. Instead it
4153 was generating a temporary error. It now behaves as specified.
4155 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4156 (the order specifies the preference for clients). The order is now AES256,
4157 AES128, 3DES, ARCFOUR128.
4159 31. Small patch to Sieve code - explicitly set From: when generating an
4162 32. Exim crashed if a remote delivery caused a very long error message to be
4163 recorded - for instance if somebody sent an entire SpamAssassin report back
4164 as a large number of 550 error lines. This bug was coincidentally fixed by
4165 increasing the size of one of Exim's internal buffers (big_buffer) that
4166 happened as part of the Exiscan merge. However, to be on the safe side, I
4167 have made the code more robust (and fixed the comments that describe what
4170 33. Some experimental protocols are using DNS PTR records for new purposes. The
4171 keys for these records are domain names, not reversed IP addresses. The
4172 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4173 leaves it alone. Component reversal etc. now happens only for IP addresses.
4176 34. The host_aton() function is supposed to be passed a string that is known
4177 to be a valid IP address. However, in the case of IPv6 addresses, it was
4178 not checking this. This is a hostage to fortune. Exim now panics and dies
4179 if the condition is not met. A case was found where this could be provoked
4180 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4181 components; fortuitously, this particular loophole had already been fixed
4182 by change 4.50/55 or 4.44/33 above.
4184 If there are any other similar loopholes, the new check in host_aton()
4185 itself should stop them being exploited. The report I received stated that
4186 data on the command line could provoke the exploit when Exim was running as
4187 exim, but did not say which command line option was involved. All I could
4188 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4189 running as the user.
4192 35. There was a buffer overflow vulnerability in the SPA authentication code
4193 (which came originally from the Samba project). I have added a test to the
4194 spa_base64_to_bits() function which I hope fixes it.
4197 36. The daemon start-up calls getloadavg() while still root for those OS that
4198 need the first call to be done as root, but it missed one case: when
4199 deliver_queue_load_max is set with deliver_drop_privilege. This is
4200 necessary for the benefit of the queue runner, because there is no re-exec
4201 when deliver_drop_privilege is set.
4203 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4204 was in use, was not putting the data itself into the right store pool;
4205 consequently, it could be overwritten for a subsequent message in the same
4206 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4209 38. Sometimes the final signoff response after QUIT could fail to get
4210 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4211 < 0 before doing a fflush(). This bug looks as though it goes back to the
4212 introduction of TLS in release 3.20, but "sometimes" must have been rare
4213 because the tests only now provoked it.
4219 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4220 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4221 Now it returns an integer. A number of calls were still expecting a T/F
4222 return. Fortuitously, in all cases, the tests worked in OK situations,
4223 which is the norm. However, things would have gone wrong on any write
4224 failures on the smtp file descriptor. This function is used when sending
4225 messages over SMTP and also when doing verify callouts.
4227 2. When Exim is called to do synchronous delivery of a locally submitted
4228 message (the -odf or -odi options), it no longer closes stderr before doing
4231 3. Implemented the mua_wrapper option.
4233 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4235 5. Implemented the functions header_remove(), header_testname(),
4236 header_add_at_position(), and receive_remove_recipient(), and exported them
4239 6. If an ACL "warn" statement specified the addition of headers, Exim already
4240 inserted X-ACL-Warn: at the start if there was no header name. However, it
4241 was not making this test for the second and subsequent header lines if
4242 there were newlines in the string. This meant that an invalid header could
4243 be inserted if Exim was badly configured.
4245 7. Allow an ACL "warn" statement to add header lines at the start or after all
4246 the Received: headers, as well as at the end.
4248 8. Added the rcpt_4xx retry error code.
4250 9. Added postmaster_mailfrom=xxx to callout verification option.
4252 10. Added mailfrom=xxxx to the callout verification option, for verify=
4255 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4256 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4258 12. Inserted some casts to stop certain compilers warning when using pointer
4259 differences as field lengths or precisions in printf-type calls (mostly
4260 affecting debugging statements).
4262 13. Added optional readline() support for -be (dynamically loaded).
4264 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4265 same clock tick as a message's arrival, so that its received time was the
4266 same as the "first fail" time on the retry record, and that message
4267 remained on the queue past the ultimate address timeout, every queue runner
4268 would try a delivery (because it was past the ultimate address timeout) but
4269 after another failure, the ultimate address timeout, which should have then
4270 bounced the address, did not kick in. This was a "< instead of <=" error;
4271 in most cases the first failure would have been in the next clock tick
4272 after the received time, and all would be well.
4274 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4275 being recognized when the domain list was tested by the match_domain
4276 condition in an expansion string.
4278 16. Added the ${str2b64: operator.
4280 17. Exim was always calling setrlimit() to set a large limit for the number of
4281 processes, without checking whether the existing limit was already
4282 adequate. (It did check for the limit on file descriptors.) Furthermore,
4283 errors from getrlimit() and setrlimit() were being ignored. Now they are
4284 logged to the main and panic logs, but Exim does carry on, to try to do its
4285 job under whatever limits there are.
4287 18. Imported PCRE 5.0.
4289 19. Trivial typo in log message " temporarily refused connection" (the leading
4292 20. If the log selector return_path_on_delivery was set and an address was
4293 redirected to /dev/null, the delivery process crashed because it assumed
4294 that a return path would always be set for a "successful" delivery. In this
4295 case, the whole delivery is bypassed as an optimization, and therefore no
4298 21. Internal re-arrangement: the function for sending a challenge and reading
4299 a response while authentication was assuming a zero-terminated challenge
4300 string. It's now changed to take a pointer and a length, to allow for
4301 binary data in such strings.
4303 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4305 23. Exim was not respecting finduser_retries when seeking the login of the
4306 uid under which it was called; it was always trying 10 times. (The default
4307 setting of finduser_retries is zero.) Also, it was sleeping after the final
4308 failure, which is pointless.
4310 24. Implemented tls_on_connect_ports.
4312 25. Implemented acl_smtp_predata.
4314 26. If the domain in control=submission is set empty, Exim assumes that the
4315 authenticated id is a complete email address when it generates From: or
4316 Sender: header lines.
4318 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4319 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4320 chown and chgrp in /bin and hostname in /usr/bin.
4322 28. Exim was keeping the "process log" file open after each use, just as it
4323 does for the main log. This opens the possibility of it remaining open for
4324 long periods when the USR1 signal hits a daemon. Occasional processlog
4325 errors were reported, that could have been caused by this. Anyway, it seems
4326 much more sensible not to leave this file open at all, so that is what now
4329 29. The long-running daemon process does not normally write to the log once it
4330 has entered its main loop, and it closes the log before doing so. This is
4331 so that log files can straightforwardly be renamed and moved. However,
4332 there are a couple of unusual error situations where the daemon does write
4333 log entries, and I had neglected to close the log afterwards.
4335 30. The text of an SMTP error response that was received during a remote
4336 delivery was being truncated at 512 bytes. This is too short for some of
4337 the long messages that one sometimes sees. I've increased the limit to
4340 31. It is now possible to make retry rules that apply only when a message has a
4341 specific sender, in particular, an empty sender.
4343 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4344 it possible to be selective about when SMTP synchronization is enforced.
4346 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4348 32. Implemented hosts_connection_nolog.
4350 33. Added an ACL for QUIT.
4352 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4355 35. Added mailbox_size and mailbox_filecount to appendfile.
4357 36. Added control = no_multiline_responses to ACLs.
4359 37. There was a bug in the logic of the code that waits for the clock to tick
4360 in the case where the clock went backwards by a substantial amount such
4361 that the microsecond fraction of "now" was more than the microsecond
4362 fraction of "then" (but the whole seconds number was less).
4364 38. Added support for the libradius Radius client library this is found on
4365 FreeBSD (previously only the radiusclient library was supported).
4371 1. When certain lookups returned multiple values in the form name=value, the
4372 quoting of the values was not always being done properly. Specifically:
4373 (a) If the value started with a double quote, but contained no whitespace,
4375 (b) If the value contained whitespace other than a space character (i.e.
4376 tabs or newlines or carriage returns) it was not quoted.
4377 This fix has been applied to the mysql and pgsql lookups by writing a
4378 separate quoting function and calling it from the lookup code. The fix
4379 should probably also be applied to nisplus, ibase and oracle lookups, but
4380 since I cannot test any of those, I have not disturbed their existing code.
4382 2. A hit in the callout cache for a specific address caused a log line with no
4383 reason for rejecting RCPT. Now it says "Previous (cached) callout
4384 verification failure".
4386 3. There was an off-by-one bug in the queryprogram router. An over-long
4387 return line was truncated at 256 instead of 255 characters, thereby
4388 overflowing its buffer with the terminating zero. As well as fixing this, I
4389 have increased the buffer size to 1024 (and made a note to document this).
4391 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4392 when Exim is waiting for an SMTP response from a remote server, Exim
4393 restarts its select() call on the socket, thereby resetting its timeout.
4394 This is not a problem when such interrupts are rare. Somebody set up a cron
4395 job to run exiwhat every 2 minutes, which is less than the normal select()
4396 timeout (5 or 10 minutes). This meant that the select() timeout never
4397 kicked in because it was always reset. I have fixed this by comparing the
4398 time when an interrupt arrives with the time at the start of the first call
4399 to select(). If more time than the timeout has elapsed, the interrupt is
4400 treated as a timeout.
4402 5. Some internal re-factoring in preparation for the addition of Sieve
4403 extensions (by MH). In particular, the "personal" test is moved to a
4404 separate function, and given an option for scanning Cc: and Bcc: (which is
4405 not set for Exim filters).
4407 6. When Exim created an email address using the login of the caller as the
4408 local part (e.g. when creating a From: or Sender: header line), it was not
4409 quoting the local part when it contained special characters such as @.
4411 7. Installed new OpenBSD configuration files.
4413 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4414 try to make them clearer.
4416 9. Callout options, other than the timeout value, were being ignored when
4417 verifying sender addresses in header lines. For example, when using
4419 verify = header_sender/callout=no_cache
4421 the cache was (incorrectly) being used.
4423 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4424 only the Cygwin environment.
4426 11. When return_path_on_delivery was set as a log selector, if different remote
4427 addresses in the same message used different return paths and parallel
4428 remote delivery occurred, the wrong values would sometimes be logged.
4429 (Whenever a remote delivery process finished, the return path value from
4430 the most recently started remote delivery process was logged.)
4432 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4433 lines when AUTH and/or TLS are in use. This is the "received protocol"
4434 field. Exim used to use "asmtp" for authenticated SMTP, without any
4435 indication (in the protocol name) for TLS use. Now it follows the RFC and
4436 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4437 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4438 names appear in log lines as well as in Received: header lines.
4440 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4441 extensions, and comparison tests, and to fix some bugs.
4443 14. Changes to the "personal" filter test:
4445 (1) The test was buggy in that it was just doing the equivalent of
4446 "contains" tests on header lines. For example, if a user's address was
4447 anne@some.where, the "personal" test would incorrectly be true for
4449 To: susanne@some.where
4451 This test is now done by extracting each address from the header in turn,
4452 and checking the entire address. Other tests that are part of "personal"
4453 are now done using regular expressions (for example, to check local parts
4454 of addresses in From: header lines).
4456 (2) The list of non-personal local parts in From: addresses has been
4457 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4458 taken from the Sieve specification recommendations.
4460 (3) If the message contains any header line starting with "List-" it is
4461 treated as non-personal.
4463 (4) The test for "circular" in the Subject: header line has been removed
4464 because it now seems ill-conceived.
4466 15. Minor typos in src/EDITME comments corrected.
4468 16. Installed latest exipick from John Jetmore.
4470 17. If headers_add on a router specified a text string that was too long for
4471 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4472 of string_sprintf() is now avoided.
4474 18. $message_body_size was not set (it was always zero) when running the DATA
4475 ACL and the local_scan() function.
4477 19. For the "mail" command in an Exim filter, no default was being set for
4478 the once_repeat time, causing a random time value to be used if "once" was
4479 specified. (If the value happened to be <= 0, no repeat happened.) The
4480 default is now 0s, meaning "never repeat". The "vacation" command was OK
4481 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4482 (I found it when inspecting the code).
4484 20. There is now an overall timeout for performing a callout verification. It
4485 defaults to 4 times the callout timeout, which applies to individual SMTP
4486 commands during the callout. The overall timeout applies when there is more
4487 than one host that can be tried. The timeout is checked before trying the
4488 next host. This prevents very long delays if there are a large number of
4489 hosts and all are timing out (e.g. when the network connections are timing
4490 out). The value of the overall timeout can be changed by specifying an
4491 additional sub-option for "callout", called "maxwait". For example:
4493 verify = sender/callout=5s,maxwait=20s
4495 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4496 to the end before writing, but this should make it even safer).
4498 22. Exim was forgetting that it had advertised PIPELINING for the second and
4499 subsequent messages on an SMTP connection. It was also not resetting its
4500 memory on STARTTLS and an internal HELO.
4502 23. When Exim logs an SMTP synchronization error within a session, it now
4503 records whether PIPELINING has been advertised or not.
4505 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4506 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4507 rather than long int.
4509 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4511 26. Added the never_mail option to autoreply.
4517 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4518 crash if the getsockname() call failed; this can happen if a connection is
4519 closed very soon after it is established. The problem was simply in the
4520 order in which certain operations were done, causing Exim to try to write
4521 to the SMTP stream before it had set up the file descriptor. The bug has
4522 been fixed by making things happen in the correct order.
4528 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4529 before the connection was closed, thus losing the rejection response.
4531 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4532 some early Solaris releases, but causes trouble in current releases where
4533 socklen_t is defined.
4535 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4538 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4541 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4542 that releasing the top of it at the end releases what was used for sub-
4543 expansions (unless the block got too big). However, discard this block if
4544 the first thing is a variable or header, so that we can use its block when
4545 it is dynamic (useful for very large $message_headers, for example).
4547 6. Lookups now cache *every* query, not just the most recent. A new, separate
4548 store pool is used for this. It can be recovered when all lookup caches are
4549 flushed. Lookups now release memory at the end of their result strings.
4550 This has involved some general refactoring of the lookup sources.
4552 7. Some code has been added to the store_xxx() functions to reduce the amount
4553 of flapping under certain conditions.
4555 8. log_incoming_interface used to affect only the <= reception log lines. Now
4556 it causes the local interface and port to be added to several more SMTP log
4557 lines, for example "SMTP connection from", and rejection lines.
4559 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4561 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4563 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4564 could be overwritten at the end of the current message (or the start of a
4565 new message if it was set in a HELO ACL). The value is now preserved for
4566 the duration of the SMTP connection.
4568 12. If a transport had a headers_rewrite setting, and a matching header line
4569 contained an unqualified address, that address was qualified, even if it
4570 did not match any rewriting rules. The underlying bug was that the values
4571 of the flags that permit the existence of unqualified sender and recipient
4572 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4573 non-local messages, and by -bnq for local messages) were not being
4574 preserved with the message after it was received.
4576 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4577 "next input=" as part of the text comprising the host identity instead of
4578 the correct text. The code was using the same buffer for two different
4579 strings. However, depending on which order the printing function evaluated
4580 its arguments, the bug did not always show up. Under Linux, for example, my
4581 test suite worked just fine.
4583 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4584 doesn't work with some older versions of Perl. It has been changed to "my",
4585 which in any case is probably the better facility to use.
4587 15. A really picky compiler found some instances of statements for creating
4588 error messages that either had too many or two few arguments for the format
4591 16. The size of the buffer for calls to the DNS resolver has been increased
4592 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4593 for addresses that have a lot of PTR records. This alleviates a problem; it
4594 does not fully solve it.
4596 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4597 buffer now truncates the list and logs the incident, which is the same
4598 action as happens when Exim is looking up a host name and its aliases.
4599 Previously in this situation something unpredictable would happen;
4600 sometimes it was "internal error: store_reset failed".
4602 18. If a server dropped the connection unexpectedly when an Exim client was
4603 using GnuTLS and trying to read a response, the client delivery process
4604 crashed while trying to generate an error log message.
4606 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4607 string, for example:
4609 warn message = H1: something\nH2: something
4611 the text was added as a single header line from Exim's point of view
4612 though it ended up OK in the delivered message. However, searching for the
4613 second and subsequent header lines using $h_h2: did not work. This has been
4614 fixed. Similarly, if a system filter added multiple headers in this way,
4615 the routers could not see them.
4617 20. Expanded the error message when iplsearch is called with an invalid key to
4618 suggest using net-iplsearch in a host list.
4620 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4621 ACLs are no longer actually imposed (and a message to that effect is
4624 22. If a "gecos" field in a passwd entry contained escaped characters, in
4625 particular, if it contained a \" sequence, Exim got it wrong when building
4626 a From: or a Sender: header from that name. A second bug also caused
4627 incorrect handling when an unquoted " was present following a character
4628 that needed quoting.
4630 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4631 was not being matched caselessly.
4633 24. Arranged for all hyphens in the exim.8 source to be escaped with
4636 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4637 properly. Recipient callout cache records were still being keyed to include
4638 the sender, even when use_sender was set false. This led to far more
4639 callouts that were necessary. The sender is no longer included in the key
4640 when use_sender is false.
4642 26. Added "control = submission" modifier to ACLs.
4644 27. Added the ${base62d: operator to decode base 62 numbers.
4646 28. dnsdb lookups can now access SRV records.
4648 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4649 the configuration file.
4651 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4652 (-v) mode. This makes the output for a verbose queue run more intelligible.
4654 31. Added a use_postmaster feature to recipient callouts.
4656 32. Added the $body_zerocount variable, containing the number of binary zero
4657 bytes in the message body.
4659 33. The time of last modification of the "new" subdirectory is now used as the
4660 "mailbox time last read" when there is a quota error for a maildir
4663 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4665 35. Added +ignore_unknown as a special item in host lists.
4667 36. Code for decoding IPv6 addresses in host lists is now included, even if
4668 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4669 address was recognized as an IP address, but was then not correctly decoded
4670 into binary, causing unexpected and incorrect effects when compared with
4677 1. Very minor rewording of debugging text in manualroute to say "list of
4678 hosts" instead of "hostlist".
4680 2. If verify=header_syntax was set, and a header line with an unqualified
4681 address (no domain) and a large number of spaces between the end of the
4682 name and the colon was received, the reception process suffered a buffer
4683 overflow, and (when I tested it) crashed. This was caused by some obsolete
4684 code that should have been removed. The fix is to remove it!
4686 3. When running in the test harness, delay a bit after writing a bounce
4687 message to get a bit more predictability in the log output.
4689 4. Added a call to search_tidyup() just before forking a reception process. In
4690 theory, someone could use a lookup in the expansion of smtp_accept_max_
4691 per_host which, without the tidyup, could leave open a database connection.
4693 5. Added the variables $recipient_data and $sender_data which get set from a
4694 lookup success in an ACL "recipients" or "senders" condition, or a router
4695 "senders" option, similar to $domain_data and $local_part_data.
4697 6. Moved the writing of debug_print from before to after the "senders" test
4700 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4701 problems for message scanning, either using a data ACL, or using
4702 local_scan() because the Received: header was not generated till after they
4703 were called (in order to set the time as the time of reception completion).
4704 I have revised the way this works. The header is now generated after the
4705 body is received, but before the ACL or local_scan() are called. After they
4706 are run, the timestamp in the header is updated.
4712 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4713 before starting a queue runner without re-exec. This happened only when
4714 deliver_drop_privilege was set or when the Exim user was set to root. The
4715 effect of the bug was that timeouts during subsequent deliveries caused
4716 crashes instead of being properly handled. The handler is now left at its
4717 default (and expected) setting.
4719 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4720 message, again when deliver_drop_privilege is set or Exim is run as root.
4721 The bug described in (1) was not present in this case, but the tidying up
4722 of the other signals was missing. I have made the two cases consistent.
4724 3. The ignore_target_hosts setting on a manualroute router was being ignored
4725 for hosts that were looked up using the /MX notation.
4727 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4730 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4731 operated on the sender address. After changing the $sender_address to <>
4732 for the sender address verify, Exim was re-instated it as the original
4733 (before rewriting) address, but remembering that it had rewritten it, so it
4734 wasn't rewriting it again. This bug also had the effect of breaking the
4735 sender address verification caching when the sender address was rewritten.
4737 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4738 This has been changed so that if the ip literal address matches
4739 ignore_target_hosts, the router declines.
4741 7. Added expansion conditions match_domain, match_address, and match_local_
4742 part (NOT match_host).
4744 8. The placeholder for the Received: header didn't have a length field set.
4746 9. Added code to Exim itself and to exim_lock to test for a specific race
4747 condition that could lead to file corruption when using MBX delivery. The
4748 issue is with the lockfile that is created in /tmp. If this file is removed
4749 after a process has opened it but before that process has acquired a lock,
4750 there is the potential for a second process to recreate the file and also
4751 acquire a lock. This could lead to two Exim processes writing to the file
4752 at the same time. The added code performs the same test as UW imapd; it
4753 checks after acquiring the lock that its file descriptor still refers to
4754 the same named file.
4756 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4757 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4760 11. Added the smtp_active_hostname option. If used, this will typically be made
4761 to depend on the incoming interface address. Because $interface_address is
4762 not set up until the daemon has forked a reception process, error responses
4763 that can happen earlier (such as "too many connections") no longer contain
4766 12. If an expansion in a condition on a "warn" statement fails because a lookup
4767 defers, the "warn" statement is abandoned, and the next ACL statement is
4768 processed. Previously this caused the whole ACL to be aborted.
4770 13. Added the iplsearch lookup type.
4772 14. Added ident_timeout as a log selector.
4774 15. Added tls_certificate_verified as a log selector.
4776 16. Added a global option tls_require_ciphers (compare the smtp transport
4777 option of the same name). This controls incoming TLS connections.
4779 17. I finally figured out how to make tls_require_ciphers do a similar thing
4780 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4781 before starting the TLS session.
4783 18. Tabs are now shown as \t in -bP output.
4785 19. If the log selector return_path_on_delivery was set, Exim crashed when
4786 bouncing a message because it had too many Received: header lines.
4788 20. If two routers both had headers_remove settings, and the first one included
4789 a superfluous trailing colon, the final name in the first list and the
4790 first name in the second list were incorrectly joined into one item (with a
4791 colon in the middle).
4797 1. Added -C and -D options to the exinext utility, mainly to make it easier
4798 to include in the automated testing, but these could be helpful when
4799 multiple configurations are in use.
4801 2. The exinext utility was not formatting the output nicely when there was
4802 an alternate port involved in the retry record key, nor when there was a
4803 message id as well (for retries that were specific to a specific message
4804 and a specific host). It was also confused by IPv6 addresses, because of
4805 the additional colons they contain. I have fixed the IPv4 problem, and
4806 patched it up to do a reasonable job for IPv6.
4808 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4809 delivery, the log line now contains "pipelined" if PIPELINING was used.
4811 4. An SMTP transport process used to panic and die if the bind() call to set
4812 an explicit outgoing interface failed. This has been changed; it is now
4813 treated in the same way as a connect() failure.
4815 5. A reference to $sender_host_name in the part of a conditional expansion
4816 that was being skipped was still causing a DNS lookup. This no longer
4819 6. The def: expansion condition was not recognizing references to header lines
4820 that used bh_ and bheader_.
4822 7. Added the _cache feature to named lists.
4824 8. The code for checking quota_filecount in the appendfile transport was
4825 allowing one more file than it should have been.
4827 9. For compatibility with Sendmail, the command line option
4835 and sets the incoming protocol and host name (for trusted callers). The
4836 host name and its colon can be omitted when only the protocol is to be set.
4837 Note the Exim already has two private options, -pd and -ps, that refer to
4838 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4839 "s", but I don't think that's a major issue.
4841 10. A number of refactoring changes to the code, none of which should affect
4844 (a) The number of logging options was getting close to filling up the
4845 32-bit word that was used as a bit map. I have split them into two classes:
4846 those that are passed in the argument to log_write(), and those that are
4847 only ever tested independently outside of that function. These are now in
4848 separate 32-bit words, so there is plenty of room for expansion again.
4849 There is no change in the user interface or the logging behaviour.
4851 (b) When building, for example, log lines, the code previously used a
4852 macro that called string_cat() twice, in order to add two strings. This is
4853 not really sufficiently general. Furthermore, there was one instance where
4854 it was actually wrong because one of the argument was used twice, and in
4855 one call a function was used. (As it happened, calling the function twice
4856 did not affect the overall behaviour.) The macro has been replaced by a
4857 function that can join an arbitrary number of extra strings onto a growing
4860 (c) The code for expansion conditions now uses a table and a binary chop
4861 instead of a serial search (which was left over from when there were very
4862 few conditions). Also, it now recognizes conditions like "pam" even when
4863 the relevant support is not compiled in: a suitably worded error message is
4864 given if an attempt is made to use such a condition.
4866 11. Added ${time_interval:xxxxx}.
4868 12. A bug was causing one of the ddress fields not to be passed back correctly
4869 from remote delivery subprocesses. The field in question was not being
4870 subsequently used, so this caused to problems in practice.
4872 13. Added new log selectors queue_time and deliver_time.
4874 14. Might have fixed a bug in maildirsizefile handling that threw up
4875 "unexpected character" debug warnings, and recalculated the data
4876 unnecessarily. In any case, I expanded the warning message to give more
4879 15. Added the message "Restricted characters in address" to the statements in
4880 the default ACL that block characters like @ and % in local parts.
4882 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4883 Three changes have been made:
4885 (a) There was a serious bug; a negative response to MAIL caused the whole
4886 recipient domain to be cached as invalid, thereby blocking all messages
4887 to all local parts at the same domain, from all senders. This bug has
4888 been fixed. The domain is no longer cached after a negative response to
4889 MAIL if the sender used is not empty.
4891 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4894 (c) A new callout option, "use_sender" has been added for people who want
4895 the modified behaviour.
4901 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4904 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4905 indeed breaks things for older releases.
4907 3. Added additional logging to the case where there is a problem reading data
4908 from a filter that is running in a subprocess using a pipe, in order to
4909 try to track down a specific problem.
4911 4. Testing facility fudge: when running in the test harness and attempting
4912 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4913 getting "No route to host". Convert this to a timeout.
4915 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4918 6. Some OS don't have socklen_t but use size_t instead. This affects the
4919 fifth argument of getsockopt() amongst other things. This is now
4920 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4921 can be set for individual OS. I have set it for SunOS5, OSF1, and
4922 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4923 some earlier ones do not.
4925 7. Change 4.30/15 was not doing the test caselessly.
4927 8. The standard form for an IPv6 address literal was being rejected by address
4928 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4929 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4930 this, as well as the form without the "IPv6" on the front (but only when
4931 address literals are enabled, of course).
4933 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4935 10. Exim crashed if a message with an empty sender address specified by -f
4936 encountered a router with an errors_to setting. This could be provoked only
4937 by a command such as
4941 where an empty string was supplied; "<>" did not hit this bug.
4943 11. Installed PCRE release 4.5.
4945 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4946 remained set. It is now erased.
4948 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4949 times from message ids (which are base 36 rather than the normal 62).
4951 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4952 were being counted as actual protocol errors, and logged if the log
4953 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4954 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4955 and DATA following a set of rejected RCPTs do not count as protocol errors.
4956 In other words, Exim assumes they were pipelined, though this may not
4957 actually be the case. Of course, in all cases the client gets an
4958 appropriate error code.
4960 15. If a lookup fails in an ACL condition, a message about the failure may
4961 be available; it is used if testing the ACL cannot continue, because most
4962 such messages specify what the cause of the deferral is. However, some
4963 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4964 that caused an old message to be retained and used if a later statement
4965 caused a defer, replacing the real cause of the deferral.
4967 16. If an IP address had so many PTR records that the DNS lookup buffer
4968 was not large enough to hold them, Exim could crash while trying to process
4969 the truncated data. It now detects and logs this case.
4971 17. Further to 4.21/58, another change has been made: if (and only if) the
4972 first line of a message (the first header line) ends with CRLF, a bare LF
4973 in a subsequent header line has a space inserted after it, so as not to
4974 terminate the header.
4976 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4977 unnecessarily, used atoi() instead of strtol(), and didn't check the
4978 termination when getting file sizes from file names by regex.
4980 19. Completely re-implemented the support for maildirsize files, in the light
4981 of a number of problems with the previous contributed implementation
4982 (4.30/29). In particular:
4984 . If the quota is zero, the maildirsize file is maintained, but no quota is
4987 . If the maildir directory does not exist, it is created before any attempt
4988 to write a maildirsize file.
4990 . The quota value in the file is just a cache; if the quota is changed in
4991 the transport, the new value overrides.
4993 . A regular expression is available for excluding directories from the
4996 20. The autoreply transport checks the characters in options that define the
4997 message's headers; it allows continued headers, but it was checking with
4998 isspace() after an embedded newline instead of explicitly looking for a
5001 21. If all the "regular" hosts to which an address was routed had passed their
5002 expiry times, and had not reached their retry times, the address was
5003 bounced, even if fallback hosts were defined. Now Exim should go on to try
5006 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5007 equivalent code in the SMTP transport. Some hosts send humungous responses
5008 to HELO/EHLO, more than 1024 it seems.
5010 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5011 gives compiler warnings in some environments. I've now done it "properly",
5014 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5015 (because of problems with the built-in one) was declared to return uschar *
5016 instead of char *, causing compiler failure.
5018 25. Fixed a file descriptor leak when processing alias/forward files.
5020 26. Fixed a minor format string issue in dbfn.c.
5022 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5024 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5025 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5028 29. When a sender address is verified, it is cached, to save repeating the test
5029 when there is more than one recipient in a message. However, when the
5030 verification involves a callout, it is possible for different callout
5031 options to be set for different recipients. It is too complicated to keep
5032 track of this in the cache, so now Exim always runs a verification when a
5033 callout is required, relying on the callout cache for the optimization.
5034 The overhead is duplication of the address routing, but this should not be
5037 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5038 to be verified with callout=postmaster, and the main callout worked but the
5039 postmaster check failed, the verification correctly failed. However, if a
5040 subsequent RCPT command asked for sender verification *without* the
5041 postmaster check, incorrect caching caused this verification also to fail,
5044 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5045 it was not caching the DNS options (qualify_single, search_parents) that
5046 were used when the lookup failed. A subsequent lookup with different
5047 options therefore always gave the same answer, though there were cases
5048 where it should not have. (Example: a "domains = !$mx_any" option on a
5049 dnslookup router: the "domains" option is always processed without any
5050 widening, but the router might have qualify_single set.) Now Exim uses the
5051 cached value only when the same options are set.
5053 32. Added John Jetmore's "exipick" utility to the distribution.
5055 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5056 than a timeout (e.g. a certificate is required, and is not provided), an
5057 Exim server now closes the connection immediately. Previously it waited for
5058 the client to close - but if the client is SSL, it seems that they each
5059 wait for each other, leading to a delay before one of them times out.
5061 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5062 maintained 0.8.x compatibility because I don't think many are using it, and
5063 it is clearly obsolete.
5065 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5068 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5069 client certificate was expired. A simple patch fixes this, though I don't
5070 understand the full logic of why the verify callback is called multiple
5073 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5074 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5075 which causes problems with some clients (such as the Certicom SSL Plus
5076 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5077 disables the coutermeasure allowing Eudora to connect."
5079 38. Exim was not checking that a write() to a log file succeeded. This could
5080 lead to Bad Things if a log got too big, in particular if it hit a file
5081 size limit. Exim now panics and dies if it cannot write to a log file, just
5082 as it does if it cannot open a log file.
5084 39. Modified OS/Makefile-Linux so that it now contains
5086 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5088 The two -D definitions ensure that Exim is compiled with large file
5089 support, which makes it possible to handle log files that are bigger than
5092 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5093 instance) a domain was checked against a named list that involved a lookup,
5094 causing $domain_data to be set, then another domain was checked against the
5095 same list, then the first domain was re-checked, the value of $domain_data
5096 after the final check could be wrong. In particular, if the second check
5097 failed, it could be set empty. This bug probably also applied to
5100 41. The strip_trailing_dot option was not being applied to the address given
5101 with the -f command-line option.
5103 42. The code for reading a message's header from the spool was incrementing
5104 $received_count, but never initializing it. This meant that the value was
5105 incorrect (doubled) while delivering a message in the same process in which
5106 it was received. In the most common configuration of Exim, this never
5107 happens - a fresh exec is done - but it can happen when
5108 deliver_drop_privilege is set.
5110 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5111 it now includes up to 150 characters of the unexpected data in the log
5114 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5115 and building data strings. The size of both of these buffers was 10 000
5116 bytes - far larger than anybody would *ever* want, thought I. Needless to
5117 say, somebody hit the limit. I have increased the maximum line length to
5118 20 000 and the maximum data length of concatenated lines to 100 000. I have
5119 also fixed two bugs, because there was no checking on these buffers. Tsk,
5120 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5121 buffer is too small.
5123 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5124 lsearch lookups. Now it does.
5126 46. When parsing a route_list item in a manualroute router, a fixed-length
5127 buffer was used for the list of hosts. I made this 1024 bytes long,
5128 thinking that nobody would ever have a list of hosts that long. Wrong.
5129 Somebody had a whole pile of complicated expansion conditions, and the
5130 string was silently truncated, leading to an expansion error. It turns out
5131 that it is easier to change to an unlimited length (owing to other changes
5132 that have happened since this code was originally written) than to build
5133 structure for giving a limitation error. The length of the item that
5134 expands into the list of hosts is now unlimited.
5136 47. The lsearch lookup could not handle data where the length of text line was
5137 more than 4095 characters. Such lines were truncated, leading to shortened
5138 data being returned. It should now handle lines of any length.
5140 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5141 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5144 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5145 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5146 no longer happens in comment lines. A list of replacements is now placed
5147 at the head of all of the source files, except those whose only change is
5148 to replace PERL_COMMAND in the very first #! line.
5150 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5151 messages on the queue, with a bottom-up merge sort, using code contributed
5152 by Michael Haardt. This should make operations like -bp somewhat faster on
5153 large queues. It won't affect queue runners, except when queue_run_in_order
5156 51. Installed eximstats 1.31 in the distribution.
5158 52. Added support for SRV lookups to the dnslookup router.
5160 53. If an ACL referred to $message_body or $message_body_end, the value was not
5161 reset for any messages that followed in the same SMTP session.
5163 54. The store-handling optimization for building very long strings was not
5164 differentiating between the different store pools. I don't think this
5165 actually made any difference in practice, but I've tidied it.
5167 55. While running the routers to verify a sender address, $sender_address
5168 was still set to the sender address. This is wrong, because when routing to
5169 send a bounce to the sender, it would be empty. Therefore, I have changed
5170 it so that, while verifying a sender address, $sender_address is set to <>.
5171 (There is no change to what happens when verifying a recipient address.)
5173 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5174 target A or AAAA records (if not already returned) without resetting the
5175 qualify_single or search_parents options of the DNS resolver. These are
5176 inappropriate in this case because the targets of MX and SRV records must
5177 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5178 target that, when qualified, matched something in the local domain. These
5179 two options are now turned off when doing these lookups.
5181 57. It seems that at least some releases of Reiserfs (which does not have the
5182 concept of a fixed number of inodes) returns zero and not -1 for the
5183 number of available inodes. This interacted badly with check_spool_inodes,
5184 which assumed that -1 was the "no such thing" setting. What I have done is
5185 to check that the total number of inodes is greater than zero before doing
5186 the test of how many are available.
5188 58. When a "warn" ACL statement has a log_message modifier, the message is
5189 remembered, and not repeated. This is to avoid a lot of repetition when a
5190 message has many recipients that cause the same warning to be written.
5191 Howewer, Exim was preserving the list of already written lines for an
5192 entire SMTP session, which doesn't seem right. The memory is now reset if a
5193 new message is started.
5195 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5196 debugging output unless log_rewrite was also set.
5198 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5199 of (int)(handle) when we know that handle contains (void *)(-1).
5201 61. The Exim daemon panic-logs an error return when it closes the incoming
5202 connection. However "connection reset by peer" seems to be common, and
5203 isn't really an error worthy of noting specially, so that particular error
5206 62. When Exim is trying to find all the local interfaces, it used to panic and
5207 die if the ioctl to get the interface flags failed. However, it seems that
5208 on at least one OS (Solaris 9) it is possible to have an interface that is
5209 included in the list of interfaces, but for which you get a failure error
5210 for this call. This happens when the interface is not "plumbed" into a
5211 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5212 failure of the "get flags" call assumes that the interface is down.
5214 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5215 makes life easier for people who are doing arithmetic on fields extracted
5216 from dates, where you often get leading zeros that should not be
5217 interpreted as octal.
5219 64. Added qualify_domain to the redirect router, to override the global
5222 65. If a pathologically long header line contained very many addresses (the
5223 report of this problem mentioned 10 000) and each of them was rewritten,
5224 Exim could use up a very large amount of memory. (It kept on making new
5225 copies of the header line as it rewrote, and never released the old ones.)
5226 At the expense of a bit more processing, the header rewriting function has
5227 been changed so that it no longer eats memory in this way.
5229 66. The generation of the Received: header has been moved from the time that a
5230 message starts to be received, to the time that it finishes. The timestamp
5231 in the Received: header should now be very close to that of the <= log
5232 line. There are two side-effects of this change:
5234 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5235 logged header lines no longer include the local Received: line, because
5236 it has not yet been created. The same applies to a copy of the message
5237 that is returned to a non-SMTP sender when a message is rejected.
5239 (b) When a filter file is tested using -bf, no additional Received: header
5240 is added to the test message. After some thought, I decided that this
5243 This change does not affect the value of $received_for. It is still set
5244 after address rewriting, but before local_scan() is called.
5246 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5248 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5249 gave an unhelpful panic error message, and a defer error. I have managed to
5250 change this behaviour so that it now rejects any supplied certificate,
5251 which seems right, as the list of acceptable certificates is empty.
5253 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5254 gave an unhelpful defer error. I have not managed to make this reject any
5255 supplied certificates, but the error message it gives is "no certificate
5256 supplied", which is not helpful.
5258 70. exigrep's output now also includes lines that are not associated with any
5259 message, but which match the given pattern. Implemented by a patch from
5260 Martin Sluka, which also tidied up the Perl a bit.
5262 71. Recipient callout verification, like sender verification, was using <> in
5263 the MAIL FROM command. This isn't really the right thing, since the actual
5264 sender may affect whether the remote host accepts the recipient or not. I
5265 have changed it to use the actual sender in the callout; this means that
5266 the cache record is now keyed on a recipient/sender pair, not just the
5267 recipient address. There doesn't seem to be a real danger of callout loops,
5268 since a callout by the remote host to check the sender would use <>.
5269 [SEE ABOVE: changed after hitting problems.]
5271 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5272 temporary errors. However, in the case of such a code being given after
5273 the end of a data transmission (i.e. after ".") Exim was failing to write
5274 a retry record for the message. (Yes, there was some broken host that was
5275 actually sending 8xx at this point.)
5277 73. An unknown lookup type in a host list could cause Exim to panic-die when
5278 the list was checked. (An example that provoked this was putting <; in the
5279 middle of a list instead of at the start.) If this happened during a DATA
5280 ACL check, a -D file could be left lying around. This kind of configuration
5281 error no longer causes Exim to die; instead it causes a defer errror. The
5282 incident is still logged to the main and panic logs.
5284 74. Buglet left over from Exim 3 conversion. The message "too many messages
5285 in one connection" was written to the rejectlog but not the mainlog, except
5286 when address rewriting (yes!) was being logged.
5288 75. Added write_rejectlog option.
5290 76. When a system filter was run not as root (that is, when system_filter_user
5291 was set), the values of the $n variables were not being returned to the
5292 main process; thus, they were not subsequently available in the $sn
5295 77. Added +return_path_on_delivery log selector.
5297 78. A connection timeout was being treated differently from recipients deferred
5298 when testing hosts_max_try with a message that was older than the host's
5299 retry timeout. (The host should not be counted, thus allowing all hosts to
5300 be tried at least once before bouncing.) This may have been the cause of an
5301 occasionally reported bug whereby a message would remain on the queue
5302 longer than the retry timeout, but would be bounced if a delivery was
5303 forced. I say "may" because I never totally pinned down the problem;
5304 setting up timeout/retry tests is difficult. See also the next item.
5306 79. The ultimate address timeout was not being applied to errors that involved
5307 a combination of host plus message (for example, a timeout on a MAIL
5308 command). When an address resolved to a number of possible hosts, and they
5309 were not all tried for each delivery (e.g. because of hosts_max_try), a
5310 message could remain on the queue longer than the retry timeout.
5312 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5315 81. Fixed an obscure SMTP outgoing bug which required at least the following
5316 conditions: (a) there was another message waiting for the same server;
5317 (b) the server returned 5xx to all RCPT commands in the first message so
5318 that the message was not completed; (c) the server dropped the connection
5319 or gave a negative response to the RSET that Exim sends to abort the
5320 transaction. The observed case was a dropped connection after DATA that had
5321 been sent in pipelining mode. That is, the server had advertised PIPELINING
5322 but was not implementing it correctly. The effect of the bug was incorrect
5323 behaviour, such as trying another host, and this could lead to a crash.
5329 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5330 and daemon.c were passed as pointers to ints; they should have been
5331 pointers to socklen_t variables (which are typically unsigned ints).
5333 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5336 3. Fixed a really odd bug that affected only the testing scheme; patching a
5337 certain fixed string in the binary changed the value of another string that
5338 happened to be identical to the end of the original first string.
5340 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5341 name", it returns that address as the IP address. On some operating
5342 systems (e.g. Solaris), it also passes back the IP address string as the
5343 "host name". However, on others (e.g. Linux), it passes back an empty
5344 string. Exim wasn't checking for this, and was changing the host name to an
5345 empty string, assuming it had been canonicized.
5347 5. Although rare, it is permitted to have more than one PTR record for a given
5348 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5349 all the names associated with an address, because they do in Solaris.
5350 However, it seems that they do not in Linux for data that comes from the
5351 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5352 I found this out when I moved to a new Linux workstation and tried to run
5353 the Exim test suite.
5355 To get round this problem I have changed the code so that it now does its
5356 own call to the DNS to look up PTR records when searching for a host name.
5357 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5358 addresses that are only in /etc/hosts are still found.
5360 This behaviour is, however, controlled by an option called host_lookup_
5361 order, which defaults to "bydns:byaddr". If people want to use the other
5362 order, or indeed, just use one or the other means of lookup, they can
5363 specify it in this variable.
5365 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5366 some operating systems, this comes back from gethostbyaddr() as an empty
5367 string, and this is what Exim used to test for. However, it seems that in
5368 other systems, "." is yielded. Exim now tests for this case too.
5370 7. The values of check_spool_space and check_log_space are now held internally
5371 as a number of kilobytes instead of an absolute number of bytes. If a
5372 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5373 kilobyte. This means that much larger values can be stored.
5375 8. Exim monitor: an attempt to get the action menu when not actually pointing
5376 at a message produces an empty menu entitled "No message selected". This
5377 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5378 no entries in it ("Shell widget menu has zero width and/or height"). So I
5379 have added a single, blank menu entry in this case.
5381 9. Added ${quote_local_part.
5383 10. MIME decoding is now applied to the contents of Subject: header lines when
5386 11. Now that a reference to $sender_host_address automatically causes a reverse
5387 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5388 host lookup before query-style lookups in lists that might use this
5389 variable. This has therefore been abolished, and the "net-" prefix is no
5390 longer necessary for query-style lookups.
5392 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5393 have been a typo for LFLAGS, so it has been changed.
5395 13. The install script calls Exim with "-C /dev/null" in order to find the
5396 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5397 to be output. Howeve, since Exim outputs its version number before the
5398 error, it didn't break the script. It just looked ugly. I fixed this by
5399 always allowing "-C /dev/null" if the caller is root.
5401 14. Ignore overlarge ACL variable number when reading spool file - insurance
5402 against a later release with more variables having written the file.
5404 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5405 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5406 this, as well as the form without the "IPv6" on the front.
5408 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5409 OS/Makefile-Darwin file.
5411 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5412 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5414 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5415 variable. However, some OS use ints for inodes. I've added cast to long int
5416 to get rid of the compiler warning.
5418 19. I had forgotten to lock out "/../" in configuration file names when
5419 ALT_CONFIG_PREFIX was set.
5421 20. Routers used for verification do not need to specify transports. However,
5422 if such a router generated a host list, and callout was configured, Exim
5423 crashed, because it could not find a port number from the (non-existent)
5424 transport. It now assumes port 25 in this circumstance.
5426 21. Added the -t option to exigrep.
5428 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5429 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5430 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5431 them is set, LOOKUP_LSEARCH is forced.
5433 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5435 24. Added sender and host information to the "rejected by local_scan()" log
5436 line; previously there was no indication of these.
5438 25. Added .include_if_exists.
5440 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5441 Linux. It turns out that not all file systems support this. Apparently some
5442 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5443 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5444 not supported on the file descriptor, is now ignored when Exim is trying to
5445 sync a directory. This applies only to Linux.
5447 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5449 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5450 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5453 29. Added support for maildirsize files from supplied patch (modified a bit).
5455 30. The use of :fail: followed by an empty string could lead Exim to respond to
5456 sender verification failures with (e.g.):
5458 550 Verification failed for <xxx>
5459 550 Sender verify failed
5461 where the first response line was missing the '-' that indicates it is not
5462 the final line of the response.
5464 31. The loop for finding the name of the user that called Exim had a hardwired
5465 limit of 10; it now uses the value of finduser_retries, which is used for
5466 all other user lookups.
5468 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5471 33. Exim was neglecting to zero errno before one call of strtol() when
5472 expanding a string and expecting an integer value. On some systems this
5473 resulted in spurious "integer overflow" errors. Also, it was casting the
5474 result into an int without checking.
5476 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5477 did not work. The code looks as if it has *never* worked, though it appears
5478 to have been documented since at least releast 1.62. I have made it work.
5480 35. The "timeout_DNS" error in retry rules, also documented since at least
5481 1.62, also never worked. As it isn't clear exactly what this means, and
5482 clearly it isn't a major issue, I have abolished the feature by treating it
5483 as "timeout", and writing a warning to the main and panic logs.
5485 36. The display of retry rules for -brt wasn't always showing the error code
5488 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5489 timeout_connect_A, timeout_connect_MX.
5491 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5492 to the empty sender.
5494 39. The daemon was not analysing the content of -oX till after it had closed
5495 stderr and disconnected from the controlling terminal. This meant that any
5496 syntax errors were only noted on the panic log, and the return code from
5497 the command was 0. By re-arranging the code a little, I've made the
5498 decoding happen first, so such errors now appear on stderr, and the return
5499 code is 1. However, the actual setting up of the sockets still happens in
5500 the disconnected process, so errors there are still only recorded on the
5503 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5504 connections (as happens on some IP stacks) was logged at start up time as
5505 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5506 it from "IPv6 and IPv4", which means that two separate sockets are being
5509 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5510 says whether AF_INET or AF_INET6 was passed as an argument.
5512 42. Exiwhat output was messed up when time zones were included in log
5515 43. Exiwhat now gives more information about the daemon's listening ports,
5516 and whether -tls-on-connect was used.
5518 44. The "port" option of the smtp transport is now expanded.
5520 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5521 silently ignored. Now an error message is written to the main and panic
5524 46. There's a new ACL modifier called "logwrite" which writes to a log file
5525 as soon as it is encountered.
5527 47. Added $local_user_uid and $local_user_gid at routing time.
5529 48. Exim crashed when trying to verify a sender address that was being
5532 49. Exim was recognizing only a space character after ".include". It now also
5533 recognizes a tab character.
5535 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5536 extracting the relevant information from the specification. The man page no
5537 longer contains scrambled data for the -d option, and I've added a section
5538 at the front about calling Exim under different names.
5540 51. Added "extra_headers" argument to the "mail" command in filter files.
5542 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5545 53. Installed eximstats 1.29.
5547 54. Added transport_filter_timeout as a generic transport option.
5549 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5550 Cc: header lines. This was required by RFC 822, but it not required by RFC
5553 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5554 incoming messages that did not have them. Now it does so only if the
5555 message originates locally, that is, if there is no associated remote host
5556 address. When Resent- header lines are present, this applies to the Resent-
5557 lines rather than the non-Resent- lines.
5559 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5560 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5562 58. Messages for configuration errors now include the name of the main
5563 configuration file - useful now that there may be more than one file in a
5564 list (.included file names were always shown).
5566 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5567 for those rare installations that do not start the daemon as root or run it
5568 setuid root. I've cut out the call to initgroups() if the daemon is not
5571 60. The Exim user and group can now be bound into the binary as text strings
5572 that are looked up at the start of Exim's processing.
5574 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5576 62. Added $mailstore_basename variable.
5578 63. Installed patch to sieve.c from Michael Haardt.
5580 64. When Exim failed to open the panic log after failing to open the main log,
5581 the original message it was trying to log was written to stderr and debug
5582 output, but if they were not available (the usual case in production), it
5583 was lost. Now it is written to syslog before the two lines that record the
5584 failures to open the logs.
5586 65. Users' Exim filters run in subprocesses under the user's uid. It is
5587 possible for a "deliver" command or an alias in a "personal" command to
5588 provoke an address rewrite. If logging of address rewriting is configured,
5589 this fails because the process is not running as root or exim. There may be
5590 a better way of dealing with this, but for the moment (because 4.30 needs
5591 to be released), I have disabled address rewrite logging when running a
5592 filter in a non-root, non-exim process.
5598 1. The buildconfig auxiliary program wasn't quoting the value set for
5599 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5600 not defined. This bug was masked in 4.22 by the effect that was fixed in
5603 2. Some messages that were rejected after a message id was allocated were
5604 shown as "incomplete" by exigrep. It no longer does this for messages that
5605 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5607 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5608 have allow_domain_literals set, the ID did not get logged in the <= line.
5609 Domain literals are now always recognized in Message-ID: header lines.
5611 4. The first argument for a ${extract expansion item is the key name or field
5612 number. Leading and trailing spaces in this item were not being ignored,
5613 causing some misleading effects.
5615 5. When deliver_drop_privilege was set, single queue runner processes started
5616 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5617 same command in the process it spins off) were not dropping privilege.
5619 6. When the daemon running as "exim" started a queue runner, it always
5620 re-executed Exim in the spun-off process. This is a waste of effort when
5621 deliver_drop_privilege is set. The new process now just calls the
5622 queue-runner function directly.
5628 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5631 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5632 ignored. Though the use of -oP was forcing the writing of a pid file, it
5633 was always written to the default place.
5635 3. If the message "no IP address found for host xxxx" is generated during
5636 incoming verification, it is now followed by identification of the incoming
5637 connection (so you can more easily find what provoked it).
5639 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5641 5. Added some features to "harden" Exim a bit more against certain attacks:
5643 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5644 be put in Local/Makefile. This is like the never_users runtime option,
5645 but it cannot be overridden. The default setting is "root".
5647 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5648 prefix string with which any file named in a -C command line option
5651 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5652 is retained for -C and -D only if the caller of Exim is root. Without
5653 it, the exim user may also use -C and -D and retain privilege.
5655 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5656 command line option is disabled.
5658 6. Macro names set by the -D option must start with an upper case letter, just
5659 like macro names defined in the configuration file.
5661 7. Added "dereference=" facility to LDAP.
5663 8. Two instances of the typo "uknown" in the source files are fixed.
5665 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5666 the Configure-Makefile script screwed up while processing it.
5668 10. Incorporated PCRE 4.4.
5670 11. The SMTP synchronization check was not operating right at the start of an
5671 SMTP session. For example, it could not catch a HELO sent before the client
5672 waited for the greeting. There is now a check for outstanding input at the
5673 point when the greeting is written. Because of the duplex, asynchronous
5674 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5675 way, but not yet received, when the check is performed.
5677 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5678 on TCP/IP sockets, because this apparently causes some broken clients to
5681 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5682 unchanged) from the Cygwin maintainer.
5684 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5685 was supported instead of testing for mbx. Effectively a typo.
5687 15. The spa authenticator server code was not checking that the input it
5688 received was valid base64.
5690 16. The debug output line for the "set" modifier in ACLs was not showing the
5691 name of the variable that was being set.
5693 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5695 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5696 lookup on its own. Something else was needed to trigger the lookup. For
5697 example, a match in host_lookup or the need for a host name in a host list.
5698 Now, if $sender_host_name is referenced and the host name has not yet been
5699 looked up, a lookup is performed. If the lookup fails, the variable remains
5700 empty, and $host_lookup_failed is set to "1".
5702 19. Added "eqi" as a case-independent comparison operator.
5704 20. The saslauthd authentication condition could segfault if neither service
5705 nor realm was specified.
5707 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5708 error message that was logged was misleading, and incoming SMTP
5709 connections were dropped. The message is now more accurate, and temporary
5710 errors are given to SMTP connections.
5712 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5713 (except RSET) with a 421 error, until QUIT is received. However, it was
5714 failing to send a response to QUIT.
5716 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5717 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5718 ACL is now run after the helo_try_verify_hosts code.
5720 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5721 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5722 case-independent, so other case variants are also recognized). Apparently
5723 some systems use these upper case variants.
5725 25. If more than two messages were waiting for the same host, and a transport
5726 filter was specified for the transport, Exim sent two messages over the
5727 same TCP/IP connection, and then failed with "socket operation on non-
5728 socket" when it tried to send the third.
5730 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5732 27. The extern definition of crypt16() in expand.c was not being excluded when
5733 the OS had its own crypt16() function.
5735 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5736 as a preferred synonym for return_size_limit, both as an option and as an
5739 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5741 30. Changed the default configuration ACL to relax the local part checking rule
5742 for addresses that are not in any local domains. For these addresses,
5743 slashes and pipe symbols are allowed within local parts, but the sequence
5744 /../ is explicitly forbidden.
5746 31. SPA server authentication was not clearing the challenge buffer before
5749 32. log_message in a "warn" ACL statement was writing to the reject log as
5750 well as to the main log, which contradicts the documentation and doesn't
5751 seem right (because no rejection is happening). So I have stopped it.
5753 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5754 However, I am unable to do any testing of this.
5756 34. Fixed an infelicity in the appendfile transport. When checking directories
5757 for a mailbox, to see if any needed to be created, it was accidentally
5758 using path names with one or more superfluous leading slashes; tracing
5759 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5761 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5762 contents are added to the log line that is written for every discarded
5763 recipient. (Previously a log_message setting was ignored.)
5765 36. The ${quote: operator now quotes the string if it is empty.
5767 37. The install script runs exim in order to find its version number. If for
5768 some reason other than non-existence or emptiness, which it checks, it
5769 could not run './exim', it was installing it with an empty version number,
5770 i.e. as "exim-". This error state is now caught, and the installation is
5773 38. An argument was missing from the function that creates an error message
5774 when Exim fails to connect to the socket for saslauthd authentication.
5775 This could cause Exim to crash, or give a corrupted message.
5777 39. Added isip, isip4, and isip6 to ${if conditions.
5779 40. The ACL variables $acl_xx are now saved with the message, and can be
5780 accessed later in routers, transports, and filters.
5782 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5783 strings in the file are not string-expanded.
5785 42. If a MAIL command specified a SIZE value that was too large to fit into an
5786 int variable, the check against message_size_limit failed. Such values are
5787 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5788 day this will have to be increased, but I don't think I want to be around
5789 when emails are that large.
5796 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5797 iconv() is not standard in FreeBSD.
5799 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5800 IPv6 enabled. The observed symptom was a segmentation fault on return from
5801 the function os_common_find_running_interfaces() in src/os.c.
5803 3. In the check_special_case() function in daemon.c I had used "errno" as an
5804 argument name, which causes warnings on some systems. This was basically a
5805 typo, since it was named "eno" in the comments!
5807 4. The code that waits for the clock to tick (at a resolution of some fraction
5808 of a second) so as to ensure message-id uniqueness was always waiting for
5809 at least one whole tick, when it could have waited for less. [This is
5810 almost certainly not relevant at current processor speeds, where it is
5811 unlikely to ever wait at all. But we try to future-proof.]
5813 5. The function that sleeps for a time interval that includes fractions of a
5814 second contained a race. It did not block SIGALRM between setting the
5815 timer, and suspending (a couple of lines later). If the interval was short
5816 and the sigsuspend() was delayed until after it had expired, the suspension
5817 never ended. On busy systems this could lead to processes getting stuck for
5820 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5821 process, before it forks any delivery processes. The open lookup caching
5822 mechanism meant that the open file or database connection was passed into
5823 the delivery process. The problem was that delivery processes always tidy
5824 up cached lookup data. This could cause a problem for the next delivery
5825 process started by the queue runner, because the external queue runner
5826 process does not know about the closure. So the next delivery process
5827 still has data in the lookup cache. In the case of a file lookup, there was
5828 no problem because closing a file descriptor in a subprocess doesn't affect
5829 the parent. However, if the lookup was caching a connection to a database,
5830 the connection was closed, and the second delivery process was likely to
5831 see errors such as "PGSQL: query failed: server closed the connection
5832 unexpectedly". The problem has been fixed by closing all cached lookups
5833 in a queue runner before running a delivery process.
5835 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5836 seem to have the "const" qualifier which it has on other OS. I've
5839 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5840 *of the same type* (client or server) with the same public name that an
5841 error should be diagnosed.
5843 9. When Exim looked up a host name for an IP address, but failed to find the
5844 original IP address when looking up the host name (a safety check), it
5845 output the message "<ip address> does not match any IP for NULL", which was
5846 confusing, to say the least. The bug was that the host name should have
5847 appeared instead of "NULL".
5849 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5850 user that is built into the binary, and the -C or -D options is used, root
5851 privilege is dropped before the configuration file is read. In addition,
5852 logging is switched to stderr instead of the normal log files. If the
5853 configuration then re-defines the Exim user, the unprivileged environment
5854 is probably not what is expected, so Exim logs a panic warning message (but
5857 However, if deliver_drop_privilege is set, the unprivileged state may well
5858 be exactly what is intended, so the warning has been cut out in that case,
5859 and Exim is allowed to try to write to its normal log files.
5865 1. smtp_return_error_details was not giving details for temporary sender
5866 or receiver verification errors.
5868 2. Diagnose a configuration error if two authenticators have the same public
5871 3. Exim used not to create the message log file for a message until the first
5872 delivery attempt. This could be confusing when incoming messages were held
5873 for policy or load reasons. The message log file is now created at the time
5874 the message is received, and an initial "Received" line is written to it.
5876 4. The automatically generated man page for command line options had a minor
5877 bug that caused no ill effects; however, a more serious problem was that
5878 the procedure for building the man page automatically didn't always
5879 operate. Consequently, release 4.20 contains an out-of-date version. This
5880 shouldn't happen again.
5882 5. When building Exim with embedded Perl support, the script that builds the
5883 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5884 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5886 6. The freeze_tell option was not being used for messages that were frozen on
5887 arrival, either by an ACL or by local_scan().
5889 7. Added the smtp_incomplete_transaction log selector.
5891 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5892 was accepting AUTH without a new EHLO.
5894 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5895 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5898 10. Logging of TCP/IP connections (when configured) now happens in the main
5899 daemon process instead of the child process, so that the TCP/IP connection
5900 count is more accurate (but it can never be perfect).
5902 11. The use of "drop" in a nested ACL was not being handled correctly in the
5903 outer ACL. Now, if condition failure induced by the nested "drop" causes
5904 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5905 or "require"), the connection is dropped.
5907 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5908 that yield "discard" can now be used with an "accept" or a "discard" verb,
5909 but an error is generated for any others (because I can't see a useful way
5910 to define what should happen).
5912 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5913 are now processed in the same way as lines in the Exim configuration file.
5914 In particular, continuation lines are supported.
5916 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5918 15. Added -ti meaning -t -i.
5920 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5921 domains, and warn by logging if others are found.
5923 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5924 returned from GIFCONF when Exim is trying to find the list of interfaces on
5925 a host. The code in os.c has been modified to copy each ifreq to an aligned
5926 structure in all cases.
5928 Also, in some cases, the returned ifreq's were being copied to a 'struct
5929 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5930 means the last couple of bytes of an IPv6 address could be chopped if the
5931 ifreq contained only a normal sockaddr (14 bytes storage).
5933 18. Named domain lists were not supported in the hosts_treat_as_local option.
5934 An entry such as +xxxx was not recognized, and was treated as a literal
5937 19. Ensure that header lines added by a DATA ACL are included in the reject log
5938 if the ACL subsequently rejects the message.
5940 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5941 MD5 (which is deprecated).
5943 21. When testing a filter file using -bf, Exim was writing a message when it
5944 took the sender from a "From " line in the message, but it was not doing so
5945 when it took $return_path from a Return-Path: header line. It now does.
5947 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5948 with a valid header line field name (a series of printing characters
5949 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5951 23. Changed "disc" in the source to "disk" to conform to the documentation and
5952 the book and for uniformity.
5954 24. Ignore Sendmail's -Ooption=value command line item.
5956 25. When execve() failed while trying to run a command in a pipe transport,
5957 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5958 could be confused with a return value of 69 from the command itself. This
5959 has been changed to 127, the value the shell returns if it is asked to run
5960 a non-existent command. The wording for the related log line suggests a
5961 non-existent command as the problem.
5963 26. If received_header_text expands to an empty string, do not add a Received:
5964 header line to the message. (Well, it adds a token one on the spool, but
5965 marks it "old" so that it doesn't get used or transmitted.)
5967 27. Installed eximstats 1.28 (addition of -nt option).
5969 28. There was no check for failure on the call to getsockname() in the daemon
5970 code. This can fail if there is a shortage of resources on the system, with
5971 ENOMEM, for example. A temporary error is now given on failure.
5973 29. Contrary to the C standard, it seems that in some environments, the
5974 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5975 program. Exim now does this explicitly; it affects the formatting of
5976 timestamps using strftime().
5978 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5979 complaints. I've now initialized all the variables, to avoid this.
5981 32. Header lines added by a system filter were not being "seen" during
5982 transport-time rewrites.
5984 33. The info_callback() function passed to OpenSSL is set up with type void
5985 (*)(SSL *, int, int), as described somewhere. However, when calling the
5986 function (actually a macro) that sets it up, the type void(*)() is
5987 expected. I've put in a cast to prevent warnings from picky compilers.
5989 34. If a DNS black list lookup found a CNAME record, but there were no A
5990 records associated with the domain it pointed at, Exim crashed.
5992 35. If a DNS black list lookup returned more than one A record, Exim ignored
5993 all but the first. It now scans all returned addresses if a particular IP
5994 value is being sought. In this situation, the contents of the
5995 $dnslist_value variable are a list of all the addresses, separated by a
5998 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5999 to accept a host name and all its aliases if the forward lookup for any of
6000 them yielded the IP address of the incoming connection. Now it accepts only
6001 those names whose forward lookup yields the correct IP address. Any other
6002 names are discarded. This closes a loophole whereby a rogue DNS
6003 administrator could create reverse DNS records to break through a
6004 wildcarded host restriction in an ACL.
6006 37. If a user filter or a system filter that ran in a subprocess used any of
6007 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6008 the wrong values were passed to the pipe command ($thisaddress had the
6009 value of $0, $0 had the value of $1, etc). This bug was introduced by
6010 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6012 38. Improved the line breaking for long SMTP error messages from ACLs.
6013 Previously, if there was no break point between 40 and 75 characters, Exim
6014 left the rest of the message alone. Two changes have been made: (a) I've
6015 reduced the minimum length to 35 characters; (b) if it can't find a break
6016 point between 35 and 75 characters, it looks ahead and uses the first one
6017 that it finds. This may give the occasional overlong line, but at least the
6018 remaining text gets split now.
6020 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6021 file descriptors might be low, and that setting 1000 would always raise it.
6022 It turns out that in some environments, the limit is already over 1000 and
6023 that lowering it causes trouble. So now Exim takes care not to decrease it.
6025 40. When delivering a message, the value of $return_path is set to $sender_
6026 address at the start of routing (routers may change the value). By an
6027 oversight, this default was not being set up when an address was tested by
6028 -bt or -bv, which affected the outcome if any router or filter referred to
6031 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6032 log only when "message" or "log_message" are set. However, if one of the
6033 conditions was an address verification, or a call to a nested ACL, the
6034 messages generated by the underlying test were being passed through. This
6035 no longer happens. The underlying message is available in $acl_verify_
6036 message for both "message" and "log_message" expansions, so it can be
6037 passed through if needed.
6039 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6040 new expansion $bh_ to give the encoded byte string without charset
6041 translation. Translation happens only if iconv() is available; HAVE_ICONV
6042 indicates this at build time. HEADERS_CHARSET gives the charset to
6043 translate to; headers_charset can change it in the configuration, and
6044 "headers charset" can change it in an individual filter file.
6046 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6047 that creates RFC 2047 encoded "words" labels them as that charset instead
6048 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6049 expansion operator; (ii) when Exim creates a From: line for a local
6050 message; (iii) when a header line is rewritten to include a "phrase" part.
6052 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6053 buggy, causing it to skip the first lines of messages whose message ID
6054 ended in 'D'. This would not have bitten before Exim release 4.14, because
6055 message IDs were unlikely to end in 'D' before then. The effect was to have
6056 incorrect size information for certain domains.
6058 45. #include "config.h" was missing at the start of the crypt16.c module. This
6059 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6062 46. If there was a timeout during a "random" callout check, Exim treated it as
6063 a failure of the random address, and carried on sending RSET and the real
6064 address. If the delay was just some slowness somewhere, the response to the
6065 original RCPT would be taken as a response to RSET and so on, causing
6066 mayhem of various kinds.
6068 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6069 when I implemented it. It didn't allow for the fact that some option values
6070 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6071 the right test for positive values.
6073 48. Domain names in DNS records are case-independent. Exim always looks them up
6074 in lower case. Some resolvers return domain names in exactly the case they
6075 appear in the zone file, that is, they may contain uppercase letters. Not
6076 all resolvers do this - some return always lower case. Exim was treating a
6077 change of case by a resolver as a change of domain, similar to a widening
6078 of a domain abbreviation. This triggered its re-routing code and so it was
6079 trying to route what was effectively the same domain again. This normally
6080 caused routing to fail (because the router wouldn't handle the domain
6081 twice). Now Exim checks for this case specially, and just changes the
6082 casing of the domain that it ultimately uses when it transmits the message
6085 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6088 50. If a filter generated a file delivery with a non-absolute name (possible if
6089 no home directory exists for the router), the forbid_file option was not
6092 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6093 the existing equality matching.
6095 52. Exim was using ints instead of ino_t variables in some places where it was
6096 dealing with inode numbers.
6098 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6099 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6100 called TMPDIR, and if it finds it is different, it changes its value.
6102 54. The smtp_printf() function is now made available to local_scan() so
6103 additional output lines can be written before returning. There is also an
6104 smtp_fflush() function to enable the detection of a dropped connection.
6105 The variables smtp_input and smtp_batched_input are exported to
6108 55. Changed the default runtime configuration: the message "Unknown user"
6109 has been removed from the ACL, and instead placed on the localuser router,
6110 using the cannot_route_message feature. This means that any verification
6111 failures that generate their own messages won't get overridden. Similarly,
6112 the "Unrouteable address" message that was in the ACL for unverifiable
6113 relay addresses has also been removed.
6115 56. Added hosts_avoid_esmtp to the smtp transport.
6117 57. The exicyclog script was not checking for the esoteric option
6118 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6119 will work only if exicyclog is run under the appropriate euid.
6121 58. Following a discussion on the list, the rules by which Exim recognises line
6122 endings on incoming messages have been changed. The -dropcr and drop_cr
6123 options are now no-ops, retained only for backwards compatibility. The
6124 following line terminators are recognized: LF CRLF CR. However, special
6125 processing applies to CR:
6127 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6128 nor a local message in the state where . is a terminator.
6130 (ii) If a bare CR is encountered in a header line, an extra space is added
6131 after the line terminator so as not to end the header. The reasoning
6132 behind this is that bare CRs in header lines are most likely either
6133 to be mistakes, or people trying to play silly games.
6135 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6136 was being incorrectly given as 18 bytes larger than it should have been.
6137 This is a VOB (very old bug).
6139 60. This may never have affected anything current, but just in case it has:
6140 When the local host is found other than at the start of a list of hosts,
6141 the local host, those with the same MX, and any that follow, are discarded.
6142 When the list in question was part of a longer list of hosts, the following
6143 hosts (not currently being processed) were also being discarded. This no
6144 longer happens. I'm not sure if this situation could ever has previously
6147 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6150 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6151 header. This is something that is recommended in a new Internet Draft, and
6152 is something that is documented as being done by Sendmail. There are two
6153 possible values. For messages generated by the autoreply transport, Exim
6156 Auto-Submitted: auto-replied
6158 whereas for all other generated messages (e.g. bounces) it adds
6160 Auto-Submitted: auto-generated
6162 63. The "personal" condition in filters now includes a test for the
6163 Auto-Submitted: header. If it contains the string "auto-" the message it
6164 not considered personal.
6166 64. Added rcpt_include_affixes as a generic transport option.
6168 65. Added queue_only_override (default true).
6170 66. Added the syslog_duplication option.
6172 67. If what should have been the first header line of a message consisted of
6173 a space followed by a colon, Exim was mis-interpreting it as a header line.
6174 It isn't of course - it is syntactically invalid and should therefore be
6175 treated as the start of the message body. The misbehaviour could have
6176 caused a number of strange effects, including loss of data in subsequent
6177 header lines, and spool format errors.
6179 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6180 client host had authenticated. This control can now be exercised by an ACL
6181 for more flexibility.
6183 69. By default, callouts do not happen when testing with -bh. There is now a
6184 variant, -bhc, which does actually run the callout code, including
6185 consulting and updating the callout cache.
6187 70. Added support for saslauthd authentication, courtesy of Alexander
6190 71. If statvfs() failed on the spool or log directories while checking their
6191 size for availability, Exim confusingly gave the error "space shortage".
6192 Furthermore, in debugging mode it crashed with a floating point exception.
6193 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6194 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6195 serious problem, Exim now writes to the main and panic logs when this
6196 happens, with details of the failure. It then refuses to accept the
6197 incoming message, giving the message "spool directory problem" or "log
6198 directory problem" with a 421 code for SMTP messages.
6200 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6201 0, 1, and 2 exist, because some OS complain for execs without them (see
6202 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6203 descriptors. However, the code omitted to check that the open succeeded,
6204 causing mysterious errors if for some reason the permissions on /dev/null
6205 got screwed. Now Exim writes a message to the main and panic logs, and
6206 bombs out if it can't open /dev/null.
6208 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6209 interact so that it is all more flexible. It is supposed to remain
6210 backwards compatible. Also added extra_local_interfaces.
6212 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6213 to bomb out with an assertion failure - to the client this appears as a
6214 connection drop. This problem occurs in the part of the code that was taken
6215 from the Samba project. Fortunately, the assertion is in a very simple
6216 function, so I have fixed this by reproducing the function inline in the
6217 one place where it is called, and arranging for authentication to fail
6218 instead of killing the process with assert().
6220 75. The SPA client code was not working when the server requested OEM rather
6221 than Unicode encoding.
6223 76. Added code to make require_files with a specific uid setting more usable in
6224 the case where statting the file as root fails - usually a non-root-mounted
6225 NFS file system. When this happens and the failure is EACCES, Exim now
6226 forks a subprocess and does the per-uid checking as the relevant uid.
6228 77. Added process_log_path.
6230 78. If log_file_path was not explicitly set, a setting of check_log_space or
6231 check_log_inodes was ignored.
6233 79. If a space check for the spool or log partitions fails, the incident is now
6234 logged. Of course, in the latter case the data may get lost...
6236 80. Added the %p formatting code to string_format() so that it can be used to
6237 print addresses in debug_print(). Adjusted all the address printing in the
6238 debugging in store.c to use %p rather than %d.
6240 81. There was a concern that a line of code in smtp_in.c could overflow a
6241 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6242 initially expressed, the concern was not well-founded, because trailing
6243 spaces are removed early. However, if the trailing spaces were followed by
6244 a NULL, they did not get removed, so the overflow was possible. Two fixes
6247 (a) I re-wrote the offending code in a cleaner fashion.
6248 (b) If an incoming SMTP command contains a NULL character, it is rejected
6251 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6252 runs initgroups(), so that if the Exim user is in any additional groups,
6253 they will be used during message reception.
6259 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob