1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and backup-only
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
74 TL/01 When running the test suite, the README says that variables such as
75 no_msglog_check are global and can be placed anywhere in a specific
76 test's script, however it was observed that placement needed to be near
77 the beginning for it to behave that way. Changed the runtest perl
78 script to read through the entire script once to detect and set these
79 variables, reset to the beginning of the script, and then run through
80 the script parsing/test process like normal.
82 TL/02 The BSD's have an arc4random API. One of the functions to induce
83 adding randomness was arc4random_stir(), but it has been removed in
84 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
85 function when detected.
87 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
88 cause callback expansion.
90 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
91 syntax errors in an expansion can be treated as a string instead of
92 logging or causing an error, due to the internal use of bool_lax
93 instead of bool when processing it.
95 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
96 server certificates when making smtp deliveries.
98 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
100 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
102 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
104 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
105 Merged patch from Sebastian Wiedenroth.
107 JH/05 Fix results-pipe from transport process. Several recipients, combined
108 with certificate use, exposed issues where response data items split
109 over buffer boundaries were not parsed properly. This eventually
110 resulted in duplicates being sent. This issue only became common enough
111 to notice due to the introduction of conection certificate information,
112 the item size being so much larger. Found and fixed by Wolfgang Breyha.
114 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
115 size buffer was used, resulting in syntax errors when an expansion
118 JH/07 Add support for directories of certificates when compiled with a GnuTLS
119 version 3.3.6 or later.
121 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
122 is EXPERIMENTAL_EVENT, the main-configuration and transport options
123 both become "event_action", the variables become $event_name, $event_data
124 and $event_defer_errno. There is a new variable $verify_mode, usable in
125 routers, transports and related events. The tls:cert event is now also
126 raised for inbound connections, if the main configuration event_action
129 TL/06 In test suite, disable OCSP for old versions of openssl which contained
130 early OCSP support, but no stapling (appears to be less than 1.0.0).
132 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
133 server certificate names available under the smtp transport option
134 "tls_verify_cert_hostname" now do not permit multi-component wildcard
137 JH/10 Time-related extraction expansions from certificates now use the main
138 option "timezone" setting for output formatting, and are consistent
139 between OpenSSL and GnuTLS compilations. Bug 1541.
141 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
142 encoded parameter in the incoming message. Bug 1558.
144 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
145 include certificate info, eximon was claiming there were spoolfile
148 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
150 JH/14 Log delivery-related information more consistently, using the sequence
151 "H=<name> [<ip>]" wherever possible.
153 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
154 are problematic for Debian distribution, omit them from the release
157 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
159 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
162 JH/17 Fix a null-indirection in certextract expansions when a nondefault
163 output list separator was used.
168 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
169 checkers that were complaining about end of non-void function with no
172 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
173 This was a regression intruduced in 4.83 by another bugfix.
175 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
177 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
178 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
184 TF/01 Correctly close the server side of TLS when forking for delivery.
186 When a message was received over SMTP with TLS, Exim failed to clear up
187 the incoming connection properly after forking off the child process to
188 deliver the message. In some situations the subsequent outgoing
189 delivery connection happened to have the same fd number as the incoming
190 connection previously had. Exim would try to use TLS and fail, logging
191 a "Bad file descriptor" error.
193 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
194 utilities have not been installed.
196 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
197 temporary space as the ACL may create new global variables.
199 TL/01 LDAP support uses per connection or global context settings, depending
200 upon the detected version of the libraries at build time.
202 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
203 to extract and use the src ip:port in logging and expansions as if it
204 were a direct connection from the outside internet. PPv2 support was
205 updated based on HAProxy spec change in May 2014.
207 JH/02 Add ${listextract {number}{list}{success}{fail}}.
209 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
210 Properly escape header and check for NULL return.
212 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
215 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
217 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
218 characters in header names, implemented as a verify condition.
219 Contributed by Michael Fischer v. Mollard.
221 TL/05 Rename SPF condition results err_perm and err_temp to standardized
222 results permerror and temperror. Previous values are deprecated but
223 still accepted. In a future release, err_perm and err_temp will be
224 completely removed, which will be a backward incompatibility if the
225 ACL tests for either of these two old results. Patch contributed by
226 user bes-internal on the mailing list.
228 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
230 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
231 selectors, in both main and reject logs.
233 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
236 JH/07 Add malware type "sock" for talking to simple daemon.
238 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
240 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
241 routers/transports under cutthrough routing.
243 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
244 numbers. Touch up "bool" conditional to keep the same definition.
246 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
248 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
250 JH/12 Expand items in router/transport headers_add or headers_remove lists
251 individually rather than the list as a whole. Bug 1452.
253 Required for reasonable handling of multiple headers_ options when
254 they may be empty; requires that headers_remove items with embedded
255 colons must have them doubled (or the list-separator changed).
257 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
258 view the policy declared in the DMARC record. Currently, $dmarc_status
259 is a combined value of both the record presence and the result of the
262 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
264 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
265 dnslookup router and the smtp transport (applying to the forward
268 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
269 of ldap servers used for a specific lookup. Patch provided by Heiko
272 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
273 New variable $lookup_dnssec_authenticated for observability.
275 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
276 Patch submitted by Lars Timman.
278 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
280 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
281 Requires trusted mode and valid format message id, aborts otherwise.
282 Patch contributed by Heiko Schlichting.
284 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
285 certextract with support for various fields. Bug 1358.
287 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
288 is requested by default, modifiable by smtp transport option
291 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
292 operate on certificate variables to give certificate fingerprints
293 Also new ${sha256:cert_variable}.
295 JH/23 The PRDR feature is moved from being Experimental into the mainline.
297 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
298 Christian Aistleitner.
300 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
302 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
303 file. Patch from Wolfgang Breyha.
305 JH/25 Expand the coverage of the delivery $host and $host_address to
306 client authenticators run in verify callout. Bug 1476.
308 JH/26 Port service names are now accepted for tls_on_connect_ports, to
309 align with daemon_smtp_ports. Bug 72.
311 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
312 support and error reporting did not work properly.
314 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
315 and is readable. Patch from Andrew Colin Kissa.
317 TL/14 Enhance documentation of ${run expansion and how it parses the
318 commandline after expansion, particularly in the case when an
319 unquoted variable expansion results in an empty value.
321 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
323 PP/02 Fix internal collision of T_APL on systems which support RFC3123
324 by renaming away from it. Addresses GH issue 15, reported by
327 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
329 TL/15 SECURITY: prevent double expansion in math comparison functions
330 (can expand unsanitized data). Not remotely exploitable.
337 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
339 PP/02 Make -n do something, by making it not do something.
340 When combined with -bP, the name of an option is not output.
342 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
345 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
346 $sender_host_name and config options to manage this, and basic check
349 PP/05 DSCP support for outbound connections and control modifier for inbound.
351 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
352 (Only plugin which currently uses this is kerberos4, which nobody should
353 be using, but we should make it available and other future plugins might
354 conceivably use it, even though it would break NAT; stuff *should* be
355 using channel bindings instead).
357 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
358 name; added for Sendmail compatibility; requires admin caller.
359 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
360 just ignore it); requires trusted caller.
361 Also parse but ignore: -Ac -Am -X<logfile>
364 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
366 TL/02 Add +smtp_confirmation as a default logging option.
368 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
369 Patch by Magnus Holmgren from 2007-02-20.
371 TL/04 Bugzilla 1281 - Spec typo.
372 Bugzilla 1283 - Spec typo.
373 Bugzilla 1290 - Spec grammar fixes.
375 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
377 TL/06 Add Experimental DMARC support using libopendmarc libraries.
379 TL/07 Fix an out of order global option causing a segfault. Reported to dev
380 mailing list by by Dmitry Isaikin.
382 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
384 JH/02 Support "G" suffix to numbers in ${if comparisons.
386 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
388 NM/01 Bugzilla 1197 - Spec typo
389 Bugzilla 1196 - Spec examples corrections
391 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
393 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
394 gnutls_enable_pkcs11, but renamed to more accurately indicate its
397 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
398 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
400 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
401 "acl {{name}{arg}...}", and optional args on acl condition
404 JH/05 Permit multiple router/transport headers_add/remove lines.
406 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
408 JH/07 Avoid using a waiting database for a single-message-only transport.
409 Performance patch from Paul Fisher. Bugzilla 1262.
411 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
414 JH/09 Add $headers_added variable, with content from use of ACL modifier
415 add_header (but not yet added to the message). Bugzilla 199.
417 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
418 Pulled from Bugzilla 817 by Wolfgang Breyha.
420 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
422 (nb: this is the same fix as in Exim 4.80.1)
424 JH/11 Add A= logging on delivery lines, and a client_set_id option on
427 JH/12 Add optional authenticated_sender logging to A= and a log_selector
430 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
432 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
433 advertise SMTP AUTH mechanism to us, instead of a generic
434 protocol violation error. Also, make Exim more robust to bad
435 data from the Dovecot auth socket.
437 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
439 When a queue runner is handling a message, Exim first routes the
440 recipient addresses, during which it prunes them based on the retry
441 hints database. After that it attempts to deliver the message to
442 any remaining recipients. It then updates the hints database using
445 So if a recipient address works intermittently, it can get repeatedly
446 deferred at routing time. The retry hints record remains fresh so the
447 address never reaches the final cutoff time.
449 This is a fairly common occurrence when a user is bumping up against
450 their storage quota. Exim had some logic in its local delivery code
451 to deal with this. However it did not apply to per-recipient defers
452 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
454 This change adds a proper retry rule check during routing so that the
455 final cutoff time is checked against the message's age. We only do
456 this check if there is an address retry record and there is not a
457 domain retry record; this implies that previous attempts to handle
458 the address had the retry_use_local_parts option turned on. We use
459 this as an approximation for the destination being like a local
460 delivery, as in LMTP.
462 I suspect this new check makes the old local delivery cutoff check
463 redundant, but I have not verified this so I left the code in place.
465 TF/02 Correct gecos expansion when From: is a prefix of the username.
467 Test 0254 submits a message to Exim with the header
471 When I ran the test suite under the user fanf2, Exim expanded
472 the header to contain my full name, whereas it should have added
473 a Resent-Sender: header. It erroneously treats any prefix of the
474 username as equal to the username.
476 This change corrects that bug.
478 GF/01 DCC debug and logging tidyup
479 Error conditions log to paniclog rather than rejectlog.
480 Debug lines prefixed by "DCC: " to remove any ambiguity.
482 TF/03 Avoid unnecessary rebuilds of lookup-related code.
484 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
485 Bug spotted by Jeremy Harris; was flawed since initial commit.
486 Would have resulted in OCSP responses post-SNI triggering an Exim
487 NULL dereference and crash.
489 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
491 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
492 Bug detection, analysis and fix by Samuel Thibault.
493 Bugzilla 1331, Debian bug #698092.
495 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
497 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
498 Server implementation by Todd Lyons, client by JH.
499 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
500 config variable "prdr_enable" controls whether the server
501 advertises the facility. If the client requests PRDR a new
502 acl_data_smtp_prdr ACL is called once for each recipient, after
503 the body content is received and before the acl_smtp_data ACL.
504 The client is controlled by bolth of: a hosts_try_prdr option
505 on the smtp transport, and the server advertisement.
506 Default client logging of deliveries and rejections involving
507 PRDR are flagged with the string "PRDR".
509 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
510 fclose(). Diagnosis by Todd Lyons.
512 PP/17 Update configure.default to handle IPv6 localhost better.
513 Patch by Alain Williams (plus minor tweaks).
516 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
517 This is now consistent with GnuTLS, and is now documented: the
518 previous undocumented portable approach to treating the option as
519 unset was to force an expansion failure. That still works, and
520 an empty string is now equivalent.
522 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
523 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
524 not performing validation itself.
526 PP/20 Added force_command boolean option to pipe transport.
527 Patch from Nick Koston, of cPanel Inc.
529 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
532 TF/04 Added udpsend ACL modifer and hexquote expansion operator
534 PP/21 Fix eximon continuous updating with timestamped log-files.
535 Broken in a format-string cleanup in 4.80, missed when I repaired the
536 other false fix of the same issue.
537 Report and fix from Heiko Schlichting.
540 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
541 Report from Prashanth Katuri.
543 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
544 It's SecureTransport, so affects any MacOS clients which use the
545 system-integrated TLS libraries, including email clients.
547 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
548 using a MIME ACL for non-SMTP local injection.
549 Report and assistance in diagnosis by Warren Baker.
551 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
553 JH/16 Fix comparisons for 64b. Bugzilla 1385.
555 TL/09 Add expansion variable $authenticated_fail_id to keep track of
556 last id that failed so it may be referenced in subsequent ACL's.
558 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
561 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
562 ldap library initialization, allowing self-signed CA's to be
563 used. Also properly sets require_cert option later in code by
564 using NULL (global ldap config) instead of ldap handle (per
565 session). Bug diagnosis and testing by alxgomz.
567 TL/12 Enhanced documentation in the ratelimit.pl script provided in
568 the src/util/ subdirectory.
570 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
571 renamed to Transport Post Delivery Action by Jeremy Harris, as
574 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
575 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
576 redis_servers = needs to be configured which will be used by the redis
577 lookup. Patch from Warren Baker, of The Packet Hub.
579 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
581 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
582 hostname or reverse DNS when processing a host list. Used suggestions
583 from multiple comments on this bug.
585 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
587 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
588 Missed a few lines, added it to make the runtest require no keyboard
591 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
592 contains upper case chars. Make router use caseful_local_part.
594 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
595 support when GnuTLS has been built with p11-kit.
601 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
603 This, or similar/improved, will also be change PP/11 of 4.82.
609 PP/01 Handle short writes when writing local log-files.
610 In practice, only affects FreeBSD (8 onwards).
611 Bugzilla 1053, with thanks to Dmitry Isaikin.
613 NM/01 Bugzilla 949 - Documentation tweak
615 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
618 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
620 PP/02 Implemented gsasl authenticator.
622 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
624 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
625 `pkg-config foo` for cflags/libs.
627 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
628 with rest of GSASL and with heimdal_gssapi.
630 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
631 `pkg-config foo` for cflags/libs for the TLS implementation.
633 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
634 properties get this fed in as external SSF. A number of robustness
635 and debugging improvements to the cyrus_sasl authenticator.
637 PP/08 cyrus_sasl server now expands the server_realm option.
639 PP/09 Bugzilla 1214 - Log authentication information in reject log.
640 Patch by Jeremy Harris.
642 PP/10 Added dbmjz lookup type.
644 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
646 PP/12 MAIL args handles TAB as well as SP, for better interop with
647 non-compliant senders.
648 Analysis and variant patch by Todd Lyons.
650 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
651 Bug report from Lars Müller <lars@samba.org> (via SUSE),
652 Patch from Dirk Mueller <dmueller@suse.com>
654 PP/13 tls_peerdn now print-escaped for spool files.
655 Observed some $tls_peerdn in wild which contained \n, which resulted
656 in spool file corruption.
658 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
659 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
660 or write after TLS renegotiation, which otherwise led to messages
663 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
664 as a tracking header (ie: a signed header comes before the signature).
665 Patch from Wolfgang Breyha.
667 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
668 comma-sep list; embedded commas doubled.
670 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
672 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
674 Report and patch from Dmitry Banschikov.
676 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
677 Removed SSL_clear() after SSL_new() which led to protocol negotiation
678 failures. We appear to now support TLS1.1+ with Exim.
680 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
681 lets Exim select keys and certificates based upon TLS SNI from client.
682 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
683 before an outbound SMTP session. New log_selector, +tls_sni.
685 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
686 NULL dereference. Report and patch from Alun Jones.
688 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
689 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
690 Not seeing resolver debug output on NetBSD, but suspect this is a
691 resolver implementation change.
693 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
694 Left warnings. Added "eximon gdb" invocation mode.
696 PP/21 Defaulting "accept_8bitmime" to true, not false.
698 PP/22 Added -bw for inetd wait mode support.
700 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
701 locate the relevant includes and libraries. Made this the default.
703 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
704 Bugzilla 1246, report and most of solution from Tomasz Kusy.
706 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
707 This may cause build issues on older platforms.
709 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
710 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
711 gnutls_require_mac & gnutls_require_protocols (no longer supported).
712 Added SNI support via GnuTLS too.
713 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
715 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
717 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
718 Applied second patch from Janne, implementing suggestion to default
719 multiple-strings-in-record handling to match SPF spec.
721 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
723 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
724 read-only, out of scope).
725 Patch from Wolfgang Breyha, report from Stuart Northfield.
727 PP/29 Fix three issues highlighted by clang analyser static analysis.
728 Only crash-plausible issue would require the Cambridge-specific
729 iplookup router and a misconfiguration.
730 Report from Marcin Mirosław.
732 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
734 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
735 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
736 As part of this, removing so much warning spew let me fix some minor
737 real issues in debug logging.
739 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
740 assignment on my part. Fixed.
742 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
743 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
744 Janne Snabb (who went above and beyond: thank you).
746 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
747 string otherwise requires a connection and a bunch more work and it's
748 relatively easy to get wrong. Should also expose TLS library linkage
751 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
752 64-bit ${eval} (JH/03).
754 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
755 GNU libc to support some of the 64-bit stuff, should not lead to
756 conflicts. Defined before os.h is pulled in, so if a given platform
757 needs to override this, it can.
759 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
760 protection layer was required, which is not implemented.
761 Bugzilla 1254, patch from Wolfgang Breyha.
763 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
764 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
765 tls_dhparam take prime identifiers. Also unbreak combination of
766 OpenSSL+DH_params+TLSSNI.
768 PP/39 Disable SSLv2 by default in OpenSSL support.
774 PP/01 Solaris build fix for Oracle's LDAP libraries.
775 Bugzilla 1109, patch from Stephen Usher.
777 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
779 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
782 TF/02 Fix a couple more cases where we did not log the error message
783 when unlink() failed. See also change 4.74-TF/03.
785 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
786 lock up or crash if it happened to be inside a call to libc when it
787 got a SIGUSR1 from exiwhat.
789 The SIGUSR1 handler appends the current process status to the process
790 log which is later printed by exiwhat. It used to use the general
791 purpose logging code to do this, but several functions it calls are
792 not safe for signals.
794 The new output code in the SIGUSR1 handler is specific to the process
795 log, and simple enough that it's easy to inspect for signal safety.
796 Removing some special cases also simplifies the general logging code.
797 Removing the spurious timestamps from the process log simplifies
800 TF/04 Improved ratelimit ACL condition.
802 The /noupdate option has been deprecated in favour of /readonly which
803 has clearer semantics. The /leaky, /strict, and /readonly update modes
804 are mutually exclusive. The update mode is no longer included in the
805 database key; it just determines when the database is updated. (This
806 means that when you upgrde Exim will forget old rate measurements.)
808 Exim now checks that the per_* options are used with an update mode that
809 makes sense for the current ACL. For example, when Exim is processing a
810 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
811 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
812 must specify per_mail/readonly. If you omit the update mode it defaults to
813 /leaky where that makes sense (as before) or /readonly where required.
815 The /noupdate option is now undocumented but still supported for
816 backwards compatibility. It is equivalent to /readonly except that in
817 ACLs where /readonly is required you may specify /leaky/noupdate or
818 /strict/noupdate which are treated the same as /readonly.
820 A useful new feature is the /count= option. This is a generalization
821 of the per_byte option, so that you can measure the throughput of other
822 aggregate values. For example, the per_byte option is now equivalent
823 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
825 The per_rcpt option has been generalized using the /count= mechanism
826 (though it's more complicated than the per_byte equivalence). When it is
827 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
828 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
829 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
830 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
831 Note that using per_rcpt with a non-readonly update mode in more than
832 one ACL will cause the recipients to be double-counted. (The per_mail
833 and per_byte options don't have this problem.)
835 The handling of very low rates has changed slightly. If the computed rate
836 is less than the event's count (usually one) then this event is the first
837 after a long gap. In this case the rate is set to the same as this event's
838 count, so that the first message of a spam run is counted properly.
840 The major new feature is a mechanism for counting the rate of unique
841 events. The new per_addr option counts the number of different
842 recipients that someone has sent messages to in the last time period. It
843 behaves like per_rcpt if all the recipient addresses are different, but
844 duplicate recipient addresses do not increase the measured rate. Like
845 the /count= option this is a general mechanism, so the per_addr option
846 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
847 example, measure the rate that a client uses different sender addresses
848 with the options per_mail/unique=$sender_address. There are further
849 details in the main documentation.
851 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
853 TF/06 Removed a few PCRE remnants.
855 TF/07 Automatically extract Exim's version number from tags in the git
856 repository when doing development or release builds.
858 PP/02 Raise smtp_cmd_buffer_size to 16kB.
859 Bugzilla 879. Patch from Paul Fisher.
861 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
862 Heavily based on revision 40f9a89a from Simon Arlott's tree.
865 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
867 PP/05 Variable $av_failed, true if the AV scanner deferred.
868 Bugzilla 1078. Patch from John Horne.
870 PP/06 Stop make process more reliably on build failure.
871 Bugzilla 1087. Patch from Heiko Schlittermann.
873 PP/07 Make maildir_use_size_file an _expandable_ boolean.
874 Bugzilla 1089. Patch from Heiko Schlittermann.
876 PP/08 Handle ${run} returning more data than OS pipe buffer size.
877 Bugzilla 1131. Patch from Holger Weiß.
879 PP/09 Handle IPv6 addresses with SPF.
880 Bugzilla 860. Patch from Wolfgang Breyha.
882 PP/10 GnuTLS: support TLS 1.2 & 1.1.
884 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
887 PP/11 match_* no longer expand right-hand-side by default.
888 New compile-time build option, EXPAND_LISTMATCH_RHS.
889 New expansion conditions, "inlist", "inlisti".
891 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
893 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
895 PP/14 fix log_write() format string regression from TF/03.
896 Bugzilla 1152. Patch from Dmitry Isaikin.
902 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
904 PP/02 Harmonised TLS library version reporting; only show if debugging.
905 Layout now matches that introduced for other libraries in 4.74 PP/03.
907 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
909 PP/04 New "dns_use_edns0" global option.
911 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
914 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
915 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
917 TK/01 Updated PolarSSL code to 0.14.2.
918 Bugzilla 1097. Patch from Andreas Metzler.
920 PP/07 Catch divide-by-zero in ${eval:...}.
923 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
926 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
927 format-string attack -- SECURITY: remote arbitrary code execution.
929 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
930 time unintentionally subject to list matching rules, letting the header
931 cause arbitrary Exim lookups (of items which can occur in lists, *not*
932 arbitrary string expansion). This allowed for information disclosure.
934 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
935 INT_MIN/-1 -- value coerced to INT_MAX.
941 NM/01 Workround for PCRE version dependency in version reporting
944 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
945 This fixes portability to compilers other than gcc, notably
946 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
948 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
949 makefiles for portability to HP-UX and POSIX correctness.
951 PP/01 Permit LOOKUP_foo enabling on the make command-line.
952 Also via indented variable definition in the Makefile.
953 (Debugging by Oliver Heesakkers).
955 PP/02 Restore caching of spamd results with expanded spamd_address.
956 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
958 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
959 Improves build reliability. Fix from: Frank Elsner
961 NM/02 Fix wide character breakage in the rfc2047 coding
962 Fixes bug 1064. Patch from Andrey N. Oktyabrski
964 NM/03 Allow underscore in dnslist lookups
965 Fixes bug 1026. Patch from Graeme Fowler
967 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
968 Code patches from Adam Ciarcinski of NetBSD.
970 NM/04 Fixed exiqgrep to cope with mailq missing size issue
973 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
974 is logged, to avoid truncation. Patch from John Horne.
976 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
977 Patch from Jakob Hirsch.
979 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
980 SQL string expansion failure details.
981 Patch from Andrey Oktyabrski.
983 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
984 Patch from Simon Arlott.
986 PP/09 New lookups functionality failed to compile on old gcc which rejects
987 extern declarations in function scope.
988 Patch from Oliver Fleischmann
990 PP/10 Use sig_atomic_t for flags set from signal handlers.
991 Check getgroups() return and improve debugging.
992 Fixed developed for diagnosis in bug 927 (which turned out to be
995 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
996 Patch from Mark Zealey.
998 PP/12 Bugzilla 1056: Improved spamd server selection.
999 Patch from Mark Zealey.
1001 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1002 Based on patch from Heiko Schlittermann.
1004 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1005 Patch from Uwe Doering, sign-off by Michael Haardt.
1007 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1008 variable declaration deep within a block. Bug and patch from
1011 PP/15 lookups-Makefile IRIX compatibilty coercion.
1013 PP/16 Make DISABLE_DKIM build knob functional.
1015 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1016 Patch by Simon Arlott
1018 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1019 variable argument macros. Our copy now differs from upstream.
1025 TF/01 Failure to get a lock on a hints database can have serious
1026 consequences so log it to the panic log.
1028 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1029 controlled using the smtp_confirmation log selector.
1031 TF/03 Include the error message when we fail to unlink a spool file.
1033 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1034 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1035 for maintaining out-of-tree patches for some time.
1037 PP/01 Bugzilla 139: Documentation and portability issues.
1038 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1039 Handle per-OS dynamic-module compilation flags.
1041 PP/02 Let /dev/null have normal permissions.
1042 The 4.73 fixes were a little too stringent and complained about the
1043 permissions on /dev/null. Exempt it from some checks.
1044 Reported by Andreas M. Kirchwitz.
1046 PP/03 Report version information for many libraries, including
1047 Exim version information for dynamically loaded libraries. Created
1048 version.h, now support a version extension string for distributors
1049 who patch heavily. Dynamic module ABI change.
1051 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1052 privilege escalation vulnerability whereby the Exim run-time user
1053 can cause root to append content of the attacker's choosing to
1056 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1059 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1060 If dropping privileges for untrusted macros, we disabled normal logging
1061 on the basis that it would fail; for the Exim run-time user, this is not
1062 the case, and it resulted in successful deliveries going unlogged.
1063 Fixed. Reported by Andreas Metzler.
1069 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1070 only prepend for the Resent-* case. Fixes regression introduced in
1071 Exim 4.70 by NM/22 for Bugzilla 607.
1073 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1074 increasing numbers of administrators be bitten by this.
1076 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1078 PP/03 Bugzilla 994: added openssl_options main configuration option.
1080 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1082 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1084 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1086 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1087 without a peer certificate, leading to a segfault because of an
1088 assumption that peers always have certificates. Be a little more
1089 paranoid. Problem reported by Martin Tscholak.
1091 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1092 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1093 NB: ClamAV planning to remove STREAM in "middle of 2010".
1094 CL also introduces -bmalware, various -d+acl logging additions and
1095 more caution in buffer sizes.
1097 PP/09 Implemented reverse_ip expansion operator.
1099 PP/10 Bugzilla 937: provide a "debug" ACL control.
1101 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1103 PP/12 Bugzilla 973: Implement --version.
1105 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1107 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1109 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1111 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1112 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1113 ignore trailing whitespace.
1115 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1117 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1120 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1121 Notification from Dr Andrew Aitchison.
1123 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1124 ExtendedDetectionInfo response format.
1125 Notification from John Horne.
1127 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1130 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1131 XSL and documented dependency on system catalogs, with examples of how
1134 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1137 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1138 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1139 configuration file which is writeable by the Exim user or group.
1141 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1142 of configuration files to cover files specified with the -C option if
1143 they are going to be used with root privileges, not just the default
1146 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1147 option (effectively making it always true).
1149 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1150 files to be used while preserving root privileges.
1152 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1153 that rogue child processes cannot use them.
1155 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1156 run-time user, instead of root.
1158 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1159 Exim run-time user without dropping privileges.
1161 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1162 result string, instead of calling string_vformat() twice with the same
1165 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1166 for other users. Others should always drop root privileges if they use
1167 -C on the command line, even for a whitelisted configure file.
1169 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1171 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1177 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1178 $data_path, and $header_path variables; fixed documentation bugs and
1181 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1182 exipick to access non-standard spools, including the "frozen" queue
1185 NM/01 Bugzilla 965: Support mysql stored procedures.
1186 Patch from Alain Williams
1188 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1190 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1191 Patch from Andreas Metzler
1193 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1194 Patch from Kirill Miazine
1196 NM/05 Bugzilla 671: Added umask to procmail example.
1198 JJ/03 installed exipick 20100323.0, fixing doc bug
1200 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1201 directory. Notification and patch from Dan Rosenberg.
1203 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1205 TK/02 Improve log output when DKIM signing operation fails.
1207 MH/01 Treat the transport option dkim_domain as a colon separated
1208 list, not as a single string, and sign the message with each element,
1209 omitting multiple occurences of the same signer.
1211 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1212 Bugzilla 985, 986. Patch by Simon Arlott
1214 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1215 Patch by Simon Arlott
1217 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1218 MBX locking. Notification from Dan Rosenberg.
1224 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1226 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1228 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1230 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1232 NM/04 Bugzilla 915: Flag broken perl installation during build.
1238 TK/01 Added patch by Johannes Berg that expands the main option
1239 "spamd_address" if it starts with a dollar sign.
1241 TK/02 Write list of recipients to X-Envelope-Sender header when building
1242 the mbox-format spool file for content scanning (suggested by Jakob
1245 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1246 (http://www.dcc-servers.net/) support via dccifd. Activated by
1247 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1249 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1250 by Mark Daniel Reidel <mr@df.eu>.
1252 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1253 When building exim an external PCRE library is now needed -
1254 PCRE is a system library on the majority of modern systems.
1255 See entry on PCRE_LIBS in EDITME file.
1257 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1258 conversation. Added nologin parameter to request.
1259 Patch contributed by Kirill Miazine.
1261 TF/01 Do not log submission mode rewrites if they do not change the address.
1263 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1265 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1266 log files in place. Contributed by Roberto Lima.
1268 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1270 TF/03 Bugzilla 615: When checking the local_parts router precondition
1271 after a local_part_suffix or local_part_prefix option, Exim now
1272 does not use the address's named list lookup cache, since this
1273 contains cached lookups for the whole local part.
1275 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1276 Robert Millan. Documentation is in experimental-spec.txt.
1278 TF/04 Bugzilla 668: Fix parallel build (make -j).
1280 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1282 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1283 Patch provided by Jan Srzednicki.
1285 TF/05 Leading white space used to be stripped from $spam_report which
1286 wrecked the formatting. Now it is preserved.
1288 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1289 that they are available at delivery time.
1291 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1293 TF/08 TLS error reporting now respects the incoming_interface and
1294 incoming_port log selectors.
1296 TF/09 Produce a more useful error message if an SMTP transport's hosts
1297 setting expands to an empty string.
1299 NM/06 Bugzilla 744: EXPN did not work under TLS.
1300 Patch provided by Phil Pennock.
1302 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1303 Patch provided by Richard Godbee.
1305 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1306 acl_smtp_notquit, added index entry.
1308 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1309 Patch provided by Eugene Bujak.
1311 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1312 accept(). Patch provided by Maxim Dounin.
1314 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1315 Patch provided by Phil Pennock.
1317 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1319 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1320 Patch provided by Brad "anomie" Jorsch.
1322 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1323 Patch provided by Dean Brooks.
1325 TK/05 Add native DKIM support (does not depend on external libraries).
1327 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1328 Patch provided by Graeme Fowler.
1330 NM/16 Bugzilla 851: Documentation example syntax fix.
1332 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1334 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1337 NM/19 Bugzilla 745: TLS version reporting.
1338 Patch provided by Phil Pennock.
1340 NM/20 Bugzilla 167: bool: condition support.
1341 Patch provided by Phil Pennock.
1343 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1344 clients. Patch provided by Phil Pennock.
1346 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1347 Patch provided by Brad "anomie" Jorsch.
1349 NM/23 Bugzilla 687: Fix misparses in eximstats.
1350 Patch provided by Heiko Schlittermann.
1352 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1353 Patch provided by Heiko Schlittermann.
1355 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1356 plus update to original patch.
1358 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1360 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1361 Patch provided by David Brownlee.
1363 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1365 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1367 NM/30 Bugzilla 888: TLS documentation bugfixes.
1369 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1371 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1372 Unlike the original bugzilla I have changed all shell scripts in src tree.
1374 NM/33 Bugzilla 898: Transport filter timeout fix.
1375 Patch by Todd Rinaldo.
1377 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1378 Patch by Serge Demonchaux.
1380 NM/35 Bugzilla 39: Base64 decode bug fixes.
1381 Patch by Jakob Hirsch.
1383 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1385 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1387 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1389 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1395 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1396 ALT-N's libdkim that I have put here:
1397 http://duncanthrax.net/exim-experimental/
1399 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1400 were called 'true' and it seems that C99 defines that as a reserved
1401 keyword to be used with 'bool' variable types. That means you could
1402 not include C99-style headers which use bools without triggering
1403 build errors in sieve.c.
1405 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1406 as mailq or other aliases. Changed the --help handling significantly
1407 to do whats expected. exim_usage() emits usage/help information.
1409 SC/01 Added the -bylocaldomain option to eximstats.
1411 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1413 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1415 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1421 PH/01 Another patch from the Sieve maintainer.
1423 PH/02 When an IPv6 address is converted to a string for single-key lookup
1424 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1425 dots are used instead of colons so that keys in lsearch files need not
1426 contain colons. This was done some time before quoting was made available
1427 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1428 (notated using the quote facility) so as to distinguish them from IPv4
1429 keys. This meant that lookups for IP addresses in host lists did not work
1430 for iplsearch lookups.
1432 This has been fixed by arranging for IPv6 addresses to be expressed with
1433 colons if the lookup type is iplsearch. This is not incompatible, because
1434 previously such lookups could never work.
1436 The situation is now rather anomolous, since one *can* have colons in
1437 ordinary lsearch keys. However, making the change in all cases is
1438 incompatible and would probably break a number of configurations.
1440 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1443 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1444 conversion specification without a maximum field width, thereby enabling
1445 a rogue spamd server to cause a buffer overflow. While nobody in their
1446 right mind would setup Exim to query an untrusted spamd server, an
1447 attacker that gains access to a server running spamd could potentially
1448 exploit this vulnerability to run arbitrary code as the Exim user.
1450 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1451 $primary_hostname instead of what libspf2 thinks the hosts name is.
1453 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1454 a directory entry by the name of the lookup key. Previously, if a
1455 symlink pointed to a non-existing file or a file in a directory that
1456 Exim lacked permissions to read, a lookup for a key matching that
1457 symlink would fail. Now it is enough that a matching directory entry
1458 exists, symlink or not. (Bugzilla 503.)
1460 PH/03 The body_linecount and body_zerocount variables are now exported in the
1463 PH/04 Added the $dnslist_matched variable.
1465 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1466 This means they are set thereafter only if the connection becomes
1469 PH/06 Added the client_condition to authenticators so that some can be skipped
1470 by clients under certain conditions.
1472 PH/07 The error message for a badly-placed control=no_multiline_responses left
1473 "_responses" off the end of the name.
1475 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1477 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1478 (without spaces) instead of just copying the configuration text.
1480 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1482 PH/11 Added $max_received_linelength.
1484 PH/12 Added +ignore_defer and +include_defer to host lists.
1486 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1487 way in which PCRE > 7.0 is built.
1489 PH/14 Implemented queue_only_load_latch.
1491 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1492 MAIL command. The effect was to mangle the value on 64-bit systems.
1494 PH/16 Another patch from the Sieve maintainer.
1496 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1498 PH/18 If a system quota error occurred while trying to create the file for
1499 a maildir delivery, the message "Mailbox is full" was not appended to the
1500 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1501 only to a quota excession during the actual writing of the file.
1503 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1504 characters?) which causes problems in log lines. The DN values are now
1505 passed through string_printing() before being added to log lines.
1507 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1508 and InterBase are left for another time.)
1510 PH/21 Added message_body_newlines option.
1512 PH/22 Guard against possible overflow in moan_check_errorcopy().
1514 PH/23 POSIX allows open() to be a macro; guard against that.
1516 PH/24 If the recipient of an error message contained an @ in the local part
1517 (suitably quoted, of course), incorrect values were put in $domain and
1518 $local_part during the evaluation of errors_copy.
1524 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1525 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1528 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1529 issue a MAIL command.
1531 PH/02 In an ACL statement such as
1533 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1535 if a client was not listed at all, or was listed with a value other than
1536 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1537 the condition was not true (as it should be), so access was not denied.
1538 The bug was that the ! inversion was incorrectly passed on to the second
1539 item. This has been fixed.
1541 PH/03 Added additional dnslists conditions == and =& which are different from
1542 = and & when the dns lookup returns more than one IP address.
1544 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1545 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1547 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1548 FSYNC, which compiles an option called disable_fsync that allows for
1549 bypassing fsync(). The documentation is heavily laced with warnings.
1551 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1553 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1554 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1555 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1556 including adding "make clean"; (3) Added -fPIC when compiling the test
1557 dynamically loaded module, to get rid of a warning.
1559 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1560 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1561 = 0s. The bug is otherwise harmless.
1563 PH/07 There was a bug in the dovecot authenticator such that the value of
1564 $auth1 could be overwritten, and so not correctly preserved, after a
1565 successful authentication. This usually meant that the value preserved by
1566 the server_setid option was incorrect.
1568 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1570 PH/09 Installed PCRE release 7.0.
1572 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1573 run for batched SMTP input. It is now run at the start of every message
1574 in the batch. While fixing this I discovered that the process information
1575 (output by running exiwhat) was not always getting set for -bs and -bS
1576 input. This is fixed, and it now also says "batched" for BSMTP.
1578 PH/11 Added control=no_pipelining.
1580 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1581 patch, slightly modified), and move the expansion of helo_data till after
1582 the connection is made in the smtp transport (so it can use these
1585 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1587 PH/14 Added log_selector = +pid.
1589 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1591 PH/16 Add ${if forany and ${if forall.
1593 PH/17 Added dsn_from option to vary the From: line in DSNs.
1595 PH/18 Flush SMTP output before performing a callout, unless control =
1596 no_callout_flush is set.
1598 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1599 was true (the default) a successful delivery failed to delete the retry
1600 item, thus causing premature timeout of the address. The bug is now
1603 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1605 PH/21 Long custom messages for fakedefer and fakereject are now split up
1606 into multiline reponses in the same way that messages for "deny" and
1607 other ACL rejections are.
1609 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1610 with slight modification.
1612 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1613 draft, changing the syntax and factoring some duplicate code".
1615 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1616 for deliveries of the second and subsequent messages over the same SMTP
1619 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1620 ${reduce, with only minor "tidies".
1622 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1624 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1625 expansion side effects.
1627 PH/27 When a message times out after an over-quota error from an Exim-imposed
1628 quota, the bounce message says "mailbox is full". This message was not
1629 being given when it was a system quota that was exceeded. It now should
1632 MH/03 Made $recipients available in local_scan(). local_scan() already has
1633 better access to the recipient list through recipients_list[], but
1634 $recipients can be useful in postmaster-provided expansion strings.
1636 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1637 in the case of a MAIL command with additional options following the
1638 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1639 were accidentally chopped off.
1641 PH/29 SMTP synchronization checks are implemented when a command is read -
1642 there is a check that no more input is waiting when there shouldn't be
1643 any. However, for some commands, a delay in an ACL can mean that it is
1644 some time before the response is written. In this time, more input might
1645 arrive, invalidly. So now there are extra checks after an ACL has run for
1646 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1647 pipelining has not been advertised.
1649 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1651 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1652 returned in the SMTP response when smtp_return_error_details was false.
1653 This has been fixed.
1655 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1656 instead of the C I/O that was originally supplied, because problems were
1657 reported on Solaris.
1659 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1660 Exim which did not show up earlier: it was assuming that a call to
1661 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1662 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1663 was a macro that became an assignment, so it seemed to work. This has
1664 changed to a proper function call with a void return, hence the compile
1665 error. Exim's code has been fixed.
1667 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1670 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1672 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1674 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1675 because it turns out to be too big or there is a timeout) there is no
1676 "Completed" line in the log. When some messages of this type were
1677 selected by exigrep, they were listed as "not completed". Others were
1678 picked up by some special patterns. I have improved the selection
1679 criteria to be more general.
1681 PH/38 The host_find_failed option in the manualroute router can now be set
1682 to "ignore", to completely ignore a host whose IP address cannot be
1683 found. If all hosts are ignored, the behaviour is controlled by the new
1684 host_all_ignored option.
1686 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1687 homing or because of multiple MX records with /mx) generated more than
1688 one IP address, and the following item turned out to be the local host,
1689 all the secondary addresses of the first item were incorrectly removed
1690 from the list, along with the local host and any following hosts (which
1691 is what is supposed to happen).
1693 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1694 whoever called Exim into the -H file. In the case of the daemon it was
1695 behaving confusingly. When first started, it used values for whoever
1696 started the daemon, but after a SIGHUP it used the Exim user (because it
1697 calls itself on a restart). I have changed the code so that it now always
1700 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1701 message are rejected with the same error (e.g. no authentication or bad
1702 sender address), and a DATA command is nevertheless sent (as can happen
1703 with PIPELINING or a stupid MUA), the error message that was given to the
1704 RCPT commands is included in the rejection of the DATA command. This is
1705 intended to be helpful for MUAs that show only the final error to their
1708 PH/42 Another patch from the Sieve maintainer.
1710 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1711 Eximstats - Fixed some broken HTML links and added missing column headers
1713 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1714 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1716 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1718 PH/43 Yet another patch from the Sieve maintainer.
1720 PH/44 I found a way to check for a TCP/IP connection going away before sending
1721 the response to the final '.' that terminates a message, but only in the
1722 case where the client has not sent further data following the '.'
1723 (unfortunately, this is allowed). However, in many cases there won't be
1724 any further data because there won't be any more messages to send. A call
1725 to select() can be used: if it shows that the input is "ready", there is
1726 either input waiting, or the socket has been closed. An attempt to read
1727 the next input character can distinguish the two cases. Previously, Exim
1728 would have sent an OK response which the client would never have see.
1729 This could lead to message repetition. This fix should cure that, at
1730 least in a lot of common cases.
1732 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1733 advertised in response to EHLO.
1739 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1740 fixed by 4.65/MH/01 (is this a record?) are fixed:
1742 (i) An empty string was always treated as zero by the numeric comparison
1743 operators. This behaviour has been restored.
1745 (ii) It is documented that the numeric comparison operators always treat
1746 their arguments as decimal numbers. This was broken in that numbers
1747 starting with 0 were being interpreted as octal.
1749 While fixing these problems I realized that there was another issue that
1750 hadn't been noticed. Values of message_size_limit (both the global option
1751 and the transport option) were treated as octal if they started with 0.
1752 The documentation was vague. These values are now always treated as
1753 decimal, and I will make that clear in the documentation.
1759 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1760 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1763 MH/01 Don't check that the operands of numeric comparison operators are
1764 integers when their expansion is in "skipping" mode (fixes bug
1765 introduced by 4.64-PH/07).
1767 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1768 child addresses, Exim now panics and dies. Previously, because the count
1769 is held in a short int, deliveries were likely to be lost. As such a
1770 large number of recipients for a single message is ridiculous
1771 (performance will be very, very poor), I have chosen to impose a limit
1772 rather than extend the field.
1778 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1779 leftover -K file (the existence of which was triggered by #402).
1780 While we were at it, introduced process PID as part of the -K
1781 filename. This should rule out race conditions when creating
1784 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1785 processing considerably. Previous code took too long for large mails,
1786 triggering a timeout which in turn triggers #401.
1788 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1789 in the DK code in transports.c. sendfile() is not really portable,
1790 hence the _LINUX specificness.
1792 TF/01 In the add_headers option to the mail command in an Exim filter,
1793 there was a bug that Exim would claim a syntax error in any
1794 header after the first one which had an odd number of characters
1797 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1798 callout verification, Exim cached a "reject" for the entire domain. This
1799 is correct for most verifications, but it is not correct for a recipient
1800 verification with use_sender or use_postmaster set, because in that case
1801 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1802 case of MAIL FROM:<> rejection from other early rejections (e.g.
1803 rejection of HELO). When verifying a recipient using a non-null MAIL
1804 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1805 Whatever the result of the callout, the value of the domain cache is
1806 left unchanged (for any other kind of callout, getting as far as trying
1807 RCPT means that the domain itself is ok).
1809 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1812 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1813 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1814 session, the daemon ignores ECONNECT errors and logs others; it now
1815 ignores EPIPE as well.
1817 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1818 (quoted-printable decoding).
1820 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1821 later the small subsequent patch to fix an introduced bug.
1823 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1825 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1827 PH/08 An error is now given if message_size_limit is specified negative.
1829 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1830 to be given (somewhat) arbitrary names.
1832 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1835 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1836 miscellaneous code fixes
1838 PH/10 Added the log_reject_target ACL modifier to specify where to log
1841 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1842 hostname. This is wrong, because it relates to the incoming message (and
1843 probably the interface on which it is arriving) and not to the outgoing
1844 callout (which could be using a different interface). This has been
1845 changed to use the value of the helo_data option from the smtp transport
1846 instead - this is what is used when a message is actually being sent. If
1847 there is no remote transport (possible with a router that sets up host
1848 addresses), $smtp_active_hostname is used.
1850 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1851 tweaks were necessary in order to get it to work (see also 21 below):
1852 (a) The code assumed that strncpy() returns a negative number on buffer
1853 overflow, which isn't the case. Replaced with Exim's string_format()
1855 (b) There were several signed/unsigned issues. I just did the minimum
1856 hacking in of casts. There is scope for a larger refactoring.
1857 (c) The code used strcasecmp() which is not a standard C function.
1858 Replaced with Exim's strcmpic() function.
1859 (d) The code set only $1; it now sets $auth1 as well.
1860 (e) A simple test gave the error "authentication client didn't specify
1861 service in request". It would seem that Dovecot has changed its
1862 interface. Fortunately there's a specification; I followed it and
1863 changed what the client sends and it appears to be working now.
1865 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1868 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1869 address A is aliased to B and C, where B exists and C does not. Without
1870 -v the output is "A verified" because verification stops after a
1871 successful redirection if more than one address is generated. However,
1872 with -v the child addresses are also verified. Exim was outputting "A
1873 failed to verify" and then showing the successful verification for C,
1874 with its parentage. It now outputs "B failed to verify", showing B's
1875 parentage before showing the successful verification of C.
1877 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1878 look up a TXT record in a specific list after matching in a combined
1881 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1882 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1883 they consult the DNS. I had assumed they would set it the way they
1884 wanted; and indeed my experiments on Linux seem to show that in some
1885 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1886 To be on the safe side, however, I have now made the interface to
1887 host_find_byname() similar to host_find_bydns(), with an argument
1888 containing the DNS resolver options. The host_find_byname() function now
1889 sets these options at its start, just as host_find_bydns() does. The smtp
1890 transport options dns_qualify_single and dns_search_parents are passed to
1891 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1892 of host_find_byname() use the default settings of RES_DEFNAMES
1893 (qualify_single) but not RES_DNSRCH (search_parents).
1895 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1896 spool_read_header() do less string testing, by means of a preliminary
1897 switch on the second character of optional "-foo" lines. (This is
1898 overdue, caused by the large number of possibilities that now exist.
1899 Originally there were few.) While I was there, I also converted the
1900 str(n)cmp tests so they don't re-test the leading "-" and the first
1901 character, in the hope this might squeeze out yet more improvement.
1903 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1904 flag allowing group syntax was set by the header_syntax check but not
1905 turned off, possible causing trouble later; (2) The flag was not being
1906 set at all for the header_verify test, causing "group"-style headers to
1907 be rejected. I have now set it in this case, and also caused header_
1908 verify to ignore an empty address taken from a group. While doing this, I
1909 came across some other cases where the code for allowing group syntax
1910 while scanning a header line wasn't quite right (mostly, not resetting
1911 the flag correctly in the right place). These bugs could have caused
1912 trouble for malformed header lines. I hope it is now all correct.
1914 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1915 with the "reply" argument non-NULL. The code, however (which originally
1916 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1917 but it didn't always do it. This confused somebody who was copying the
1918 code for some other use. I have removed all the tests.
1920 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1921 feature that was used to support insecure browsers during the U.S. crypto
1922 embargo. It requires special client support, and Exim is probably the
1923 only MTA that supported it -- and would never use it because real RSA is
1924 always available. This code has been removed, because it had the bad
1925 effect of slowing Exim down by computing (never used) parameters for the
1926 RSA_EXPORT functionality.
1928 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1929 authenticator to fail if there's a tab character in the incoming data
1930 (there should never be unless someone is messing about, as it's supposed
1931 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1932 if the connection is using TLS or if the remote IP is the same as the
1933 local IP, and the "valid-client-cert option" if a client certificate has
1936 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1937 authenticators. This can be used for authorization after authentication
1938 succeeds. (In the case of plaintext, it servers for both authentication
1941 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1942 if any retry times were supplied.
1944 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1945 connection, where there is no client IP address to check. In this
1946 situation, the verify now always succeeds.
1948 PH/25 Applied John Jetmore's -Mset patch.
1950 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1952 PH/27 In a string expansion for a processed (not raw) header when multiple
1953 headers of the same name were present, leading whitespace was being
1954 removed from all of them, but trailing whitespace was being removed only
1955 from the last one. Now trailing whitespace is removed from each header
1956 before concatenation. Completely empty headers in a concatenation (as
1957 before) are ignored.
1959 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1960 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1962 PH/29 [Removed. This was a change that I later backed out, and forgot to
1963 correct the ChangeLog entry (that I had efficiently created) before
1964 committing the later change.]
1966 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1967 address errors (4xx response to RCPT) over the same connection as other
1968 messages routed to the same hosts. Such deliveries are always "forced",
1969 so retry times are not inspected. This resulted in far too many retries
1970 for the affected addresses. The effect occurred only when there were more
1971 hosts than the hosts_max_try setting in the smtp transport when it had
1972 the 4xx errors. Those hosts that it had tried were not added to the list
1973 of hosts for which the message was waiting, so if all were tried, there
1974 was no problem. Two fixes have been applied:
1976 (i) If there are any address or message errors in an SMTP delivery, none
1977 of the hosts (tried or untried) are now added to the list of hosts
1978 for which the message is waiting, so the message should not be a
1979 candidate for sending over the same connection that was used for a
1980 successful delivery of some other message. This seems entirely
1981 reasonable: after all the message is NOT "waiting for some host".
1982 This is so "obvious" that I'm not sure why it wasn't done
1983 previously. Hope I haven't missed anything, but it can't do any
1984 harm, as the worst effect is to miss an optimization.
1986 (ii) If, despite (i), such a delivery is accidentally attempted, the
1987 routing retry time is respected, so at least it doesn't keep
1988 hammering the server.
1990 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1991 in ${readsocket because some servers need this prod.
1993 PH/32 Added some extra debug output when updating a wait-xxx database.
1995 PH/33 The hint "could be header name not terminated by colon", which has been
1996 given for certain expansion errors for a long time, was not being given
1997 for the ${if def:h_colon_omitted{... case.
1999 PH/34 The spec says: "With one important exception, whenever a domain list is
2000 being scanned, $domain contains the subject domain." There was at least
2001 one case where this was not true.
2003 PH/35 The error "getsockname() failed: connection reset by peer" was being
2004 written to the panic log as well as the main log, but it isn't really
2005 panic-worthy as it just means the connection died rather early on. I have
2006 removed the panic log writing for the ECONNRESET error when getsockname()
2009 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2010 runs only) independently of the message's sender address. This meant
2011 that, if the 4xx error was in fact related to the sender, a different
2012 message to the same recipient with a different sender could confuse
2013 things. In particualar, this can happen when sending to a greylisting
2014 server, but other circumstances could also provoke similar problems.
2015 I have changed the default so that the retry time for these errors is now
2016 based a combination of the sender and recipient addresses. This change
2017 can be overridden by setting address_retry_include_sender=false in the
2020 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2021 remote server are returned as part of bounce messages. This was not
2022 happening for LMTP over a pipe (the lmtp transport), but now it is the
2023 same for both kinds of LMTP.
2025 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2026 in header lines that were in fact CNAMEs. This is no longer the case.
2028 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2029 and queue runs started by the daemon processed all messages. This has
2030 been fixed so that -R and -S can now usefully be given with -q<time>.
2032 PH/40 Import PCRE release 6.7 (fixes some bugs).
2034 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2036 PH/42 Give an error if -q is specified more than once.
2038 PH/43 Renamed the variables $interface_address and $interface_port as
2039 $received_ip_address and $received_port, to make it clear that these
2040 values apply to message reception, and not to the outgoing interface when
2041 a message is delivered. (The old names remain recognized, of course.)
2043 PH/44 There was no timeout on the connect() call when using a Unix domain
2044 socket in the ${readsocket expansion. There now is.
2046 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2047 be meaningful with "accept".
2049 SC/01 Eximstats V1.43
2050 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2052 SC/02 Eximstats V1.44
2053 Use a glob alias rather than an array ref in the generated
2054 parser. This improves both readability and performance.
2056 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2057 Collect SpamAssassin and rejection statistics.
2058 Don't display local sender or destination tables unless
2059 there is data to show.
2060 Added average volumes into the top table text output.
2062 SC/04 Eximstats V1.46
2063 Collect data on the number of addresses (recipients)
2064 as well as the number of messages.
2066 SC/05 Eximstats V1.47
2067 Added 'Message too big' to the list of mail rejection
2068 reasons (thanks to Marco Gaiarin).
2070 SC/06 Eximstats V1.48
2071 Mainlog lines which have GMT offsets and are too short to
2072 have a flag are now skipped.
2074 SC/07 Eximstats V1.49 (Alain Williams)
2075 Added the -emptyok flag.
2077 SC/08 Eximstats V1.50
2078 Fixes for obtaining the IP address from reject messages.
2080 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2081 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2082 whitesspace changes from 4.64-PH/27
2084 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2087 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2088 are found, allow negative numbers in numeric criteria)
2090 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2092 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2093 to match changes made in 4.64-PH/43
2095 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2097 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2098 "domain" is actually an IP address, and force a failure. This locks out
2099 those revolvers/nameservers that support "A-for-A" lookups, in
2100 contravention of the specifications.
2102 PH/48 When a host name was looked up from an IP address, and the subsequent
2103 forward lookup of the name timed out, the host name was left in
2104 $sender_host_name, contrary to the specification.
2106 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2107 restricted to single-key lookups, Exim was not diagnosing an error if
2108 * or *@ was used with a query-style lookup.
2110 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2112 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2113 long ago, but noone interested enough thought of it. Let's just say that
2114 the "1.1" means that there are some new functions that weren't there at
2115 some point in the past.
2117 PH/51 Error processing for expansion failure of helo_data from an smtp
2118 transport during callout processing was broken.
2120 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2121 tested/used via the -bh/-bhc/-bs options.
2123 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2124 bug, fixed in subsequent PCRE releases).
2126 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2127 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2129 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2135 SC/01 Use a glob alias rather than an array ref in eximstats generated
2136 parser. This improves both readability and performance.
2138 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2139 Don't display local sender or destination tables in eximstats unless
2140 there is data to show.
2141 Added average volumes into the eximstats top table text output.
2143 SC/03 Collect data on the number of addresses (recipients) as well
2144 as the number of messages in eximstats.
2146 TF/01 Correct an error in the documentation for the redirect router. Exim
2147 does (usually) call initgroups() when daemonizing.
2149 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2150 with consistent privilege compared to when running as a daemon.
2152 TF/03 Note in the spec that $authenticated_id is not set for local
2153 submissions from trusted users.
2155 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2156 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2158 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2159 by adding some example configuration directives to the default
2160 configuration file. A little bit of work is required to uncomment the
2161 directives and define how usernames and passwords are checked, but
2162 there is now a framework to start from.
2164 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2165 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2166 without this. I don't know how relevant this is to other LDAP libraries.
2168 PH/02 Add the verb name to the "unknown ACL verb" error.
2170 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2172 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2174 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2175 directory not expanded when it should be if an expanded home directory
2176 was set for the address (which is overridden by the transport).
2178 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2181 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2182 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2183 because it is too late at that time, and has no effect.
2185 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2186 security issue with \' (bugzilla #107). I could not use the
2187 PQescapeStringConn() function, because it needs a PGconn value as one of
2190 PH/08 When testing addresses using -bt, indicate those final addresses that
2191 are duplicates that would not cause an additional delivery. At least one
2192 person was confused, thinking that -bt output corresponded to deliveries.
2193 (Suppressing duplicates isn't a good idea as you lose the information
2194 about possibly different redirections that led to the duplicates.)
2196 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2197 systems where poll() doesn't work, in particular OS X.
2199 PH/10 Added more information to debugging output for retry time not reached.
2201 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2202 operations in malware.c.
2204 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2207 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2208 syslog_duplication set false, log lines that would normally be written
2209 both the the main log and to the reject log were not written to syslog at
2212 PH/14 In the default configuration, change the use of "message" in ACL warn
2213 statements to "add_header".
2215 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2216 not followed by a command (e.g. "seen endif").
2218 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2219 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2222 PH/17 Added extra conditions to the default value of delay_warning_condition
2226 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2227 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2228 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2231 The Auto-Submitted: and various List- headers are standardised, whereas I
2232 don't think Precedence: ever was.
2234 PH/18 Refactored debugging code in route_finduser() to show more information,
2235 in particular, the error code if getpwnam() issues one.
2237 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2238 This is apparently needed in addition to the PH/07 change above to avoid
2239 any possible encoding problems.
2241 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2242 but not after initializing Perl.
2244 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2245 output them only if debugging. By default they are written stderr,
2246 apparently, which is not desirable.
2248 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2251 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2254 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2256 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2257 authenticated or an ident call has been made. Suppress the default
2258 values for $authenticated_id and $authenticated_sender (but permit -oMai
2259 and -oMas) when testing with -bh.
2261 PH/24 Re-jigged the order of the tests in the default configuration so that the
2262 tests for valid domains and recipients precede the DNS black list and CSA
2263 tests, on the grounds that those ones are more expensive.
2265 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2266 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2267 This bug exists in every version of Exim that I still have, right back to
2270 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2271 However, an attempt to turn on case-sensitivity in a regex key by
2272 including (?-i) didn't work because the subject string was already
2273 lowercased, and the effects were non-intuitive. It turns out that a
2274 one-line patch can be used to allow (?-i) to work as expected.
2280 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2281 other effects) broke the use of negated acl sub-conditions.
2283 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2286 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2287 "Deny" causes Exim to reject the incoming connection with a 554 error.
2288 Unfortunately, if there is a major crisis, such as a disk failure,
2289 tcp-wrappers gives "deny", whereas what one would like would be some
2290 kind of temporary error. A kludge has been added to help with this.
2291 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2292 554 error is used if errno is still zero or contains ENOENT (which occurs
2293 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2296 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2298 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2299 errors. Otherwise a message that provokes a temporary error (when other
2300 messages do not) can cause a whole host to time out.
2302 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2303 addresses were routed directly to files or pipes from a redirect router.
2304 File deliveries just didn't batch; pipe deliveries might have suffered
2307 PH/06 A failure to get a lock for a hints database would erroneously always say
2308 "Failed to get write lock", even when it was really a read lock.
2310 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2311 of 0600. This has been changed to use the value of the lockfile_mode
2312 option (which defaults to 0600).
2314 PH/08 Applied small patch from the Sieve maintainer.
2316 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2317 folder from quota calculations, a direct delivery into this folder messed
2318 up the contents of the maildirsize file. This was because the regex was
2319 used only to exclude .Trash (or whatever) when the size of the mailbox
2320 was calculated. There was no check that a delivery was happening into an
2321 excluded directory. This bug has been fixed by ignoring all quota
2322 processing for deliveries into excluded directories.
2324 PH/10 Added the maildirfolder_create_regex option to appendfile.
2330 PH/01 The code for finding all the local interface addresses on a FreeBSD
2331 system running IPv6 was broken. This may well have applied to all BSD
2332 systems, as well as to others that have similar system calls. The broken
2333 code found IPv4 interfaces correctly, but gave incorrect values for the
2334 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2335 that it would not match correctly against @[] and not recognize the IPv6
2338 PH/02 The ipliteral router was not recognizing addresses of the form user@
2339 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2341 PH/03 Added disable_ipv6.
2343 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2344 decoded form, because it is most often used to construct To: headers
2345 lines in autoreplies, and the decoded form may well be syntactically
2346 invalid. However, $reply_address has leading white space removed, and all
2347 newlines turned into spaces so that the autoreply transport does not
2350 PH/05 If group was specified without a user on a router, and no group or user
2351 was specified on a transport, the group from the router was ignored.
2353 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2354 for visible compile-time settings that can be used to change these
2355 numbers, for those that want even more. Backwards compatibility with old
2356 spool files has been maintained. However, going back to a previous Exim
2357 release will lost any variables that are in spool files.
2359 PH/07 Two small changes when running in the test harness: increase delay when
2360 passing a TCP/IP connection to a new process, in case the original
2361 process has to generate a bounce, and remove special handling of
2362 127.0.0.2 (sic), which is no longer necessary.
2364 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2365 be the same on different OS.
2367 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2370 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2371 whether --show-vars was specified or not
2373 JJ/02 exipick: Added support for new ACL variable spool format introduced
2376 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2377 syntactically invalid From: or Reply-to: line, and a filter used this to
2378 generate an autoreply, and therefore failed to obtain an address for the
2379 autoreply, Exim could try to deliver to a non-existent relative file
2380 name, causing unrelated and misleading errors. What now happens is that
2381 it logs this as a hard delivery error, but does not attempt to create a
2384 PH/11 The exinext utility has a -C option for testing purposes, but although
2385 the given file was scanned by exinext itself; it wasn't being passed on
2386 when Exim was called.
2388 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2389 an end-of-file indication when reading a command response.
2391 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2392 compiled. In many other places in Exim, IPv6 addresses are always
2393 recognized, so I have changed this. It also means that IPv4 domain
2394 literals of the form [IPV4:n.n.n.n] are now always recognized.
2396 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2397 used if the router is not running as root, for example, when verifying at
2398 ACL time, or when using -bh. The debugging output from this situation was
2399 non-existent - all you got was a failure to exec. I have made two
2402 (a) Failures to set uid/gid, the current directory, or a process leader
2403 in a subprocess such as that created by queryprogram now generate
2404 suitable debugging ouput when -d is set.
2406 (b) The queryprogram router detects when it is not running as root,
2407 outputs suitable debugging information if -d is set, and then runs
2408 the subprocess without attempting to change uid/gid.
2410 PH/15 Minor change to Makefile for building test_host (undocumented testing
2413 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2414 additional section of a DNS packet that returns MX or SRV records.
2415 Instead, it always explicitly searches for A/AAAA records. This avoids
2416 major problems that occur when a DNS server includes only records of one
2417 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2418 fixed another bug: if SRV records were looked up and the corresponding
2419 address records were *not* found in the additional section, the port
2420 values from the SRV records were lost.
2422 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2423 using the correct key (the original address) when searching the retry
2424 rules in order to find which one to use for generating the retry hint.
2426 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2427 adding the default one. Similarly, if it contains a Reply-To: header, the
2428 errors_reply_to option, if set, is not used.
2430 PH/19 When calculating a retry time, Exim used to measure the "time since
2431 failure" by looking at the "first failed" field in the retry record. Now
2432 it does not use this if it is later than than the arrival time of the
2433 message. Instead it uses the arrival time. This makes for better
2434 behaviour in cases where some deliveries succeed, thus re-setting the
2435 "first failed" field. An example is a quota failure for a huge message
2436 when small messages continue to be delivered. Without this change, the
2437 "time since failure" will always be short, possible causing more frequent
2438 delivery attempts for the huge message than are intended.
2439 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2441 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2442 $1, $2, $3) because the numerical variables can be reset during some
2443 expansion items (e.g. "match"), thereby losing the authentication data.
2445 PH/21 Make -bV show the size of off_t variables so that the test suite can
2446 decide whether to run tests for quotas > 2G.
2448 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2449 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2450 filecount value is greater than 2G or if a quota value is greater than 2G
2451 on a system where the size of off_t is not greater than 4, a panic error
2454 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2455 never match. The debug and -bh output now contains an explicit error
2456 message indicating a malformed IPv4 address or mask.
2458 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2459 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2460 PH/23 above applies.
2462 PH/25 Do not write to syslog when running in the test harness. The only
2463 occasion when this arises is a failure to open the main or panic logs
2464 (for which there is an explicit test).
2466 PH/26 Added the /no_tell option to "control=freeze".
2468 PH/27 If a host name lookup failed very early in a connection, for example, if
2469 the IP address matched host_lookup and the reverse lookup yielded a name
2470 that did not have a forward lookup, an error message of the form "no IP
2471 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2472 could be logged. Now it outputs the IP address instead of "NULL".
2474 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2475 allows the sender and the authenticated sender to be set when
2476 submitting a message from within Exim. Since child_open_exim() is
2477 documented for local_scan(), the new function should be too.
2479 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2480 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2481 results in an empty string is now treated as unset.
2483 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2485 PH/31 Added sender_verify_fail logging option.
2487 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2488 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2489 tidied the source and removed it altogether.
2491 PH/33 When a queue run was abandoned because the load average was too high, a
2492 log line was always written; now it is written only if the queue_run log
2493 selector is set. In addition, the log line for abandonment now contains
2494 information about the queue run such as the pid. This is always present
2495 in "start" and "stop" lines but was omitted from the "abandon" line.
2497 PH/34 Omit spaces between a header name and the colon in the error message that
2498 is given when verify = headers_syntax fails (if there are lots of them,
2499 the message gets confusing).
2501 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2502 names, as there are now some PTR records that contain slashes. This check
2503 is only to protect against broken name servers that fall over on strange
2504 characters, so the fact that it applies to all lookups doesn't matter.
2506 PH/36 Now that the new test suite is complete, we can remove some of the
2507 special code in Exim that was needed for the old test suite. For example,
2508 sorting DNS records because real resolvers return them in an arbitrary
2509 order. The new test suite's fake resolver always returns records in the
2512 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2513 bounces) except when queue_only is set, to avoid logging races between
2514 the different processes.
2516 PH/38 Panic-die if .include specifies a non-absolute path.
2518 PH/39 A tweak to the "H" retry rule from its user.
2520 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2521 a label. They prevented compilation on older perls.
2523 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2524 a warning to be raised on newish perls.
2526 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2527 on queue. Changes to match documented behaviour of showing count of
2528 messages matching specified criteria.
2530 PH/40 Changed the default ident timeout from 30s to 5s.
2532 PH/41 Added support for the use of login_cap features, on those BSD systems
2533 that have them, for controlling the resources used by pipe deliveries.
2535 PH/42 The content-scanning code uses fopen() to create files in which to put
2536 message data. Previously it was not paying any attention to the mode of
2537 the files. Exim runs with umask(0) because the rest of the code creates
2538 files with open(), and sets the required mode explicitly. Thus, these
2539 files were ending up world-writeable. This was not a big issue, because,
2540 being within the spool directory, they were not world-accessible. I have
2541 created a function called modefopen, which takes an additional mode
2542 argument. It sets umask(777), creates the file, chmods it to the required
2543 mode, then resets the umask. All the relevant calls to fopen() in the
2544 content scanning code have been changed to use this function.
2546 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2547 to 24 hours. This avoids potential overflow problems when processing G
2548 and H retry rules. I suspect nobody ever tinkers with this value.
2550 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2552 PH/45 When the plaintext authenticator is running as a client, the server's
2553 challenges are checked to ensure they are valid base64 strings. By
2554 default, the authentication attempt is cancelled if an invalid string is
2555 received. Setting client_ignore_invalid_base64 true ignores these errors.
2556 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2557 they are received. Thus, the responses can be made to depend on the
2558 challenges. If an invalid string is ignored, an empty string is placed in
2561 PH/46 Messages that are created by the autoreply transport now contains a
2562 References: header, in accordance with RFCs 2822 and 3834.
2564 PH/47 Added authenticated_sender_force to the smtp transport.
2566 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2568 PH/49 Installed latest patch from the Sieve maintainer.
2570 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2571 was also set, the appendfile transport was unnecessarily scanning a
2572 directory of message files (e.g. for maildir delivery) to find the count
2573 of files (along with the size), even though it did not need this
2574 information. It now does the scan only if it needs to find either the
2575 size of the count of files.
2577 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2579 PH/52 Two bugs concerned with error handling when the smtp transport is
2582 (i) Exim was not creating retry information for temporary errors given
2583 for individual recipients after the DATA command when the smtp transport
2584 was used in LMTP mode. This meant that they could be retried too
2585 frequently, and not timed out correctly.
2587 (ii) Exim was setting the flag that allows error details to be returned
2588 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2589 recipients that were returned after the DATA command.
2591 PH/53 This is related to PH/52, but is more general: for any failing address,
2592 when detailed error information was permitted to be returned to the
2593 sender, but the error was temporary, then after the final timeout, only
2594 "retry timeout exceeded" was returned. Now it returns the full error as
2595 well as "retry timeout exceeded".
2597 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2598 do this, and (what is worse) MTAs that accept it.
2600 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2601 will now be deprecated.
2603 PH/56 New os.c-cygwin from the Cygwin maintainer.
2605 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2606 formats (previously only available in exim formats via -bpr, -bpru,
2607 and -bpra. Now also available in native and exiqgrep formats)
2609 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2610 with very large, slow to parse queues
2612 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2614 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2616 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2617 responses to authentication challenges, though it was showing the
2618 challenges; (ii) I've removed the CR characters from the debug output for
2621 PH/58 Allow for the insertion of a newline as well as a space when a string
2622 is turned into more than one encoded-word during RFC 2047 encoding. The
2623 Sieve code now uses this.
2625 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2626 data_4xx, lost_connection, tls_required.
2628 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2629 message was being sent as an SMTP response.
2631 PH/61 Add -l and -k options to exicyclog.
2633 PH/62 When verifying, if an address was redirected to one new address, so that
2634 verification continued, and the new address failed or deferred after
2635 having set something in $address_data, the value of $address_data was not
2636 passed back to the ACL. This was different to the case when no
2637 redirection occurred. The value is now passed back in both cases.
2639 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2640 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2641 use login_cap.h, so on its own it isn't the distinguishing feature. The
2642 new name refers directly to the setclassresources() function.
2644 PH/65 Added configuration files for NetBSD3.
2646 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2648 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2649 is preferred over IPv4.
2651 PH/68 The bounce_return_message and bounce_return_body options were not being
2652 honoured for bounces generated during the reception of non-SMTP messages.
2653 In particular, this applied to messages rejected by the ACL. This bug has
2654 been fixed. However, if bounce_return_message is true and bounce_return_
2655 body is false, the headers that are returned for a non-SMTP message
2656 include only those that have been read before the error was detected.
2657 (In the case of an ACL rejection, they have all been read.)
2659 PH/69 The HTML version of the specification is now built in a directory called
2660 spec_html instead of spec.html, because the latter looks like a path with
2661 a MIME-type, and this confuses some software.
2663 PH/70 Catch two compiler warnings in sieve.c.
2665 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2666 function verify_get_ident() calls ip_connect() to connect a socket, but
2667 if the "connect()" function timed out, ip_connect() used to close the
2668 socket. However, verify_get_ident() also closes the socket later, and in
2669 between Exim writes to the log, which may get opened at this point. When
2670 the socket was closed in ip_connect(), the log could get the same file
2671 descriptor number as the socket. This naturally causes chaos. The fix is
2672 not to close the socket in ip_connect(); the socket should be closed by
2673 the function that creates it. There was only one place in the code where
2674 this was missing, in the iplookup router, which I don't think anybody now
2675 uses, but I've fixed it anyway.
2677 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2678 well as to direct DNS lookups. Otherwise the handling of names in host
2679 lists is inconsistent and therefore confusing.
2685 PH/01 Two changes to the default runtime configuration:
2687 (1) Move the checks for relay_from_hosts and authenticated clients from
2688 after to before the (commented out) DNS black list checks.
2690 (2) Add control=submission to the relay_from_hosts and authenticated
2691 clients checks, on the grounds that messages accepted by these
2692 statements are most likely to be submissions.
2694 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2696 (1) Generate an error if the third argument for the ${prvs expansion is
2699 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2702 (3) Reset the variables that are obtained from the first argument of
2703 ${prvscheck and used in the second argument before leaving the code,
2704 because their memory is reclaimed, so using them afterwards may do
2707 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2708 one (it's much easier than Tom thought :-).
2710 (5) Because of (4), we can now allow for the use of $prvscheck_result
2711 inside the third argument.
2713 PH/03 For some reason, the default setting of PATH when running a command from
2714 a pipe transport was just "/usr/bin". I have changed it to
2717 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2718 anything to be listed in the output from -bV.
2720 PH/05 When a filter generated an autoreply, the entire To: header line was
2721 quoted in the delivery log line, like this:
2723 => >A.N.Other <ano@some.domain> <original@ddress> ...
2725 This has been changed so that it extracts the operative address. There
2726 may be more than one such address. If so, they are comma-separated, like
2729 => >ano@some.domain,ona@other.domain <original@ddress> ...
2731 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2732 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2733 not being looked up in the rDNS to get a host name, Exim was showing the
2734 IP address twice in Received: lines, even though the IP addresses were
2735 identical. For example:
2737 Received: from [1.2.3.4] (helo=[1.2.3.4])
2739 However, if the real host name was known, it was omitting the HELO data
2740 if it matched the actual IP address. This has been tidied up so that it
2741 doesn't show the same IP address twice.
2743 PH/07 When both +timestamp and +memory debugging was on, the value given by
2744 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2745 called by the debug printing, thereby overwriting the timestamp buffer.
2746 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2748 PH/08 When the original message was included in an autoreply transport, it
2749 always said "this is a copy of the message, including all the headers",
2750 even if body_only or headers_only was set. It now gives an appropriate
2753 PH/09 Applied a patch from the Sieve maintainer which:
2755 o fixes some comments
2756 o adds the (disabled) notify extension core
2757 o adds some debug output for the result of if/elsif tests
2758 o points to the current vacation draft in the documentation
2759 and documents the missing references header update
2763 o fixes a bug in processing the envelope test (when testing
2764 multiple envelope elements, the last element determinted the
2767 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2768 Electronic Mail") by including:
2770 Auto-submitted: auto-generated
2772 in the messages that it generates (bounce messages and others, such as
2773 warnings). In the case of bounce messages for non-SMTP mesages, there was
2774 also a typo: it was using "Auto_submitted" (underscore instead of
2775 hyphen). Since every message generated by Exim is necessarily in response
2776 to another message, thes have all been changed to:
2778 Auto-Submitted: auto-replied
2780 in accordance with these statements in the RFC:
2782 The auto-replied keyword:
2784 - SHOULD be used on messages sent in direct response to another
2785 message by an automatic process,
2787 - MUST NOT be used on manually-generated messages,
2789 - MAY be used on Delivery Status Notifications (DSNs) and Message
2790 Disposition Notifications (MDNs),
2792 - MUST NOT be used on messages generated by automatic or periodic
2793 processes, except for messages which are automatic responses to
2796 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2797 to the default Received: header definition.
2799 PH/12 Added log selector acl_warn_skipped (default on).
2801 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2802 variables because (a) they are in the wrong storage pool and (b) even if
2803 they were copied, it wouldn't work properly because of the caching.
2805 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2806 checking when decoding. Apparently there are clients that generate
2807 overlong encoded strings. Why am I not surprised?
2809 PH/15 If the first argument of "${if match_address" was not empty, but did not
2810 contain an "@" character, Exim crashed. Now it writes a panic log message
2811 and treats the condition as false.
2813 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2815 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2816 extension "envelope-auth". The code is finished and in agreement with
2817 other implementations, but there is no documentation so far and in fact,
2818 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2819 not changing the active code.
2821 Print executed "if" and "elsif" statements when debugging is used. This
2822 helps a great deal to understand what a filter does.
2824 Document more things not specified clearly in RFC3028. I had all this
2825 sorted out, when out of a sudden new issues came to my mind. Oops."
2827 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2830 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2831 canonical form (as documented). However, after a host name lookup from
2832 the IP address, check_host() was doing a simple string comparison with
2833 addresses acquired from the DNS when checking that the found name did
2834 have the original IP as one of its addresses. Since any found IPv6
2835 addresses are likely to be in abbreviated form, the comparison could
2836 fail. Luckily, there already exists a function for doing the comparison
2837 by converting both addresses to binary, so now that is used instead of
2838 the text comparison.
2840 PH/20 There was another similar case to PH/19, when a complete host name was
2841 given in a host list; looking up its IP address could give an abbreviated
2842 form, whereas the current host's name might or might not be abbreviated.
2843 The same fix has been applied.
2849 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2850 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2853 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2854 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2856 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2858 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2859 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2860 Also, for "Auto-Submitted", treat anything other than "no" as
2861 non-personal, in accordance with RFC 3834. (Previously it treated
2862 anything starting "auto-" as non-personal.)
2864 TF/01 The control=submission/name=... option had a problem with syntax
2865 errors if the name included a slash character. The /name= option
2866 now slurps the rest of the string, so it can include any characters
2867 but it must come last in the list of options (after /sender_retain
2870 PH/05 Some modifications to the interface to the fake nameserver for the new
2878 TK/01 Added the "success_on_redirect" address verification option. See
2879 NewStuff for rationale and an example.
2881 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2883 PH/02 Patch to exigrep to allow it to work on syslog lines.
2885 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2886 fread() to skip over the body file's header line, because in Cygwin the
2887 header line is locked and is inaccessible.
2889 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2890 co-exist for some time) to make it clear that it is the Exim ID that is
2891 referenced, not the Message-ID: header line.
2893 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2894 string_format() function, because snprintf() does not exist on all
2897 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2898 ${stat: expansion item.
2900 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2901 protocol synchronization error", to keep the pedants happy.
2903 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2904 well as for IRIX systems, when gcc is being used. See the host.c source
2907 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2909 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2912 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2913 transport and to the smtp transport in LMTP mode.
2915 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2917 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2918 run a filter in a subprocess. This could lead to confusion in subsequent
2919 lookups in the parent process. There should also be a search_tidyup() at
2920 the end of the subprocess.
2922 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2923 only if the host matched helo_try_verify_hosts, which caused the
2924 verification to occur when the EHLO/HELO command was issued. The ACL just
2925 tested the remembered result. Now, if a previous verification attempt has
2926 not happened, "verify = helo" does it there and then.
2928 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2930 TK/03 Fix log output including CR from clamd.
2932 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2933 exist provoked a memory error which could cause a segfault.
2935 PH/15 Installed PCRE 6.2
2937 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2939 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2940 of the problem. Specifically, suggested +O2 rather than +O1 for the
2943 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2945 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2946 record, the port was not correctly logged when the outgoing_port log
2947 selector was set (it logged the transort's default port).
2949 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2950 fallback_hosts, and "hosts" in the smtp transport.
2952 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2953 host errors such as "Connection refused".
2955 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2956 authentication with radiusclient 0.4.9:
2958 - Error returned from rc_read_config was caught wrongly
2959 - Username/password not passed on to radius server due to wrong length.
2961 The presumption is that some radiusclient API changes for 4.51/PH/17
2962 were not taken care of correctly. The code is still untested by me (my
2963 Linux distribution still has 0.3.2 of radiusclient), but it was
2964 contributed by a Radius user.
2966 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2967 expanding the "port" option of the smtp transport.
2969 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2970 while reading a MIME header. Thanks to Tom Hughes for a patch.
2972 PH/24 Include config.h inside local_scan.h so that configuration settings are
2975 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2976 that in an ACL for RCPT (for example), you can examine exactly what was
2979 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2980 commands, but it was not correctly comparing the address with the actual
2981 client host address. Thus, it would show the EHLO address in Received:
2982 header lines when this was not necessary.
2984 PH/27 Added the % operator to ${eval:}.
2986 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2987 it should be ignoring failures (because with -C, for example, it has lost
2988 privilege). It wasn't ignoring creation failures other than "already
2991 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2994 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2995 because an input line was too long, either on its own, or by virtue of
2996 too many continuations, the temporary file was not being removed, and the
2997 return code was incorrect.
2999 PH/31 Missing "BOOL" in function definition in filtertest.c.
3001 PH/32 Applied Sieve patches from the maintainer.
3003 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3005 PH/33 Added "verify = not_blind".
3007 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3008 Local/Makefile (with some defaults set). These are used in built scripts
3009 such as exicyclog, but they have never been used in the exim_install
3010 script (though there are many overriding facilities there). I have
3011 arranged that the exim_install script now takes note of these two
3014 PH/35 Installed configuration files for Dragonfly.
3016 PH/36 When a locally submitted message by a trusted user did not contain a
3017 From: header, and the sender address was obtained from -f or from an SMTP
3018 MAIL command, and the trusted user did not use -F to supply a sender
3019 name, $originator_name was incorrectly used when constructing a From:
3020 header. Furthermore, $originator_name was used for submission mode
3021 messages from external hosts without From: headers in a similar way,
3022 which is clearly wrong.
3024 PH/37 Added control=suppress_local_fixups.
3026 PH/38 When log_selector = +received_sender was set, and the addition of the
3027 sender made the log line's construction buffer exactly full, or one byte
3028 less than full, an overflow happened when the terminating "\n" was
3031 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3032 when the result of a list match is failure because a DNS lookup failed.
3034 PH/40 RM_COMMAND is now used in the building process.
3036 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3037 the "build-* directories that it finds.
3039 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3040 address in a domain literal was a prefix of an interface address.
3042 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3043 when verifying a sender address, unless rewrite_headers is false.
3045 PH/44 Wrote a long comment about why errors_to addresses are verified as
3046 recipients, not senders.
3048 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3049 the ratelimit ACL was added.
3051 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3053 PH/46 Added extra information about PostgreSQL errors to the error string.
3055 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3056 avoiding the need to install special zones in a real server. This is
3057 backwards compatible; if it can't find the fake resolver, it drops back.
3058 Thus, both old and new test suites can be run.
3060 TF/02 Added util/ratelimit.pl
3062 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3063 clock is set back in time.
3065 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3066 Candler <B.Candler@pobox.com>.
3068 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3069 OK for addresses that are the result of redirections.
3071 PH/48 A number of further additions for the benefit of the new test suite,
3072 including a fake gethostbyname() that interfaces to the fake DNS resolver
3075 TF/06 The fix for widen_domains has also been applied to qualify_single and
3076 search_parents which are the other dnslookup options that can cause
3079 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3082 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3084 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3085 service" if it encountered an MX record with an empty target hostname.
3086 The message is now "an MX or SRV record indicated no SMTP service".
3088 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3089 if the DNS of the sending site is misconfigured. This is quite a
3090 common situation. This change restores the behaviour of treating a
3091 helo verification defer as a failure.
3093 PH/51 If self=fail was set on a router, the bounce message did not include the
3094 actual error message.
3100 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3102 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3103 command itself ended in error, the underlying message about the transport
3104 filter timeout was being overwritten with the pipe command error. Now the
3105 underlying error message should be appended to the second error message.
3107 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3109 PH/02 Reduce the amount of output that "make" produces by default. Full output
3110 can still be requested.
3112 PH/03 The warning log line about a condition test deferring for a "warn" verb
3113 was being output only once per connection, rather than after each
3114 occurrence (because it was using the same function as for successful
3115 "warn" verbs). This seems wrong, so I have changed it.
3117 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3118 it should not have, which might have caused a crash in the right
3119 circumstances, but probably never did.
3121 PH/04 Installed a modified version of Tony Finch's patch to make submission
3122 mode fix the return path as well as the Sender: header line, and to
3123 add a /name= option so that you can make the user's friendly name appear
3126 TF/03 Added the control = fakedefer ACL modifier.
3128 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3129 Mark Lowes for thorough testing.
3131 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3133 TK/03 Merged latest SRS patch from Miles Wilton.
3135 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3136 with the definition in sysexits.h (which is #included earlier).
3137 Fortunately, Exim does not actually use EX_OK. The code used to try to
3138 preserve the sysexits.h value, by assumimg that macro definitions were
3139 scanned for macro replacements. I have been disabused of this notion,
3140 so now the code just undefines EX_OK before #including unistd.h.
3142 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3143 in the smtp transport. When a block could not be written in a single
3144 write() function, the timeout was being re-applied to each part-write.
3145 This seems wrong - if the receiver was accepting one byte at a time it
3146 would take for ever. The timeout is now adjusted when this happens. It
3147 doesn't have to be particularly precise.
3149 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3150 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3152 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3153 without a domain if the check to <postmaster@domain> fails.
3155 SC/01 Eximstats: added -xls and the ability to specify output files
3156 (patch written by Frank Heydlauf).
3158 SC/02 Eximstats: use FileHandles for outputing results.
3160 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3162 SC/04 Eximstats: fixed display of large numbers with -nvr option
3164 SC/05 Eximstats: fixed merging of reports with empty tables.
3166 SC/06 Eximstats: added the -include_original_destination flag
3168 SC/07 Eximstats: removed tabs and trailing whitespace.
3170 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3172 TK/06 MBOX spool code: Add real "From " MBOX separator line
3173 so the .eml file is really in mbox format (even though
3174 most programs do not really care). Patch from Alex Miller.
3176 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3177 The latter is generated from $received_to and is only set if the
3178 message has one envelope recipient. SA can use these headers,
3179 obviously out-of-the-box. Patch from Alex Miller.
3181 PH/08 The ${def test on a variable was returning false if the variable's
3182 value was "0", contrary to what the specification has always said!
3183 The result should be true unless the variable is empty.
3185 PH/09 The syntax error of a character other than { following "${if
3186 def:variable_name" (after optional whitespace) was not being diagnosed.
3187 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3188 accidental colon was present, for example, could give incorrect results.
3190 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3191 result is used (not including appendfile, where other changes are about
3194 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3195 This involved changing a lot of size variables from int to off_t. It
3196 should work with maildirs and everything.
3198 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3199 spamd dying while we are connected to it.
3201 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3204 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3205 function for BDB 4.3.
3207 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3209 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3210 the sources for PCRE have been re-arranged and more files are now
3213 PH/15 The code I had for printing potentially long long variables in PH/11
3214 above was not the best (it lost precision). The length of off_t variables
3215 is now inspected at build time, and an appropriate printing format (%ld
3216 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3217 to be "long long int" or "long int". This is needed for the internal
3218 formatting function string_vformat().
3220 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3221 the configuration file to be ":syslog", then the script "guesses" where
3222 the logs files are, rather than using the compiled in default. In our
3223 case the guess is not the same as the compiled default, so the script
3224 suddenly stopped working when I started to use syslog. The patch checks
3225 to see if log_file_path is "". If so, it attempts to read it from exim
3226 with no configuration file to get the compiled in version, before it
3227 falls back to the previous guessing code."
3229 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3230 implementing BATV in an Exim configuration. See NewStuff for the gory
3233 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3234 Makefile that are specific to HP-UX.
3236 PH/18 If the "use_postmaster" option was set for a recipient callout together
3237 with the "random" option, the postmaster address was used as the MAIL
3238 FROM address for the random test, but not for the subsequent recipient
3239 test. It is now used for both.
3241 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3242 patch removes a few documentation additions to RFC 3028, because the
3243 latest draft now contains them. It adds the new en;ascii-case comparator
3244 and a new error check for 8bit text in MIME parts. Comparator and
3245 require names are now matched exactly. I enabled the subaddress
3246 extension, but it is not well tested yet (read: it works for me)."
3248 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3249 rework some of the code of TK/09 above to avoid the hardwired use of
3250 "%lld" and "long long". Replaced the call to snprintf() with a call to
3253 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3254 records point to non-existent hosts", "retry timeout exceeded", and
3255 "retry time not reached for any host after a long failure period".
3257 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3258 experimental DomainKeys support:
3260 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3261 (2) On an error such as an illegally used "control", the wrong name for
3262 the control was given.
3264 These problems did NOT occur unless DomainKeys support was compiled.
3266 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3268 PH/24 Added ${if match_ip condition.
3270 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3271 databases so that it will be absolutely obvious if a crash occurs in the
3272 DB library. This is a regular occurrence (often caused by mis-matched
3275 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3276 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3277 picked up on a user's system that detects such things. There doesn't seem
3278 to be a gcc warning option for this - only an attribute that has to be
3279 put on the function's prototype. It seems that in Fedora Core 4 they have
3280 set this on a number of new functions. No doubt there will be more in due
3283 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3284 specify a transport. However, if an address that was verified by such a
3285 router was the subject of a callout, Exim crashed because it tried to
3286 read the rcpt_include_affixes from the non-existent transport. Now it
3287 just assumes that the setting of that option is false. This bug was
3288 introduced by 4.51/PH/31.
3290 PH/28 Changed -d+all to exclude +memory, because that information is very
3291 rarely of interest, but it makes the output a lot bigger. People tend to
3292 do -d+all out of habit.
3294 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3295 code in os-type was giving problems when libc.so lives in lib64, like on
3298 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3299 aren't the modern standard, and it seems that some systems' include files
3300 don't always have them. Exim was already checking for some of the newer
3301 ones like T_AAAA, and defining it itself. I've added checks for all the
3302 record types that Exim uses.
3304 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3305 not automatically generating a new one, as it is supposed to. This
3306 prevented TLS from working. If the file did exist, but contained invalid
3307 data, a new version was generated, as expected. It was only the case of a
3308 non-existent file that was broken.
3310 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3311 with a change in libdomainkeys > 0.64.
3313 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3314 from DNS. If the selector record carries the flag, it now has
3315 precedence over the domain-wide flag.
3317 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3319 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3320 the use of a port name that isn't defined in /etc/services) occurred, the
3321 message was deferred as in a normal delivery, and thus remained on the
3322 spool, instead of being failed because of the mua_wrapper setting. This
3323 is now fixed, and I tidied up some of the mua_wrapper messages at the
3326 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3327 the messages in a hash of arrays rather than using individual hashes.
3328 This is a bit cleaner and results in dramatic memory savings, albeit
3329 at a slight CPU cost.
3331 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3332 as requested by Marc Sherman.
3334 SC/10 Eximstats: added histograms for user specified patterns as requested
3337 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3339 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3340 fopen() in the content-scanning modules that did not already have it.
3346 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3347 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3349 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3351 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3353 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3354 to test Sieve filters that use "vacation".
3356 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3357 that changes the way the GnuTLS parameters are stored in the cache file.
3358 The new format can be generated externally. For backward compatibility,
3359 if the data in the cache doesn't make sense, Exim assumes it has read an
3360 old-format file, and it generates new data and writes a new file. This
3361 means that you can't go back to an older release without removing the
3364 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3365 work if there are any delivery delays because "one_time" forces the
3366 parent to be marked "delivered", so its unseen clone is never tried
3367 again. For this reason, Exim now forbids the simultaneous setting of
3370 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3371 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3372 entry if you want to know the details. The fix, however, neglected to
3373 consider the case when local delivery batching is involved. The test for
3374 "previously delivered" was not happening when checking to see if an
3375 address could be batched with a previous (undelivered) one; under
3376 certain circumstances this could lead to multiple deliveries to the same
3379 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3380 in its include files, and this causes problems building Exim.
3382 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3383 header_syntax) but Exim was just ignoring anything given after a slash.
3384 In particular, this caused confusion with an attempt to use "verify =
3385 reverse_host_lookup/defer_ok". An error is now given when options are
3386 supplied for verify items that do not have them. (Maybe reverse_host_
3387 lookup should have a defer_ok option, but that's a different point.)
3389 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3390 defined by RFC 821) to 2048, because there were problems with some AUTH
3391 commands, and RFC 1869 says the size should be increased for extended
3392 SMTP commands that take arguments.
3394 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3397 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3398 "unknown" error; now it says that the functionality isn't in the binary.
3400 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3401 an address' error message when a string expansion fails (syntax or
3402 whatever). Otherwise the password may appear in the log. Following change
3403 PH/42 below, there is no longer a chance of it appearing in a bounce
3406 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3408 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3409 of its addresses was ever tried. (Bugzilla bug #2.)
3411 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3412 the result incorrectly in the debug output. (It correctly added a newline
3413 to what was transported.)
3415 TF/01 Added $received_time.
3417 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3418 commented out examples of how to interface to a virus scanner and to
3419 SpamAssassin. Also added commented examples of av_scanner and
3420 spamd_address settings.
3422 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3423 and controls are allowed in which ACLs. There were a couple of minor
3424 errors. Some of the entries in the conditions table (which is a table of
3425 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3426 negation of where the condition IS allowed.
3428 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3430 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3431 header file does not have a version number, so I've had to invent a new
3432 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3433 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3434 radiusclient), but it was contributed by a Radius user.
3436 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3437 files or directories, for OpenSSL.
3439 PH/19 When an Exim process that is running as root has to create an Exim log
3440 file, it does so in a subprocess that runs as exim:exim so as to get the
3441 ownership right at creation (otherwise, other Exim processes might see
3442 the file with the wrong ownership). There was no test for failure of this
3443 fork() call, which would lead to the process getting stuck as it waited
3444 for a non-existent subprocess. Forks do occasionally fail when resources
3445 run out. I reviewed all the other calls to fork(); they all seem to check
3448 PH/20 When checking for unexpected SMTP input at connect time (before writing
3449 the banner), Exim was not dealing correctly with a non-positive return
3450 from the read() function. If the client had disconnected by this time,
3451 the result was a log entry for a synchronization error with an empty
3452 string after "input=" when read() returned zero. If read() returned -1
3453 (an event I could not check), uninitialized data bytes were printed.
3454 There were reports of junk text (parts of files, etc) appearing after
3457 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3459 PH/22 Added support for macro redefinition, and (re)definition in between
3460 driver and ACL definitions.
3462 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3463 forgetting to use the resulting value; it was using the unexpanded value.
3465 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3466 hadn't been configured. The fix is from Juergen Kreileder, who
3467 understands it better than I do:
3469 "Here's what I see happening with three configured cyrus_sasl
3470 authenticators configured (plain, login, cram-md5):
3472 On startup auth_cyrus_sasl_init() gets called for each of these.
3473 This means three calls to sasl_listmech() without a specified mech_list.
3474 => SASL tests which mechs of all available mechs actually work
3475 => three warnings about OTP not working
3476 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3478 With the patch, sasl_listmech() also gets called three times. But now
3479 SASL's mech_list option is set to the server_mech specified in the the
3480 authenticator. Or in other words, the answer from sasl_listmech()
3481 gets limited to just the mech you're testing for (which is different
3483 => the return list contains just 'plain' or 'login', 'cram-md5' or
3484 nothing depending on the value of ob->server_mech.
3486 I've just tested the patch: Authentication still works fine,
3487 unavailable mechs specified in the exim configuration are still
3488 caught, and the auth.log warnings about OTP are gone."
3490 PH/25 When debugging is enabled, the contents of the command line are added
3491 to the debugging output, even when log_selector=+arguments is not
3494 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3495 answer is "GNU", and only if the return is "GNU/something" is the answer
3498 PH/27 $acl_verify_message is now set immediately after the failure of a
3499 verification in an ACL, and so is available in subsequent modifiers. In
3500 particular, the message can be preserved by coding like this:
3502 warn !verify = sender
3503 set acl_m0 = $acl_verify_message
3505 Previously, $acl_verify_message was set only while expanding "message"
3506 and "log_message" when a very denied access.
3508 PH/28 Modified OS/os.c-Linux with
3510 -#ifndef OS_LOAD_AVERAGE
3511 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3513 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3514 nomenclature these days.)
3516 PH/29 Installed patch from the Sieve maintainer that adds the options
3517 sieve_useraddress and sieve_subaddress to the redirect router.
3519 PH/30 In these circumstances:
3520 . Two addresses routed to the same list of hosts;
3521 . First host does not offer TLS;
3522 . First host accepts first address;
3523 . First host gives temporary error to second address;
3524 . Second host offers TLS and a TLS session is established;
3525 . Second host accepts second address.
3526 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3527 and peerdn, if requested) that were in fact used only for the second
3530 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3531 attention to any local part prefix or suffix that was matched by the
3532 router that accepted the address. It now behaves in the same way as it
3533 does for delivery: the affixes are removed from the local part unless
3534 rcpt_include_affixes is set on the transport.
3536 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3537 timeout during the DATA phase of an incoming message.
3539 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3540 applied a patch sanctioned by the Sieve maintainer.
3542 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3543 the uid or gid is negative. A case of a negative gid caused this to be
3544 noticed. The fix allows for either to be negative.
3546 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3547 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3550 PH/36 The change PH/12 above was broken. Fixed it.
3552 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3553 the grounds that routing the same address twice would always produce the
3554 same answer. This might have been true once, but it is certainly no
3555 longer true now. Routing a child address may depend on the previous
3556 routing that produced that child. Some complicated redirection strategies
3557 went wrong when messages had multiple recipients, and made Exim's
3558 behaviour dependent on the order in which the addresses were given.
3560 I have moved the duplicate checking until after the routing is complete.
3561 Exim scans the addresses that are assigned to local and remote
3562 transports, and removes any duplicates. This means that more work will be
3563 done, as duplicates will always all be routed, but duplicates are
3564 presumably rare, so I don't expect this is of any significance.
3566 For deliveries to pipes, files, and autoreplies, the duplicate checking
3567 still happens during the routing process, since they are not going to be
3570 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3571 It corrects a timeout issue with spamd. This is Ian's comment: "The
3572 background is that sometimes spamd either never reads data from a
3573 connection it has accepted, or it never writes response data. The exiscan
3574 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3575 blindly assumes that writes won't block so it may never time out."
3577 PH/39 Allow G after quota size as well as K and M.
3579 PH/40 The value set for $authenticated_id in an authenticator may not contain
3580 binary zeroes or newlines because the value is written to log lines and
3581 to spool files. There was no check on this. Now the value is run through
3582 the string_printing() function so that such characters are converted to
3583 printable escape sequences.
3585 PH/41 $message_linecount is a new variable that contains the total number of
3586 lines in the message. Compare $body_linecount, which is the count for the
3589 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3590 bounce and delay warning messages, except in certain special cases, which
3593 (a) An SMTP error message from a remote host;
3594 (b) A message specified in a :fail: redirection;
3595 (c) A message specified in a "fail" command in a system filter;
3596 (d) A message specified in a FAIL return from the queryprogram router;
3597 (e) A message specified by the cannot_route_message router option.
3599 In these cases only, Exim does include the error details in bounce and
3600 warning messages. There are also a few cases where bland messages such
3601 as "unrouteable address" or "local delivery error" are given.
3603 PH/43 $value is now also set for the "else" part of a ${run expansion.
3605 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3606 being worked on, but at least Exim now implements the latest version to
3609 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3610 process to complete was treated as a delivery failure, a timeout while
3611 writing the message to the pipe was logged, but erroneously treated as a
3612 successful delivery. Such timeouts include transport filter timeouts. For
3613 consistency with the overall process timeout, these timeouts are now
3614 treated as errors, giving rise to delivery failures by default. However,
3615 there is now a new Boolean option for the pipe transport called
3616 timeout_defer, which, if set TRUE, converts the failures into defers for
3617 both kinds of timeout. A transport filter timeout is now identified in
3620 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3621 systems where "make" and "gmake" are different, calling "gmake" at top
3622 level broke things. I've arranged for the value of $(MAKE) to be passed
3623 from the Makefile to this script so that it can call the same version of
3627 A note about Exim versions 4.44 and 4.50
3628 ----------------------------------------
3630 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3631 changes of various kinds. As a consequence, a big documentation update was
3632 needed. This delayed the release for rather longer than seemed good, especially
3633 in the light of a couple of (minor) security issues. Therefore, the changes
3634 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3635 release. So 4.44 and 4.50 are in effect two different branches that both start
3638 I have left the 4.50 change log unchanged; it contains all the changes since
3639 4.43. The change log for 4.44 is below; many of its items are identical to
3640 those for 4.50. This seems to be the most sensible way to preserve the
3641 historical information.
3647 1. Minor wording change to the doc/README.SIEVE file.
3649 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3650 computation of the current number of files was incorrect.
3652 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3653 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3656 4. Give more explanation in the error message when the command for a transport
3657 filter fails to execute.
3659 5. There are several places where Exim runs a non-Exim command in a
3660 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3661 was being done only for the command run by the queryprogram router. It is
3662 now done for all such subprocesses. The other cases are: ${run, transport
3663 filters, and the commands run by the lmtp and pipe transports.
3665 6. Added CONFIGURE_GROUP build-time option.
3667 7. Some older OS have a limit of 256 on the maximum number of file
3668 descriptors. Exim was using setrlimit() to set 1000 as a large value
3669 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3670 systems. I've change it so that if it can't get 1000, it tries for 256.
3672 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3673 was an oversight, and furthermore, ever since the addition of extra
3674 controls (e.g. 4.43/32), the checks on when to allow different forms of
3675 "control" were broken. There should now be diagnostics for all cases when a
3676 control that does not make sense is encountered.
3678 9. Added the /retain_sender option to "control=submission".
3680 10. $recipients is now available in the predata ACL (oversight).
3682 11. Tidy the search cache before the fork to do a delivery from a message
3683 received from the command line. Otherwise the child will trigger a lookup
3684 failure and thereby defer the delivery if it tries to use (for example) a
3685 cached ldap connection that the parent has called unbind on.
3687 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3688 of $address_data from the recipient verification was clobbered by the
3689 sender verification.
3691 13. The value of address_data from a sender verification is now available in
3692 $sender_address_data in subsequent conditions in the ACL statement.
3694 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3696 15. Added a new option "connect=<time>" to callout options, to set a different
3699 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3700 was its contents. (It was OK if the option was not defined at all.)
3702 17. A "Completed" log line is now written for messages that are removed from
3703 the spool by the -Mrm option.
3705 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3706 information about exactly what failed.
3708 19. Added -dd to debug only the daemon process.
3710 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3711 handles timeouts, both on the server side and network timeouts. Renamed the
3712 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3714 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3715 It is now set to "smtps".
3717 22. $host_address is now set to the target address during the checking of
3718 ignore_target_hosts.
3720 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3721 being passed; this would have caused $sender_host_name to have been used if
3722 matching the list had actually called for a host name (not very likely,
3723 since this list is usually IP addresses). A host name is now passed as
3726 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3727 code that specifies a non-restarting handler (typically sigaction() in
3728 modern systems) in an attempt to fix a rare and obscure crash bug.
3730 25. Narrowed the window for a race in the daemon that could cause it to ignore
3731 SIGCHLD signals. This is not a major problem, because they are used only to
3732 wake it up if nothing else does.
3734 26. A malformed maildirsize file could cause Exim to calculate negative values
3735 for the mailbox size or file count. Odd effects could occur as a result.
3736 The maildirsize information is now recalculated if the size or filecount
3739 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3740 support for a long time. Removed HAVE_SYS_VFS_H.
3742 28. Installed the latest version of exipick from John Jetmore.
3744 29. In an address list, if the pattern was not a regular expression, an empty
3745 subject address (from a bounce message) matched only if the pattern was an
3746 empty string. Non-empty patterns were not even tested. This was the wrong
3747 because it is perfectly reasonable to use an empty address as part of a
3748 database query. An empty address is now tested by patterns that are
3749 lookups. However, all the other forms of pattern expect the subject to
3750 contain a local part and a domain, and therefore, for them, an empty
3751 address still always fails if the pattern is not itself empty.
3753 30. Exim went into a mad DNS loop when attempting to do a callout where the
3754 host was specified on an smtp transport, and looking it up yielded more
3755 than one IP address.
3757 31. Re-factored the code for checking spool and log partition space into a
3758 function that finds that data and another that does the check. The former
3759 is then used to implement four new variables: $spool_space, $log_space,
3760 $spool_inodes, and $log_inodes.
3762 32. The RFC2047 encoding function was originally intended for short strings
3763 such as real names; it was not keeping to the 75-character limit for
3764 encoded words that the RFC imposes. It now respects the limit, and
3765 generates multiple encoded words if necessary. To be on the safe side, I
3766 have increased the buffer size for the ${rfc2047: expansion operator from
3769 33. It is now permitted to omit both strings after an "if" condition; if the
3770 condition is true, the result is "true". As before, when the second string
3771 is omitted, a false condition yields an empty string. This makes it less
3772 cumbersome to write custom ACL and router conditions.
3774 34. Failure to deliver a bounce message always caused it to be frozen, even if
3775 there was an errors_to setting on the router. The errors_to setting is now
3778 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3779 canonical form (fully expanded) before being placed in
3780 $sender_host_address.
3782 36. The table in the code that translates DNS record types into text (T_A to
3783 "A" for instance) was missing entries for NS and CNAME. It is just possible
3784 that this could have caused confusion if both these types were looked up
3785 for the same domain, because the text type is used as part of Exim's
3786 per-process caching. But the chance of anyone hitting this buglet seems
3789 37. The dnsdb lookup has been extended in a number of ways.
3791 (1) There is a new type, "zns", which walks up the domain tree until it
3792 finds some nameserver records. It should be used with care.
3794 (2) There is a new type, "mxh", which is like "mx" except that it returns
3795 just the host names, not the priorities.
3797 (3) It is now possible to give a list of domains (or IP addresses) to be
3798 looked up. The behaviour when one of the lookups defers can be
3799 controlled by a keyword.
3801 (4) It is now possible to specify the separator character for use when
3802 multiple records are returned.
3804 38. The dnslists ACL condition has been extended: it is now possible to supply
3805 a list of IP addresses and/or domains to be looked up in a particular DNS
3808 39. Added log_selector=+queue_time_overall.
3810 40. When running the queue in the test harness, wait just a tad after forking a
3811 delivery process, to get repeatability of debugging output.
3813 41. Include certificate and key file names in error message when GnuTLS fails
3814 to set them up, because the GnuTLS error message doesn't include the name
3815 of the failing file when there is a problem reading it.
3817 42. Allow both -bf and -bF in the same test run.
3819 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3821 44. The "Exiscan patch" is now merged into the mainline Exim source.
3823 45. Sometimes the final signoff response after QUIT could fail to get
3824 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3825 < 0 before doing a fflush(). This bug looks as though it goes back to the
3826 introduction of TLS in release 3.20, but "sometimes" must have been rare
3827 because the tests only now provoked it.
3829 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3830 (this can affect the format of dates).
3832 47. exim_tidydb, when checking for the continued existence of a message for
3833 which it has found a message-specific retry record, was not finding
3834 messages that were in split spool directories. Consequently, it was
3835 deleting retry records that should have stayed in existence.
3837 48. Steve fixed some bugs in eximstats.
3839 49. The SPA authentication driver was not abandoning authentication and moving
3840 on to the next authenticator when an expansion was forced to fail,
3841 contradicting the general specification for all authenticators. Instead it
3842 was generating a temporary error. It now behaves as specified.
3844 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3845 (the order specifies the preference for clients). The order is now AES256,
3846 AES128, 3DES, ARCFOUR128.
3848 51. Small patch to Sieve code - explicitly set From: when generating an
3851 52. Exim crashed if a remote delivery caused a very long error message to be
3852 recorded - for instance if somebody sent an entire SpamAssassin report back
3853 as a large number of 550 error lines. This bug was coincidentally fixed by
3854 increasing the size of one of Exim's internal buffers (big_buffer) that
3855 happened as part of the Exiscan merge. However, to be on the safe side, I
3856 have made the code more robust (and fixed the comments that describe what
3859 53. Now that there can be additional text after "Completed" in log lines (if
3860 the queue_time_overall log selector is set), a one-byte patch to exigrep
3861 was needed to allow it to recognize "Completed" as not the last thing in
3864 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3865 patch that reportedly fixes this has been added. I am not expert enough to
3866 create a test for it. This is what the patch creator wrote:
3868 "I found a little strange behaviour of ldap code when working with
3869 Windows 2003 AD Domain, where users was placed in more than one
3870 Organization Units. When I tried to give exim partial DN, the exit code
3871 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3872 But simultaneously result of request was absolutely normal ldap result,
3873 so I produce this patch..."
3875 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3876 REFERENCE, so I have modified the code to exclude the patch when that macro
3879 55. Some experimental protocols are using DNS PTR records for new purposes. The
3880 keys for these records are domain names, not reversed IP addresses. The
3881 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3882 leaves it alone. Component reversal etc. now happens only for IP addresses.
3885 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3887 57. Double the size of the debug message buffer (to 2048) so that more of very
3888 long debug lines gets shown.
3890 58. The exicyclog utility now does better if the number of log files to keep
3891 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3893 59. Two changes related to the smtp_active_hostname option:
3895 (1) $smtp_active_hostname is now available as a variable.
3896 (2) The default for smtp_banner uses $smtp_active_hostname instead
3897 of $primary_hostname.
3899 60. The host_aton() function is supposed to be passed a string that is known
3900 to be a valid IP address. However, in the case of IPv6 addresses, it was
3901 not checking this. This is a hostage to fortune. Exim now panics and dies
3902 if the condition is not met. A case was found where this could be provoked
3903 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3904 components; fortuitously, this particular loophole had already been fixed
3905 by change 4.50/55 above.
3907 If there are any other similar loopholes, the new check in host_aton()
3908 itself should stop them being exploited. The report I received stated that
3909 data on the command line could provoke the exploit when Exim was running as
3910 exim, but did not say which command line option was involved. All I could
3911 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3912 running as the user.
3915 61. There was a buffer overflow vulnerability in the SPA authentication code
3916 (which came originally from the Samba project). I have added a test to the
3917 spa_base64_to_bits() function which I hope fixes it.
3920 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3921 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3923 63. The daemon start-up calls getloadavg() while still root for those OS that
3924 need the first call to be done as root, but it missed one case: when
3925 deliver_queue_load_max is set with deliver_drop_privilege. This is
3926 necessary for the benefit of the queue runner, because there is no re-exec
3927 when deliver_drop_privilege is set.
3929 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3930 This has been fixed.
3932 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3933 was in use, was not putting the data itself into the right store pool;
3934 consequently, it could be overwritten for a subsequent message in the same
3935 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3938 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3940 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3941 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3942 the function were treating the return as a boolean value, which happened to
3943 work because 0=false and not-0=true, but is not correct code.
3945 68. The host_aton() function was not handling scoped IPv6 addresses (those
3946 with, for example, "%eth0" on the end) correctly.
3948 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3949 negated items (that is, ~something) in unsigned ints. Some compilers
3950 apparently mutter when there is no cast.
3952 70. If an address verification called from an ACL failed, and did not produce a
3953 user-specific message (i.e. there was only a "system" message), nothing was
3954 put in $acl_verify_message. In this situation, it now puts the system
3957 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3958 session; change 4.31/43 added the unwanted input to the log line - except
3959 that it did not do this in the start of session case. It now does.
3961 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3962 This is wrong and can cause the other end to generate a synchronization
3963 error if it is another Exim or anything else that does the synchronization
3964 check. A QUIT command is no longer sent after a timeout.
3966 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3967 during host lookups.
3969 74. The defer_ok option of callout verification was not working if it was used
3970 when verifying addresses in header lines, that is, for this case:
3972 verify = header_sender/callout=defer_ok
3974 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3975 those file descriptors could be used for SMTP connections. If anything
3976 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3977 could be sent to the SMTP client, causing chaos. The daemon now opens
3978 stdin, stdout, and stderr to /dev/null when it puts itself into the
3981 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3982 log by default. The user can override this with suitable Perl magic.
3984 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3985 the log message when discard triggers, was not working for the DATA ACL or
3986 for the non-SMTP ACL.
3988 78. Error message wording change in sieve.c.
3990 79. If smtp_accept_max_per_host was set, the number of connections could be
3991 restricted to fewer than expected, because the daemon was trying to set up
3992 a new connection before checking whether the processes handling previous
3993 connections had finished. The check for completed processes is now done
3994 earlier. On busy systems, this bug wouldn't be noticed because something
3995 else would have woken the daemon, and it would have reaped the completed
3998 80. If a message was submitted locally by a user whose login name contained one
3999 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4000 It caused a spool format error. I have fixed the spool reading code. A
4001 related problem was that the "from" clause in the Received: line became
4002 illegal because of the space(s). It is now covered by ${quote_local_part.
4004 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4007 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4009 83. Patch from Sieve maintainer to fix unterminated string problem in
4010 "vacation" handling.
4012 84. Some minor changes to the Linux configuration files to help with other
4013 OS variants using glibc.
4015 85. One more patch for Sieve to update vacation handling to latest spec.
4018 ----------------------------------------------------
4019 See the note above about the 4.44 and 4.50 releases.
4020 ----------------------------------------------------
4026 1. Change 4.43/35 introduced a bug that caused file counts to be
4027 incorrectly computed when quota_filecount was set in an appendfile
4030 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4031 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4034 3. Give more explanation in the error message when the command for a transport
4035 filter fails to execute.
4037 4. There are several places where Exim runs a non-Exim command in a
4038 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4039 was being done only for the command run by the queryprogram router. It is
4040 now done for all such subprocesses. The other cases are: ${run, transport
4041 filters, and the commands run by the lmtp and pipe transports.
4043 5. Some older OS have a limit of 256 on the maximum number of file
4044 descriptors. Exim was using setrlimit() to set 1000 as a large value
4045 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4046 systems. I've change it so that if it can't get 1000, it tries for 256.
4048 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4049 was an oversight, and furthermore, ever since the addition of extra
4050 controls (e.g. 4.43/32), the checks on when to allow different forms of
4051 "control" were broken. There should now be diagnostics for all cases when a
4052 control that does not make sense is encountered.
4054 7. $recipients is now available in the predata ACL (oversight).
4056 8. Tidy the search cache before the fork to do a delivery from a message
4057 received from the command line. Otherwise the child will trigger a lookup
4058 failure and thereby defer the delivery if it tries to use (for example) a
4059 cached ldap connection that the parent has called unbind on.
4061 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4062 of $address_data from the recipient verification was clobbered by the
4063 sender verification.
4065 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4066 was its contents. (It was OK if the option was not defined at all.)
4068 11. A "Completed" log line is now written for messages that are removed from
4069 the spool by the -Mrm option.
4071 12. $host_address is now set to the target address during the checking of
4072 ignore_target_hosts.
4074 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4075 being passed; this would have caused $sender_host_name to have been used if
4076 matching the list had actually called for a host name (not very likely,
4077 since this list is usually IP addresses). A host name is now passed as
4080 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4081 code that specifies a non-restarting handler (typically sigaction() in
4082 modern systems) in an attempt to fix a rare and obscure crash bug.
4084 15. Narrowed the window for a race in the daemon that could cause it to ignore
4085 SIGCHLD signals. This is not a major problem, because they are used only to
4086 wake it up if nothing else does.
4088 16. A malformed maildirsize file could cause Exim to calculate negative values
4089 for the mailbox size or file count. Odd effects could occur as a result.
4090 The maildirsize information is now recalculated if the size or filecount
4093 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4094 support for a long time. Removed HAVE_SYS_VFS_H.
4096 18. Updated exipick to current release from John Jetmore.
4098 19. Allow an empty sender to be matched against a lookup in an address list.
4099 Previously the only cases considered were a regular expression, or an
4102 20. Exim went into a mad DNS lookup loop when doing a callout where the
4103 host was specified on the transport, if the DNS lookup yielded more than
4106 21. The RFC2047 encoding function was originally intended for short strings
4107 such as real names; it was not keeping to the 75-character limit for
4108 encoded words that the RFC imposes. It now respects the limit, and
4109 generates multiple encoded words if necessary. To be on the safe side, I
4110 have increased the buffer size for the ${rfc2047: expansion operator from
4113 22. Failure to deliver a bounce message always caused it to be frozen, even if
4114 there was an errors_to setting on the router. The errors_to setting is now
4117 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4118 canonical form (fully expanded) before being placed in
4119 $sender_host_address.
4121 24. Updated eximstats to version 1.33
4123 25. Include certificate and key file names in error message when GnuTLS fails
4124 to set them up, because the GnuTLS error message doesn't include the name
4125 of the failing file when there is a problem reading it.
4127 26. Expand error message when OpenSSL has problems setting up cert/key files.
4130 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4131 (this can affect the format of dates).
4133 28. exim_tidydb, when checking for the continued existence of a message for
4134 which it has found a message-specific retry record, was not finding
4135 messages that were in split spool directories. Consequently, it was
4136 deleting retry records that should have stayed in existence.
4138 29. eximstats updated to version 1.35
4139 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4140 1.35 - bugfix such that pie charts by volume are generated correctly
4142 30. The SPA authentication driver was not abandoning authentication and moving
4143 on to the next authenticator when an expansion was forced to fail,
4144 contradicting the general specification for all authenticators. Instead it
4145 was generating a temporary error. It now behaves as specified.
4147 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4148 (the order specifies the preference for clients). The order is now AES256,
4149 AES128, 3DES, ARCFOUR128.
4151 31. Small patch to Sieve code - explicitly set From: when generating an
4154 32. Exim crashed if a remote delivery caused a very long error message to be
4155 recorded - for instance if somebody sent an entire SpamAssassin report back
4156 as a large number of 550 error lines. This bug was coincidentally fixed by
4157 increasing the size of one of Exim's internal buffers (big_buffer) that
4158 happened as part of the Exiscan merge. However, to be on the safe side, I
4159 have made the code more robust (and fixed the comments that describe what
4162 33. Some experimental protocols are using DNS PTR records for new purposes. The
4163 keys for these records are domain names, not reversed IP addresses. The
4164 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4165 leaves it alone. Component reversal etc. now happens only for IP addresses.
4168 34. The host_aton() function is supposed to be passed a string that is known
4169 to be a valid IP address. However, in the case of IPv6 addresses, it was
4170 not checking this. This is a hostage to fortune. Exim now panics and dies
4171 if the condition is not met. A case was found where this could be provoked
4172 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4173 components; fortuitously, this particular loophole had already been fixed
4174 by change 4.50/55 or 4.44/33 above.
4176 If there are any other similar loopholes, the new check in host_aton()
4177 itself should stop them being exploited. The report I received stated that
4178 data on the command line could provoke the exploit when Exim was running as
4179 exim, but did not say which command line option was involved. All I could
4180 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4181 running as the user.
4184 35. There was a buffer overflow vulnerability in the SPA authentication code
4185 (which came originally from the Samba project). I have added a test to the
4186 spa_base64_to_bits() function which I hope fixes it.
4189 36. The daemon start-up calls getloadavg() while still root for those OS that
4190 need the first call to be done as root, but it missed one case: when
4191 deliver_queue_load_max is set with deliver_drop_privilege. This is
4192 necessary for the benefit of the queue runner, because there is no re-exec
4193 when deliver_drop_privilege is set.
4195 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4196 was in use, was not putting the data itself into the right store pool;
4197 consequently, it could be overwritten for a subsequent message in the same
4198 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4201 38. Sometimes the final signoff response after QUIT could fail to get
4202 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4203 < 0 before doing a fflush(). This bug looks as though it goes back to the
4204 introduction of TLS in release 3.20, but "sometimes" must have been rare
4205 because the tests only now provoked it.
4211 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4212 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4213 Now it returns an integer. A number of calls were still expecting a T/F
4214 return. Fortuitously, in all cases, the tests worked in OK situations,
4215 which is the norm. However, things would have gone wrong on any write
4216 failures on the smtp file descriptor. This function is used when sending
4217 messages over SMTP and also when doing verify callouts.
4219 2. When Exim is called to do synchronous delivery of a locally submitted
4220 message (the -odf or -odi options), it no longer closes stderr before doing
4223 3. Implemented the mua_wrapper option.
4225 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4227 5. Implemented the functions header_remove(), header_testname(),
4228 header_add_at_position(), and receive_remove_recipient(), and exported them
4231 6. If an ACL "warn" statement specified the addition of headers, Exim already
4232 inserted X-ACL-Warn: at the start if there was no header name. However, it
4233 was not making this test for the second and subsequent header lines if
4234 there were newlines in the string. This meant that an invalid header could
4235 be inserted if Exim was badly configured.
4237 7. Allow an ACL "warn" statement to add header lines at the start or after all
4238 the Received: headers, as well as at the end.
4240 8. Added the rcpt_4xx retry error code.
4242 9. Added postmaster_mailfrom=xxx to callout verification option.
4244 10. Added mailfrom=xxxx to the callout verification option, for verify=
4247 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4248 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4250 12. Inserted some casts to stop certain compilers warning when using pointer
4251 differences as field lengths or precisions in printf-type calls (mostly
4252 affecting debugging statements).
4254 13. Added optional readline() support for -be (dynamically loaded).
4256 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4257 same clock tick as a message's arrival, so that its received time was the
4258 same as the "first fail" time on the retry record, and that message
4259 remained on the queue past the ultimate address timeout, every queue runner
4260 would try a delivery (because it was past the ultimate address timeout) but
4261 after another failure, the ultimate address timeout, which should have then
4262 bounced the address, did not kick in. This was a "< instead of <=" error;
4263 in most cases the first failure would have been in the next clock tick
4264 after the received time, and all would be well.
4266 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4267 being recognized when the domain list was tested by the match_domain
4268 condition in an expansion string.
4270 16. Added the ${str2b64: operator.
4272 17. Exim was always calling setrlimit() to set a large limit for the number of
4273 processes, without checking whether the existing limit was already
4274 adequate. (It did check for the limit on file descriptors.) Furthermore,
4275 errors from getrlimit() and setrlimit() were being ignored. Now they are
4276 logged to the main and panic logs, but Exim does carry on, to try to do its
4277 job under whatever limits there are.
4279 18. Imported PCRE 5.0.
4281 19. Trivial typo in log message " temporarily refused connection" (the leading
4284 20. If the log selector return_path_on_delivery was set and an address was
4285 redirected to /dev/null, the delivery process crashed because it assumed
4286 that a return path would always be set for a "successful" delivery. In this
4287 case, the whole delivery is bypassed as an optimization, and therefore no
4290 21. Internal re-arrangement: the function for sending a challenge and reading
4291 a response while authentication was assuming a zero-terminated challenge
4292 string. It's now changed to take a pointer and a length, to allow for
4293 binary data in such strings.
4295 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4297 23. Exim was not respecting finduser_retries when seeking the login of the
4298 uid under which it was called; it was always trying 10 times. (The default
4299 setting of finduser_retries is zero.) Also, it was sleeping after the final
4300 failure, which is pointless.
4302 24. Implemented tls_on_connect_ports.
4304 25. Implemented acl_smtp_predata.
4306 26. If the domain in control=submission is set empty, Exim assumes that the
4307 authenticated id is a complete email address when it generates From: or
4308 Sender: header lines.
4310 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4311 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4312 chown and chgrp in /bin and hostname in /usr/bin.
4314 28. Exim was keeping the "process log" file open after each use, just as it
4315 does for the main log. This opens the possibility of it remaining open for
4316 long periods when the USR1 signal hits a daemon. Occasional processlog
4317 errors were reported, that could have been caused by this. Anyway, it seems
4318 much more sensible not to leave this file open at all, so that is what now
4321 29. The long-running daemon process does not normally write to the log once it
4322 has entered its main loop, and it closes the log before doing so. This is
4323 so that log files can straightforwardly be renamed and moved. However,
4324 there are a couple of unusual error situations where the daemon does write
4325 log entries, and I had neglected to close the log afterwards.
4327 30. The text of an SMTP error response that was received during a remote
4328 delivery was being truncated at 512 bytes. This is too short for some of
4329 the long messages that one sometimes sees. I've increased the limit to
4332 31. It is now possible to make retry rules that apply only when a message has a
4333 specific sender, in particular, an empty sender.
4335 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4336 it possible to be selective about when SMTP synchronization is enforced.
4338 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4340 32. Implemented hosts_connection_nolog.
4342 33. Added an ACL for QUIT.
4344 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4347 35. Added mailbox_size and mailbox_filecount to appendfile.
4349 36. Added control = no_multiline_responses to ACLs.
4351 37. There was a bug in the logic of the code that waits for the clock to tick
4352 in the case where the clock went backwards by a substantial amount such
4353 that the microsecond fraction of "now" was more than the microsecond
4354 fraction of "then" (but the whole seconds number was less).
4356 38. Added support for the libradius Radius client library this is found on
4357 FreeBSD (previously only the radiusclient library was supported).
4363 1. When certain lookups returned multiple values in the form name=value, the
4364 quoting of the values was not always being done properly. Specifically:
4365 (a) If the value started with a double quote, but contained no whitespace,
4367 (b) If the value contained whitespace other than a space character (i.e.
4368 tabs or newlines or carriage returns) it was not quoted.
4369 This fix has been applied to the mysql and pgsql lookups by writing a
4370 separate quoting function and calling it from the lookup code. The fix
4371 should probably also be applied to nisplus, ibase and oracle lookups, but
4372 since I cannot test any of those, I have not disturbed their existing code.
4374 2. A hit in the callout cache for a specific address caused a log line with no
4375 reason for rejecting RCPT. Now it says "Previous (cached) callout
4376 verification failure".
4378 3. There was an off-by-one bug in the queryprogram router. An over-long
4379 return line was truncated at 256 instead of 255 characters, thereby
4380 overflowing its buffer with the terminating zero. As well as fixing this, I
4381 have increased the buffer size to 1024 (and made a note to document this).
4383 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4384 when Exim is waiting for an SMTP response from a remote server, Exim
4385 restarts its select() call on the socket, thereby resetting its timeout.
4386 This is not a problem when such interrupts are rare. Somebody set up a cron
4387 job to run exiwhat every 2 minutes, which is less than the normal select()
4388 timeout (5 or 10 minutes). This meant that the select() timeout never
4389 kicked in because it was always reset. I have fixed this by comparing the
4390 time when an interrupt arrives with the time at the start of the first call
4391 to select(). If more time than the timeout has elapsed, the interrupt is
4392 treated as a timeout.
4394 5. Some internal re-factoring in preparation for the addition of Sieve
4395 extensions (by MH). In particular, the "personal" test is moved to a
4396 separate function, and given an option for scanning Cc: and Bcc: (which is
4397 not set for Exim filters).
4399 6. When Exim created an email address using the login of the caller as the
4400 local part (e.g. when creating a From: or Sender: header line), it was not
4401 quoting the local part when it contained special characters such as @.
4403 7. Installed new OpenBSD configuration files.
4405 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4406 try to make them clearer.
4408 9. Callout options, other than the timeout value, were being ignored when
4409 verifying sender addresses in header lines. For example, when using
4411 verify = header_sender/callout=no_cache
4413 the cache was (incorrectly) being used.
4415 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4416 only the Cygwin environment.
4418 11. When return_path_on_delivery was set as a log selector, if different remote
4419 addresses in the same message used different return paths and parallel
4420 remote delivery occurred, the wrong values would sometimes be logged.
4421 (Whenever a remote delivery process finished, the return path value from
4422 the most recently started remote delivery process was logged.)
4424 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4425 lines when AUTH and/or TLS are in use. This is the "received protocol"
4426 field. Exim used to use "asmtp" for authenticated SMTP, without any
4427 indication (in the protocol name) for TLS use. Now it follows the RFC and
4428 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4429 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4430 names appear in log lines as well as in Received: header lines.
4432 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4433 extensions, and comparison tests, and to fix some bugs.
4435 14. Changes to the "personal" filter test:
4437 (1) The test was buggy in that it was just doing the equivalent of
4438 "contains" tests on header lines. For example, if a user's address was
4439 anne@some.where, the "personal" test would incorrectly be true for
4441 To: susanne@some.where
4443 This test is now done by extracting each address from the header in turn,
4444 and checking the entire address. Other tests that are part of "personal"
4445 are now done using regular expressions (for example, to check local parts
4446 of addresses in From: header lines).
4448 (2) The list of non-personal local parts in From: addresses has been
4449 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4450 taken from the Sieve specification recommendations.
4452 (3) If the message contains any header line starting with "List-" it is
4453 treated as non-personal.
4455 (4) The test for "circular" in the Subject: header line has been removed
4456 because it now seems ill-conceived.
4458 15. Minor typos in src/EDITME comments corrected.
4460 16. Installed latest exipick from John Jetmore.
4462 17. If headers_add on a router specified a text string that was too long for
4463 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4464 of string_sprintf() is now avoided.
4466 18. $message_body_size was not set (it was always zero) when running the DATA
4467 ACL and the local_scan() function.
4469 19. For the "mail" command in an Exim filter, no default was being set for
4470 the once_repeat time, causing a random time value to be used if "once" was
4471 specified. (If the value happened to be <= 0, no repeat happened.) The
4472 default is now 0s, meaning "never repeat". The "vacation" command was OK
4473 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4474 (I found it when inspecting the code).
4476 20. There is now an overall timeout for performing a callout verification. It
4477 defaults to 4 times the callout timeout, which applies to individual SMTP
4478 commands during the callout. The overall timeout applies when there is more
4479 than one host that can be tried. The timeout is checked before trying the
4480 next host. This prevents very long delays if there are a large number of
4481 hosts and all are timing out (e.g. when the network connections are timing
4482 out). The value of the overall timeout can be changed by specifying an
4483 additional sub-option for "callout", called "maxwait". For example:
4485 verify = sender/callout=5s,maxwait=20s
4487 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4488 to the end before writing, but this should make it even safer).
4490 22. Exim was forgetting that it had advertised PIPELINING for the second and
4491 subsequent messages on an SMTP connection. It was also not resetting its
4492 memory on STARTTLS and an internal HELO.
4494 23. When Exim logs an SMTP synchronization error within a session, it now
4495 records whether PIPELINING has been advertised or not.
4497 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4498 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4499 rather than long int.
4501 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4503 26. Added the never_mail option to autoreply.
4509 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4510 crash if the getsockname() call failed; this can happen if a connection is
4511 closed very soon after it is established. The problem was simply in the
4512 order in which certain operations were done, causing Exim to try to write
4513 to the SMTP stream before it had set up the file descriptor. The bug has
4514 been fixed by making things happen in the correct order.
4520 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4521 before the connection was closed, thus losing the rejection response.
4523 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4524 some early Solaris releases, but causes trouble in current releases where
4525 socklen_t is defined.
4527 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4530 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4533 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4534 that releasing the top of it at the end releases what was used for sub-
4535 expansions (unless the block got too big). However, discard this block if
4536 the first thing is a variable or header, so that we can use its block when
4537 it is dynamic (useful for very large $message_headers, for example).
4539 6. Lookups now cache *every* query, not just the most recent. A new, separate
4540 store pool is used for this. It can be recovered when all lookup caches are
4541 flushed. Lookups now release memory at the end of their result strings.
4542 This has involved some general refactoring of the lookup sources.
4544 7. Some code has been added to the store_xxx() functions to reduce the amount
4545 of flapping under certain conditions.
4547 8. log_incoming_interface used to affect only the <= reception log lines. Now
4548 it causes the local interface and port to be added to several more SMTP log
4549 lines, for example "SMTP connection from", and rejection lines.
4551 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4553 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4555 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4556 could be overwritten at the end of the current message (or the start of a
4557 new message if it was set in a HELO ACL). The value is now preserved for
4558 the duration of the SMTP connection.
4560 12. If a transport had a headers_rewrite setting, and a matching header line
4561 contained an unqualified address, that address was qualified, even if it
4562 did not match any rewriting rules. The underlying bug was that the values
4563 of the flags that permit the existence of unqualified sender and recipient
4564 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4565 non-local messages, and by -bnq for local messages) were not being
4566 preserved with the message after it was received.
4568 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4569 "next input=" as part of the text comprising the host identity instead of
4570 the correct text. The code was using the same buffer for two different
4571 strings. However, depending on which order the printing function evaluated
4572 its arguments, the bug did not always show up. Under Linux, for example, my
4573 test suite worked just fine.
4575 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4576 doesn't work with some older versions of Perl. It has been changed to "my",
4577 which in any case is probably the better facility to use.
4579 15. A really picky compiler found some instances of statements for creating
4580 error messages that either had too many or two few arguments for the format
4583 16. The size of the buffer for calls to the DNS resolver has been increased
4584 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4585 for addresses that have a lot of PTR records. This alleviates a problem; it
4586 does not fully solve it.
4588 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4589 buffer now truncates the list and logs the incident, which is the same
4590 action as happens when Exim is looking up a host name and its aliases.
4591 Previously in this situation something unpredictable would happen;
4592 sometimes it was "internal error: store_reset failed".
4594 18. If a server dropped the connection unexpectedly when an Exim client was
4595 using GnuTLS and trying to read a response, the client delivery process
4596 crashed while trying to generate an error log message.
4598 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4599 string, for example:
4601 warn message = H1: something\nH2: something
4603 the text was added as a single header line from Exim's point of view
4604 though it ended up OK in the delivered message. However, searching for the
4605 second and subsequent header lines using $h_h2: did not work. This has been
4606 fixed. Similarly, if a system filter added multiple headers in this way,
4607 the routers could not see them.
4609 20. Expanded the error message when iplsearch is called with an invalid key to
4610 suggest using net-iplsearch in a host list.
4612 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4613 ACLs are no longer actually imposed (and a message to that effect is
4616 22. If a "gecos" field in a passwd entry contained escaped characters, in
4617 particular, if it contained a \" sequence, Exim got it wrong when building
4618 a From: or a Sender: header from that name. A second bug also caused
4619 incorrect handling when an unquoted " was present following a character
4620 that needed quoting.
4622 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4623 was not being matched caselessly.
4625 24. Arranged for all hyphens in the exim.8 source to be escaped with
4628 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4629 properly. Recipient callout cache records were still being keyed to include
4630 the sender, even when use_sender was set false. This led to far more
4631 callouts that were necessary. The sender is no longer included in the key
4632 when use_sender is false.
4634 26. Added "control = submission" modifier to ACLs.
4636 27. Added the ${base62d: operator to decode base 62 numbers.
4638 28. dnsdb lookups can now access SRV records.
4640 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4641 the configuration file.
4643 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4644 (-v) mode. This makes the output for a verbose queue run more intelligible.
4646 31. Added a use_postmaster feature to recipient callouts.
4648 32. Added the $body_zerocount variable, containing the number of binary zero
4649 bytes in the message body.
4651 33. The time of last modification of the "new" subdirectory is now used as the
4652 "mailbox time last read" when there is a quota error for a maildir
4655 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4657 35. Added +ignore_unknown as a special item in host lists.
4659 36. Code for decoding IPv6 addresses in host lists is now included, even if
4660 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4661 address was recognized as an IP address, but was then not correctly decoded
4662 into binary, causing unexpected and incorrect effects when compared with
4669 1. Very minor rewording of debugging text in manualroute to say "list of
4670 hosts" instead of "hostlist".
4672 2. If verify=header_syntax was set, and a header line with an unqualified
4673 address (no domain) and a large number of spaces between the end of the
4674 name and the colon was received, the reception process suffered a buffer
4675 overflow, and (when I tested it) crashed. This was caused by some obsolete
4676 code that should have been removed. The fix is to remove it!
4678 3. When running in the test harness, delay a bit after writing a bounce
4679 message to get a bit more predictability in the log output.
4681 4. Added a call to search_tidyup() just before forking a reception process. In
4682 theory, someone could use a lookup in the expansion of smtp_accept_max_
4683 per_host which, without the tidyup, could leave open a database connection.
4685 5. Added the variables $recipient_data and $sender_data which get set from a
4686 lookup success in an ACL "recipients" or "senders" condition, or a router
4687 "senders" option, similar to $domain_data and $local_part_data.
4689 6. Moved the writing of debug_print from before to after the "senders" test
4692 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4693 problems for message scanning, either using a data ACL, or using
4694 local_scan() because the Received: header was not generated till after they
4695 were called (in order to set the time as the time of reception completion).
4696 I have revised the way this works. The header is now generated after the
4697 body is received, but before the ACL or local_scan() are called. After they
4698 are run, the timestamp in the header is updated.
4704 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4705 before starting a queue runner without re-exec. This happened only when
4706 deliver_drop_privilege was set or when the Exim user was set to root. The
4707 effect of the bug was that timeouts during subsequent deliveries caused
4708 crashes instead of being properly handled. The handler is now left at its
4709 default (and expected) setting.
4711 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4712 message, again when deliver_drop_privilege is set or Exim is run as root.
4713 The bug described in (1) was not present in this case, but the tidying up
4714 of the other signals was missing. I have made the two cases consistent.
4716 3. The ignore_target_hosts setting on a manualroute router was being ignored
4717 for hosts that were looked up using the /MX notation.
4719 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4722 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4723 operated on the sender address. After changing the $sender_address to <>
4724 for the sender address verify, Exim was re-instated it as the original
4725 (before rewriting) address, but remembering that it had rewritten it, so it
4726 wasn't rewriting it again. This bug also had the effect of breaking the
4727 sender address verification caching when the sender address was rewritten.
4729 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4730 This has been changed so that if the ip literal address matches
4731 ignore_target_hosts, the router declines.
4733 7. Added expansion conditions match_domain, match_address, and match_local_
4734 part (NOT match_host).
4736 8. The placeholder for the Received: header didn't have a length field set.
4738 9. Added code to Exim itself and to exim_lock to test for a specific race
4739 condition that could lead to file corruption when using MBX delivery. The
4740 issue is with the lockfile that is created in /tmp. If this file is removed
4741 after a process has opened it but before that process has acquired a lock,
4742 there is the potential for a second process to recreate the file and also
4743 acquire a lock. This could lead to two Exim processes writing to the file
4744 at the same time. The added code performs the same test as UW imapd; it
4745 checks after acquiring the lock that its file descriptor still refers to
4746 the same named file.
4748 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4749 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4752 11. Added the smtp_active_hostname option. If used, this will typically be made
4753 to depend on the incoming interface address. Because $interface_address is
4754 not set up until the daemon has forked a reception process, error responses
4755 that can happen earlier (such as "too many connections") no longer contain
4758 12. If an expansion in a condition on a "warn" statement fails because a lookup
4759 defers, the "warn" statement is abandoned, and the next ACL statement is
4760 processed. Previously this caused the whole ACL to be aborted.
4762 13. Added the iplsearch lookup type.
4764 14. Added ident_timeout as a log selector.
4766 15. Added tls_certificate_verified as a log selector.
4768 16. Added a global option tls_require_ciphers (compare the smtp transport
4769 option of the same name). This controls incoming TLS connections.
4771 17. I finally figured out how to make tls_require_ciphers do a similar thing
4772 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4773 before starting the TLS session.
4775 18. Tabs are now shown as \t in -bP output.
4777 19. If the log selector return_path_on_delivery was set, Exim crashed when
4778 bouncing a message because it had too many Received: header lines.
4780 20. If two routers both had headers_remove settings, and the first one included
4781 a superfluous trailing colon, the final name in the first list and the
4782 first name in the second list were incorrectly joined into one item (with a
4783 colon in the middle).
4789 1. Added -C and -D options to the exinext utility, mainly to make it easier
4790 to include in the automated testing, but these could be helpful when
4791 multiple configurations are in use.
4793 2. The exinext utility was not formatting the output nicely when there was
4794 an alternate port involved in the retry record key, nor when there was a
4795 message id as well (for retries that were specific to a specific message
4796 and a specific host). It was also confused by IPv6 addresses, because of
4797 the additional colons they contain. I have fixed the IPv4 problem, and
4798 patched it up to do a reasonable job for IPv6.
4800 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4801 delivery, the log line now contains "pipelined" if PIPELINING was used.
4803 4. An SMTP transport process used to panic and die if the bind() call to set
4804 an explicit outgoing interface failed. This has been changed; it is now
4805 treated in the same way as a connect() failure.
4807 5. A reference to $sender_host_name in the part of a conditional expansion
4808 that was being skipped was still causing a DNS lookup. This no longer
4811 6. The def: expansion condition was not recognizing references to header lines
4812 that used bh_ and bheader_.
4814 7. Added the _cache feature to named lists.
4816 8. The code for checking quota_filecount in the appendfile transport was
4817 allowing one more file than it should have been.
4819 9. For compatibility with Sendmail, the command line option
4827 and sets the incoming protocol and host name (for trusted callers). The
4828 host name and its colon can be omitted when only the protocol is to be set.
4829 Note the Exim already has two private options, -pd and -ps, that refer to
4830 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4831 "s", but I don't think that's a major issue.
4833 10. A number of refactoring changes to the code, none of which should affect
4836 (a) The number of logging options was getting close to filling up the
4837 32-bit word that was used as a bit map. I have split them into two classes:
4838 those that are passed in the argument to log_write(), and those that are
4839 only ever tested independently outside of that function. These are now in
4840 separate 32-bit words, so there is plenty of room for expansion again.
4841 There is no change in the user interface or the logging behaviour.
4843 (b) When building, for example, log lines, the code previously used a
4844 macro that called string_cat() twice, in order to add two strings. This is
4845 not really sufficiently general. Furthermore, there was one instance where
4846 it was actually wrong because one of the argument was used twice, and in
4847 one call a function was used. (As it happened, calling the function twice
4848 did not affect the overall behaviour.) The macro has been replaced by a
4849 function that can join an arbitrary number of extra strings onto a growing
4852 (c) The code for expansion conditions now uses a table and a binary chop
4853 instead of a serial search (which was left over from when there were very
4854 few conditions). Also, it now recognizes conditions like "pam" even when
4855 the relevant support is not compiled in: a suitably worded error message is
4856 given if an attempt is made to use such a condition.
4858 11. Added ${time_interval:xxxxx}.
4860 12. A bug was causing one of the ddress fields not to be passed back correctly
4861 from remote delivery subprocesses. The field in question was not being
4862 subsequently used, so this caused to problems in practice.
4864 13. Added new log selectors queue_time and deliver_time.
4866 14. Might have fixed a bug in maildirsizefile handling that threw up
4867 "unexpected character" debug warnings, and recalculated the data
4868 unnecessarily. In any case, I expanded the warning message to give more
4871 15. Added the message "Restricted characters in address" to the statements in
4872 the default ACL that block characters like @ and % in local parts.
4874 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4875 Three changes have been made:
4877 (a) There was a serious bug; a negative response to MAIL caused the whole
4878 recipient domain to be cached as invalid, thereby blocking all messages
4879 to all local parts at the same domain, from all senders. This bug has
4880 been fixed. The domain is no longer cached after a negative response to
4881 MAIL if the sender used is not empty.
4883 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4886 (c) A new callout option, "use_sender" has been added for people who want
4887 the modified behaviour.
4893 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4896 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4897 indeed breaks things for older releases.
4899 3. Added additional logging to the case where there is a problem reading data
4900 from a filter that is running in a subprocess using a pipe, in order to
4901 try to track down a specific problem.
4903 4. Testing facility fudge: when running in the test harness and attempting
4904 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4905 getting "No route to host". Convert this to a timeout.
4907 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4910 6. Some OS don't have socklen_t but use size_t instead. This affects the
4911 fifth argument of getsockopt() amongst other things. This is now
4912 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4913 can be set for individual OS. I have set it for SunOS5, OSF1, and
4914 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4915 some earlier ones do not.
4917 7. Change 4.30/15 was not doing the test caselessly.
4919 8. The standard form for an IPv6 address literal was being rejected by address
4920 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4921 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4922 this, as well as the form without the "IPv6" on the front (but only when
4923 address literals are enabled, of course).
4925 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4927 10. Exim crashed if a message with an empty sender address specified by -f
4928 encountered a router with an errors_to setting. This could be provoked only
4929 by a command such as
4933 where an empty string was supplied; "<>" did not hit this bug.
4935 11. Installed PCRE release 4.5.
4937 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4938 remained set. It is now erased.
4940 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4941 times from message ids (which are base 36 rather than the normal 62).
4943 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4944 were being counted as actual protocol errors, and logged if the log
4945 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4946 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4947 and DATA following a set of rejected RCPTs do not count as protocol errors.
4948 In other words, Exim assumes they were pipelined, though this may not
4949 actually be the case. Of course, in all cases the client gets an
4950 appropriate error code.
4952 15. If a lookup fails in an ACL condition, a message about the failure may
4953 be available; it is used if testing the ACL cannot continue, because most
4954 such messages specify what the cause of the deferral is. However, some
4955 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4956 that caused an old message to be retained and used if a later statement
4957 caused a defer, replacing the real cause of the deferral.
4959 16. If an IP address had so many PTR records that the DNS lookup buffer
4960 was not large enough to hold them, Exim could crash while trying to process
4961 the truncated data. It now detects and logs this case.
4963 17. Further to 4.21/58, another change has been made: if (and only if) the
4964 first line of a message (the first header line) ends with CRLF, a bare LF
4965 in a subsequent header line has a space inserted after it, so as not to
4966 terminate the header.
4968 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4969 unnecessarily, used atoi() instead of strtol(), and didn't check the
4970 termination when getting file sizes from file names by regex.
4972 19. Completely re-implemented the support for maildirsize files, in the light
4973 of a number of problems with the previous contributed implementation
4974 (4.30/29). In particular:
4976 . If the quota is zero, the maildirsize file is maintained, but no quota is
4979 . If the maildir directory does not exist, it is created before any attempt
4980 to write a maildirsize file.
4982 . The quota value in the file is just a cache; if the quota is changed in
4983 the transport, the new value overrides.
4985 . A regular expression is available for excluding directories from the
4988 20. The autoreply transport checks the characters in options that define the
4989 message's headers; it allows continued headers, but it was checking with
4990 isspace() after an embedded newline instead of explicitly looking for a
4993 21. If all the "regular" hosts to which an address was routed had passed their
4994 expiry times, and had not reached their retry times, the address was
4995 bounced, even if fallback hosts were defined. Now Exim should go on to try
4998 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4999 equivalent code in the SMTP transport. Some hosts send humungous responses
5000 to HELO/EHLO, more than 1024 it seems.
5002 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5003 gives compiler warnings in some environments. I've now done it "properly",
5006 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5007 (because of problems with the built-in one) was declared to return uschar *
5008 instead of char *, causing compiler failure.
5010 25. Fixed a file descriptor leak when processing alias/forward files.
5012 26. Fixed a minor format string issue in dbfn.c.
5014 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5016 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5017 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5020 29. When a sender address is verified, it is cached, to save repeating the test
5021 when there is more than one recipient in a message. However, when the
5022 verification involves a callout, it is possible for different callout
5023 options to be set for different recipients. It is too complicated to keep
5024 track of this in the cache, so now Exim always runs a verification when a
5025 callout is required, relying on the callout cache for the optimization.
5026 The overhead is duplication of the address routing, but this should not be
5029 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5030 to be verified with callout=postmaster, and the main callout worked but the
5031 postmaster check failed, the verification correctly failed. However, if a
5032 subsequent RCPT command asked for sender verification *without* the
5033 postmaster check, incorrect caching caused this verification also to fail,
5036 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5037 it was not caching the DNS options (qualify_single, search_parents) that
5038 were used when the lookup failed. A subsequent lookup with different
5039 options therefore always gave the same answer, though there were cases
5040 where it should not have. (Example: a "domains = !$mx_any" option on a
5041 dnslookup router: the "domains" option is always processed without any
5042 widening, but the router might have qualify_single set.) Now Exim uses the
5043 cached value only when the same options are set.
5045 32. Added John Jetmore's "exipick" utility to the distribution.
5047 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5048 than a timeout (e.g. a certificate is required, and is not provided), an
5049 Exim server now closes the connection immediately. Previously it waited for
5050 the client to close - but if the client is SSL, it seems that they each
5051 wait for each other, leading to a delay before one of them times out.
5053 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5054 maintained 0.8.x compatibility because I don't think many are using it, and
5055 it is clearly obsolete.
5057 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5060 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5061 client certificate was expired. A simple patch fixes this, though I don't
5062 understand the full logic of why the verify callback is called multiple
5065 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5066 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5067 which causes problems with some clients (such as the Certicom SSL Plus
5068 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5069 disables the coutermeasure allowing Eudora to connect."
5071 38. Exim was not checking that a write() to a log file succeeded. This could
5072 lead to Bad Things if a log got too big, in particular if it hit a file
5073 size limit. Exim now panics and dies if it cannot write to a log file, just
5074 as it does if it cannot open a log file.
5076 39. Modified OS/Makefile-Linux so that it now contains
5078 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5080 The two -D definitions ensure that Exim is compiled with large file
5081 support, which makes it possible to handle log files that are bigger than
5084 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5085 instance) a domain was checked against a named list that involved a lookup,
5086 causing $domain_data to be set, then another domain was checked against the
5087 same list, then the first domain was re-checked, the value of $domain_data
5088 after the final check could be wrong. In particular, if the second check
5089 failed, it could be set empty. This bug probably also applied to
5092 41. The strip_trailing_dot option was not being applied to the address given
5093 with the -f command-line option.
5095 42. The code for reading a message's header from the spool was incrementing
5096 $received_count, but never initializing it. This meant that the value was
5097 incorrect (doubled) while delivering a message in the same process in which
5098 it was received. In the most common configuration of Exim, this never
5099 happens - a fresh exec is done - but it can happen when
5100 deliver_drop_privilege is set.
5102 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5103 it now includes up to 150 characters of the unexpected data in the log
5106 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5107 and building data strings. The size of both of these buffers was 10 000
5108 bytes - far larger than anybody would *ever* want, thought I. Needless to
5109 say, somebody hit the limit. I have increased the maximum line length to
5110 20 000 and the maximum data length of concatenated lines to 100 000. I have
5111 also fixed two bugs, because there was no checking on these buffers. Tsk,
5112 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5113 buffer is too small.
5115 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5116 lsearch lookups. Now it does.
5118 46. When parsing a route_list item in a manualroute router, a fixed-length
5119 buffer was used for the list of hosts. I made this 1024 bytes long,
5120 thinking that nobody would ever have a list of hosts that long. Wrong.
5121 Somebody had a whole pile of complicated expansion conditions, and the
5122 string was silently truncated, leading to an expansion error. It turns out
5123 that it is easier to change to an unlimited length (owing to other changes
5124 that have happened since this code was originally written) than to build
5125 structure for giving a limitation error. The length of the item that
5126 expands into the list of hosts is now unlimited.
5128 47. The lsearch lookup could not handle data where the length of text line was
5129 more than 4095 characters. Such lines were truncated, leading to shortened
5130 data being returned. It should now handle lines of any length.
5132 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5133 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5136 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5137 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5138 no longer happens in comment lines. A list of replacements is now placed
5139 at the head of all of the source files, except those whose only change is
5140 to replace PERL_COMMAND in the very first #! line.
5142 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5143 messages on the queue, with a bottom-up merge sort, using code contributed
5144 by Michael Haardt. This should make operations like -bp somewhat faster on
5145 large queues. It won't affect queue runners, except when queue_run_in_order
5148 51. Installed eximstats 1.31 in the distribution.
5150 52. Added support for SRV lookups to the dnslookup router.
5152 53. If an ACL referred to $message_body or $message_body_end, the value was not
5153 reset for any messages that followed in the same SMTP session.
5155 54. The store-handling optimization for building very long strings was not
5156 differentiating between the different store pools. I don't think this
5157 actually made any difference in practice, but I've tidied it.
5159 55. While running the routers to verify a sender address, $sender_address
5160 was still set to the sender address. This is wrong, because when routing to
5161 send a bounce to the sender, it would be empty. Therefore, I have changed
5162 it so that, while verifying a sender address, $sender_address is set to <>.
5163 (There is no change to what happens when verifying a recipient address.)
5165 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5166 target A or AAAA records (if not already returned) without resetting the
5167 qualify_single or search_parents options of the DNS resolver. These are
5168 inappropriate in this case because the targets of MX and SRV records must
5169 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5170 target that, when qualified, matched something in the local domain. These
5171 two options are now turned off when doing these lookups.
5173 57. It seems that at least some releases of Reiserfs (which does not have the
5174 concept of a fixed number of inodes) returns zero and not -1 for the
5175 number of available inodes. This interacted badly with check_spool_inodes,
5176 which assumed that -1 was the "no such thing" setting. What I have done is
5177 to check that the total number of inodes is greater than zero before doing
5178 the test of how many are available.
5180 58. When a "warn" ACL statement has a log_message modifier, the message is
5181 remembered, and not repeated. This is to avoid a lot of repetition when a
5182 message has many recipients that cause the same warning to be written.
5183 Howewer, Exim was preserving the list of already written lines for an
5184 entire SMTP session, which doesn't seem right. The memory is now reset if a
5185 new message is started.
5187 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5188 debugging output unless log_rewrite was also set.
5190 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5191 of (int)(handle) when we know that handle contains (void *)(-1).
5193 61. The Exim daemon panic-logs an error return when it closes the incoming
5194 connection. However "connection reset by peer" seems to be common, and
5195 isn't really an error worthy of noting specially, so that particular error
5198 62. When Exim is trying to find all the local interfaces, it used to panic and
5199 die if the ioctl to get the interface flags failed. However, it seems that
5200 on at least one OS (Solaris 9) it is possible to have an interface that is
5201 included in the list of interfaces, but for which you get a failure error
5202 for this call. This happens when the interface is not "plumbed" into a
5203 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5204 failure of the "get flags" call assumes that the interface is down.
5206 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5207 makes life easier for people who are doing arithmetic on fields extracted
5208 from dates, where you often get leading zeros that should not be
5209 interpreted as octal.
5211 64. Added qualify_domain to the redirect router, to override the global
5214 65. If a pathologically long header line contained very many addresses (the
5215 report of this problem mentioned 10 000) and each of them was rewritten,
5216 Exim could use up a very large amount of memory. (It kept on making new
5217 copies of the header line as it rewrote, and never released the old ones.)
5218 At the expense of a bit more processing, the header rewriting function has
5219 been changed so that it no longer eats memory in this way.
5221 66. The generation of the Received: header has been moved from the time that a
5222 message starts to be received, to the time that it finishes. The timestamp
5223 in the Received: header should now be very close to that of the <= log
5224 line. There are two side-effects of this change:
5226 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5227 logged header lines no longer include the local Received: line, because
5228 it has not yet been created. The same applies to a copy of the message
5229 that is returned to a non-SMTP sender when a message is rejected.
5231 (b) When a filter file is tested using -bf, no additional Received: header
5232 is added to the test message. After some thought, I decided that this
5235 This change does not affect the value of $received_for. It is still set
5236 after address rewriting, but before local_scan() is called.
5238 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5240 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5241 gave an unhelpful panic error message, and a defer error. I have managed to
5242 change this behaviour so that it now rejects any supplied certificate,
5243 which seems right, as the list of acceptable certificates is empty.
5245 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5246 gave an unhelpful defer error. I have not managed to make this reject any
5247 supplied certificates, but the error message it gives is "no certificate
5248 supplied", which is not helpful.
5250 70. exigrep's output now also includes lines that are not associated with any
5251 message, but which match the given pattern. Implemented by a patch from
5252 Martin Sluka, which also tidied up the Perl a bit.
5254 71. Recipient callout verification, like sender verification, was using <> in
5255 the MAIL FROM command. This isn't really the right thing, since the actual
5256 sender may affect whether the remote host accepts the recipient or not. I
5257 have changed it to use the actual sender in the callout; this means that
5258 the cache record is now keyed on a recipient/sender pair, not just the
5259 recipient address. There doesn't seem to be a real danger of callout loops,
5260 since a callout by the remote host to check the sender would use <>.
5261 [SEE ABOVE: changed after hitting problems.]
5263 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5264 temporary errors. However, in the case of such a code being given after
5265 the end of a data transmission (i.e. after ".") Exim was failing to write
5266 a retry record for the message. (Yes, there was some broken host that was
5267 actually sending 8xx at this point.)
5269 73. An unknown lookup type in a host list could cause Exim to panic-die when
5270 the list was checked. (An example that provoked this was putting <; in the
5271 middle of a list instead of at the start.) If this happened during a DATA
5272 ACL check, a -D file could be left lying around. This kind of configuration
5273 error no longer causes Exim to die; instead it causes a defer errror. The
5274 incident is still logged to the main and panic logs.
5276 74. Buglet left over from Exim 3 conversion. The message "too many messages
5277 in one connection" was written to the rejectlog but not the mainlog, except
5278 when address rewriting (yes!) was being logged.
5280 75. Added write_rejectlog option.
5282 76. When a system filter was run not as root (that is, when system_filter_user
5283 was set), the values of the $n variables were not being returned to the
5284 main process; thus, they were not subsequently available in the $sn
5287 77. Added +return_path_on_delivery log selector.
5289 78. A connection timeout was being treated differently from recipients deferred
5290 when testing hosts_max_try with a message that was older than the host's
5291 retry timeout. (The host should not be counted, thus allowing all hosts to
5292 be tried at least once before bouncing.) This may have been the cause of an
5293 occasionally reported bug whereby a message would remain on the queue
5294 longer than the retry timeout, but would be bounced if a delivery was
5295 forced. I say "may" because I never totally pinned down the problem;
5296 setting up timeout/retry tests is difficult. See also the next item.
5298 79. The ultimate address timeout was not being applied to errors that involved
5299 a combination of host plus message (for example, a timeout on a MAIL
5300 command). When an address resolved to a number of possible hosts, and they
5301 were not all tried for each delivery (e.g. because of hosts_max_try), a
5302 message could remain on the queue longer than the retry timeout.
5304 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5307 81. Fixed an obscure SMTP outgoing bug which required at least the following
5308 conditions: (a) there was another message waiting for the same server;
5309 (b) the server returned 5xx to all RCPT commands in the first message so
5310 that the message was not completed; (c) the server dropped the connection
5311 or gave a negative response to the RSET that Exim sends to abort the
5312 transaction. The observed case was a dropped connection after DATA that had
5313 been sent in pipelining mode. That is, the server had advertised PIPELINING
5314 but was not implementing it correctly. The effect of the bug was incorrect
5315 behaviour, such as trying another host, and this could lead to a crash.
5321 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5322 and daemon.c were passed as pointers to ints; they should have been
5323 pointers to socklen_t variables (which are typically unsigned ints).
5325 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5328 3. Fixed a really odd bug that affected only the testing scheme; patching a
5329 certain fixed string in the binary changed the value of another string that
5330 happened to be identical to the end of the original first string.
5332 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5333 name", it returns that address as the IP address. On some operating
5334 systems (e.g. Solaris), it also passes back the IP address string as the
5335 "host name". However, on others (e.g. Linux), it passes back an empty
5336 string. Exim wasn't checking for this, and was changing the host name to an
5337 empty string, assuming it had been canonicized.
5339 5. Although rare, it is permitted to have more than one PTR record for a given
5340 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5341 all the names associated with an address, because they do in Solaris.
5342 However, it seems that they do not in Linux for data that comes from the
5343 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5344 I found this out when I moved to a new Linux workstation and tried to run
5345 the Exim test suite.
5347 To get round this problem I have changed the code so that it now does its
5348 own call to the DNS to look up PTR records when searching for a host name.
5349 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5350 addresses that are only in /etc/hosts are still found.
5352 This behaviour is, however, controlled by an option called host_lookup_
5353 order, which defaults to "bydns:byaddr". If people want to use the other
5354 order, or indeed, just use one or the other means of lookup, they can
5355 specify it in this variable.
5357 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5358 some operating systems, this comes back from gethostbyaddr() as an empty
5359 string, and this is what Exim used to test for. However, it seems that in
5360 other systems, "." is yielded. Exim now tests for this case too.
5362 7. The values of check_spool_space and check_log_space are now held internally
5363 as a number of kilobytes instead of an absolute number of bytes. If a
5364 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5365 kilobyte. This means that much larger values can be stored.
5367 8. Exim monitor: an attempt to get the action menu when not actually pointing
5368 at a message produces an empty menu entitled "No message selected". This
5369 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5370 no entries in it ("Shell widget menu has zero width and/or height"). So I
5371 have added a single, blank menu entry in this case.
5373 9. Added ${quote_local_part.
5375 10. MIME decoding is now applied to the contents of Subject: header lines when
5378 11. Now that a reference to $sender_host_address automatically causes a reverse
5379 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5380 host lookup before query-style lookups in lists that might use this
5381 variable. This has therefore been abolished, and the "net-" prefix is no
5382 longer necessary for query-style lookups.
5384 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5385 have been a typo for LFLAGS, so it has been changed.
5387 13. The install script calls Exim with "-C /dev/null" in order to find the
5388 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5389 to be output. Howeve, since Exim outputs its version number before the
5390 error, it didn't break the script. It just looked ugly. I fixed this by
5391 always allowing "-C /dev/null" if the caller is root.
5393 14. Ignore overlarge ACL variable number when reading spool file - insurance
5394 against a later release with more variables having written the file.
5396 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5397 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5398 this, as well as the form without the "IPv6" on the front.
5400 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5401 OS/Makefile-Darwin file.
5403 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5404 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5406 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5407 variable. However, some OS use ints for inodes. I've added cast to long int
5408 to get rid of the compiler warning.
5410 19. I had forgotten to lock out "/../" in configuration file names when
5411 ALT_CONFIG_PREFIX was set.
5413 20. Routers used for verification do not need to specify transports. However,
5414 if such a router generated a host list, and callout was configured, Exim
5415 crashed, because it could not find a port number from the (non-existent)
5416 transport. It now assumes port 25 in this circumstance.
5418 21. Added the -t option to exigrep.
5420 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5421 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5422 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5423 them is set, LOOKUP_LSEARCH is forced.
5425 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5427 24. Added sender and host information to the "rejected by local_scan()" log
5428 line; previously there was no indication of these.
5430 25. Added .include_if_exists.
5432 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5433 Linux. It turns out that not all file systems support this. Apparently some
5434 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5435 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5436 not supported on the file descriptor, is now ignored when Exim is trying to
5437 sync a directory. This applies only to Linux.
5439 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5441 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5442 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5445 29. Added support for maildirsize files from supplied patch (modified a bit).
5447 30. The use of :fail: followed by an empty string could lead Exim to respond to
5448 sender verification failures with (e.g.):
5450 550 Verification failed for <xxx>
5451 550 Sender verify failed
5453 where the first response line was missing the '-' that indicates it is not
5454 the final line of the response.
5456 31. The loop for finding the name of the user that called Exim had a hardwired
5457 limit of 10; it now uses the value of finduser_retries, which is used for
5458 all other user lookups.
5460 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5463 33. Exim was neglecting to zero errno before one call of strtol() when
5464 expanding a string and expecting an integer value. On some systems this
5465 resulted in spurious "integer overflow" errors. Also, it was casting the
5466 result into an int without checking.
5468 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5469 did not work. The code looks as if it has *never* worked, though it appears
5470 to have been documented since at least releast 1.62. I have made it work.
5472 35. The "timeout_DNS" error in retry rules, also documented since at least
5473 1.62, also never worked. As it isn't clear exactly what this means, and
5474 clearly it isn't a major issue, I have abolished the feature by treating it
5475 as "timeout", and writing a warning to the main and panic logs.
5477 36. The display of retry rules for -brt wasn't always showing the error code
5480 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5481 timeout_connect_A, timeout_connect_MX.
5483 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5484 to the empty sender.
5486 39. The daemon was not analysing the content of -oX till after it had closed
5487 stderr and disconnected from the controlling terminal. This meant that any
5488 syntax errors were only noted on the panic log, and the return code from
5489 the command was 0. By re-arranging the code a little, I've made the
5490 decoding happen first, so such errors now appear on stderr, and the return
5491 code is 1. However, the actual setting up of the sockets still happens in
5492 the disconnected process, so errors there are still only recorded on the
5495 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5496 connections (as happens on some IP stacks) was logged at start up time as
5497 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5498 it from "IPv6 and IPv4", which means that two separate sockets are being
5501 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5502 says whether AF_INET or AF_INET6 was passed as an argument.
5504 42. Exiwhat output was messed up when time zones were included in log
5507 43. Exiwhat now gives more information about the daemon's listening ports,
5508 and whether -tls-on-connect was used.
5510 44. The "port" option of the smtp transport is now expanded.
5512 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5513 silently ignored. Now an error message is written to the main and panic
5516 46. There's a new ACL modifier called "logwrite" which writes to a log file
5517 as soon as it is encountered.
5519 47. Added $local_user_uid and $local_user_gid at routing time.
5521 48. Exim crashed when trying to verify a sender address that was being
5524 49. Exim was recognizing only a space character after ".include". It now also
5525 recognizes a tab character.
5527 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5528 extracting the relevant information from the specification. The man page no
5529 longer contains scrambled data for the -d option, and I've added a section
5530 at the front about calling Exim under different names.
5532 51. Added "extra_headers" argument to the "mail" command in filter files.
5534 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5537 53. Installed eximstats 1.29.
5539 54. Added transport_filter_timeout as a generic transport option.
5541 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5542 Cc: header lines. This was required by RFC 822, but it not required by RFC
5545 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5546 incoming messages that did not have them. Now it does so only if the
5547 message originates locally, that is, if there is no associated remote host
5548 address. When Resent- header lines are present, this applies to the Resent-
5549 lines rather than the non-Resent- lines.
5551 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5552 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5554 58. Messages for configuration errors now include the name of the main
5555 configuration file - useful now that there may be more than one file in a
5556 list (.included file names were always shown).
5558 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5559 for those rare installations that do not start the daemon as root or run it
5560 setuid root. I've cut out the call to initgroups() if the daemon is not
5563 60. The Exim user and group can now be bound into the binary as text strings
5564 that are looked up at the start of Exim's processing.
5566 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5568 62. Added $mailstore_basename variable.
5570 63. Installed patch to sieve.c from Michael Haardt.
5572 64. When Exim failed to open the panic log after failing to open the main log,
5573 the original message it was trying to log was written to stderr and debug
5574 output, but if they were not available (the usual case in production), it
5575 was lost. Now it is written to syslog before the two lines that record the
5576 failures to open the logs.
5578 65. Users' Exim filters run in subprocesses under the user's uid. It is
5579 possible for a "deliver" command or an alias in a "personal" command to
5580 provoke an address rewrite. If logging of address rewriting is configured,
5581 this fails because the process is not running as root or exim. There may be
5582 a better way of dealing with this, but for the moment (because 4.30 needs
5583 to be released), I have disabled address rewrite logging when running a
5584 filter in a non-root, non-exim process.
5590 1. The buildconfig auxiliary program wasn't quoting the value set for
5591 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5592 not defined. This bug was masked in 4.22 by the effect that was fixed in
5595 2. Some messages that were rejected after a message id was allocated were
5596 shown as "incomplete" by exigrep. It no longer does this for messages that
5597 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5599 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5600 have allow_domain_literals set, the ID did not get logged in the <= line.
5601 Domain literals are now always recognized in Message-ID: header lines.
5603 4. The first argument for a ${extract expansion item is the key name or field
5604 number. Leading and trailing spaces in this item were not being ignored,
5605 causing some misleading effects.
5607 5. When deliver_drop_privilege was set, single queue runner processes started
5608 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5609 same command in the process it spins off) were not dropping privilege.
5611 6. When the daemon running as "exim" started a queue runner, it always
5612 re-executed Exim in the spun-off process. This is a waste of effort when
5613 deliver_drop_privilege is set. The new process now just calls the
5614 queue-runner function directly.
5620 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5623 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5624 ignored. Though the use of -oP was forcing the writing of a pid file, it
5625 was always written to the default place.
5627 3. If the message "no IP address found for host xxxx" is generated during
5628 incoming verification, it is now followed by identification of the incoming
5629 connection (so you can more easily find what provoked it).
5631 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5633 5. Added some features to "harden" Exim a bit more against certain attacks:
5635 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5636 be put in Local/Makefile. This is like the never_users runtime option,
5637 but it cannot be overridden. The default setting is "root".
5639 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5640 prefix string with which any file named in a -C command line option
5643 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5644 is retained for -C and -D only if the caller of Exim is root. Without
5645 it, the exim user may also use -C and -D and retain privilege.
5647 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5648 command line option is disabled.
5650 6. Macro names set by the -D option must start with an upper case letter, just
5651 like macro names defined in the configuration file.
5653 7. Added "dereference=" facility to LDAP.
5655 8. Two instances of the typo "uknown" in the source files are fixed.
5657 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5658 the Configure-Makefile script screwed up while processing it.
5660 10. Incorporated PCRE 4.4.
5662 11. The SMTP synchronization check was not operating right at the start of an
5663 SMTP session. For example, it could not catch a HELO sent before the client
5664 waited for the greeting. There is now a check for outstanding input at the
5665 point when the greeting is written. Because of the duplex, asynchronous
5666 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5667 way, but not yet received, when the check is performed.
5669 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5670 on TCP/IP sockets, because this apparently causes some broken clients to
5673 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5674 unchanged) from the Cygwin maintainer.
5676 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5677 was supported instead of testing for mbx. Effectively a typo.
5679 15. The spa authenticator server code was not checking that the input it
5680 received was valid base64.
5682 16. The debug output line for the "set" modifier in ACLs was not showing the
5683 name of the variable that was being set.
5685 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5687 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5688 lookup on its own. Something else was needed to trigger the lookup. For
5689 example, a match in host_lookup or the need for a host name in a host list.
5690 Now, if $sender_host_name is referenced and the host name has not yet been
5691 looked up, a lookup is performed. If the lookup fails, the variable remains
5692 empty, and $host_lookup_failed is set to "1".
5694 19. Added "eqi" as a case-independent comparison operator.
5696 20. The saslauthd authentication condition could segfault if neither service
5697 nor realm was specified.
5699 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5700 error message that was logged was misleading, and incoming SMTP
5701 connections were dropped. The message is now more accurate, and temporary
5702 errors are given to SMTP connections.
5704 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5705 (except RSET) with a 421 error, until QUIT is received. However, it was
5706 failing to send a response to QUIT.
5708 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5709 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5710 ACL is now run after the helo_try_verify_hosts code.
5712 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5713 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5714 case-independent, so other case variants are also recognized). Apparently
5715 some systems use these upper case variants.
5717 25. If more than two messages were waiting for the same host, and a transport
5718 filter was specified for the transport, Exim sent two messages over the
5719 same TCP/IP connection, and then failed with "socket operation on non-
5720 socket" when it tried to send the third.
5722 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5724 27. The extern definition of crypt16() in expand.c was not being excluded when
5725 the OS had its own crypt16() function.
5727 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5728 as a preferred synonym for return_size_limit, both as an option and as an
5731 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5733 30. Changed the default configuration ACL to relax the local part checking rule
5734 for addresses that are not in any local domains. For these addresses,
5735 slashes and pipe symbols are allowed within local parts, but the sequence
5736 /../ is explicitly forbidden.
5738 31. SPA server authentication was not clearing the challenge buffer before
5741 32. log_message in a "warn" ACL statement was writing to the reject log as
5742 well as to the main log, which contradicts the documentation and doesn't
5743 seem right (because no rejection is happening). So I have stopped it.
5745 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5746 However, I am unable to do any testing of this.
5748 34. Fixed an infelicity in the appendfile transport. When checking directories
5749 for a mailbox, to see if any needed to be created, it was accidentally
5750 using path names with one or more superfluous leading slashes; tracing
5751 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5753 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5754 contents are added to the log line that is written for every discarded
5755 recipient. (Previously a log_message setting was ignored.)
5757 36. The ${quote: operator now quotes the string if it is empty.
5759 37. The install script runs exim in order to find its version number. If for
5760 some reason other than non-existence or emptiness, which it checks, it
5761 could not run './exim', it was installing it with an empty version number,
5762 i.e. as "exim-". This error state is now caught, and the installation is
5765 38. An argument was missing from the function that creates an error message
5766 when Exim fails to connect to the socket for saslauthd authentication.
5767 This could cause Exim to crash, or give a corrupted message.
5769 39. Added isip, isip4, and isip6 to ${if conditions.
5771 40. The ACL variables $acl_xx are now saved with the message, and can be
5772 accessed later in routers, transports, and filters.
5774 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5775 strings in the file are not string-expanded.
5777 42. If a MAIL command specified a SIZE value that was too large to fit into an
5778 int variable, the check against message_size_limit failed. Such values are
5779 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5780 day this will have to be increased, but I don't think I want to be around
5781 when emails are that large.
5788 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5789 iconv() is not standard in FreeBSD.
5791 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5792 IPv6 enabled. The observed symptom was a segmentation fault on return from
5793 the function os_common_find_running_interfaces() in src/os.c.
5795 3. In the check_special_case() function in daemon.c I had used "errno" as an
5796 argument name, which causes warnings on some systems. This was basically a
5797 typo, since it was named "eno" in the comments!
5799 4. The code that waits for the clock to tick (at a resolution of some fraction
5800 of a second) so as to ensure message-id uniqueness was always waiting for
5801 at least one whole tick, when it could have waited for less. [This is
5802 almost certainly not relevant at current processor speeds, where it is
5803 unlikely to ever wait at all. But we try to future-proof.]
5805 5. The function that sleeps for a time interval that includes fractions of a
5806 second contained a race. It did not block SIGALRM between setting the
5807 timer, and suspending (a couple of lines later). If the interval was short
5808 and the sigsuspend() was delayed until after it had expired, the suspension
5809 never ended. On busy systems this could lead to processes getting stuck for
5812 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5813 process, before it forks any delivery processes. The open lookup caching
5814 mechanism meant that the open file or database connection was passed into
5815 the delivery process. The problem was that delivery processes always tidy
5816 up cached lookup data. This could cause a problem for the next delivery
5817 process started by the queue runner, because the external queue runner
5818 process does not know about the closure. So the next delivery process
5819 still has data in the lookup cache. In the case of a file lookup, there was
5820 no problem because closing a file descriptor in a subprocess doesn't affect
5821 the parent. However, if the lookup was caching a connection to a database,
5822 the connection was closed, and the second delivery process was likely to
5823 see errors such as "PGSQL: query failed: server closed the connection
5824 unexpectedly". The problem has been fixed by closing all cached lookups
5825 in a queue runner before running a delivery process.
5827 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5828 seem to have the "const" qualifier which it has on other OS. I've
5831 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5832 *of the same type* (client or server) with the same public name that an
5833 error should be diagnosed.
5835 9. When Exim looked up a host name for an IP address, but failed to find the
5836 original IP address when looking up the host name (a safety check), it
5837 output the message "<ip address> does not match any IP for NULL", which was
5838 confusing, to say the least. The bug was that the host name should have
5839 appeared instead of "NULL".
5841 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5842 user that is built into the binary, and the -C or -D options is used, root
5843 privilege is dropped before the configuration file is read. In addition,
5844 logging is switched to stderr instead of the normal log files. If the
5845 configuration then re-defines the Exim user, the unprivileged environment
5846 is probably not what is expected, so Exim logs a panic warning message (but
5849 However, if deliver_drop_privilege is set, the unprivileged state may well
5850 be exactly what is intended, so the warning has been cut out in that case,
5851 and Exim is allowed to try to write to its normal log files.
5857 1. smtp_return_error_details was not giving details for temporary sender
5858 or receiver verification errors.
5860 2. Diagnose a configuration error if two authenticators have the same public
5863 3. Exim used not to create the message log file for a message until the first
5864 delivery attempt. This could be confusing when incoming messages were held
5865 for policy or load reasons. The message log file is now created at the time
5866 the message is received, and an initial "Received" line is written to it.
5868 4. The automatically generated man page for command line options had a minor
5869 bug that caused no ill effects; however, a more serious problem was that
5870 the procedure for building the man page automatically didn't always
5871 operate. Consequently, release 4.20 contains an out-of-date version. This
5872 shouldn't happen again.
5874 5. When building Exim with embedded Perl support, the script that builds the
5875 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5876 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5878 6. The freeze_tell option was not being used for messages that were frozen on
5879 arrival, either by an ACL or by local_scan().
5881 7. Added the smtp_incomplete_transaction log selector.
5883 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5884 was accepting AUTH without a new EHLO.
5886 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5887 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5890 10. Logging of TCP/IP connections (when configured) now happens in the main
5891 daemon process instead of the child process, so that the TCP/IP connection
5892 count is more accurate (but it can never be perfect).
5894 11. The use of "drop" in a nested ACL was not being handled correctly in the
5895 outer ACL. Now, if condition failure induced by the nested "drop" causes
5896 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5897 or "require"), the connection is dropped.
5899 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5900 that yield "discard" can now be used with an "accept" or a "discard" verb,
5901 but an error is generated for any others (because I can't see a useful way
5902 to define what should happen).
5904 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5905 are now processed in the same way as lines in the Exim configuration file.
5906 In particular, continuation lines are supported.
5908 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5910 15. Added -ti meaning -t -i.
5912 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5913 domains, and warn by logging if others are found.
5915 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5916 returned from GIFCONF when Exim is trying to find the list of interfaces on
5917 a host. The code in os.c has been modified to copy each ifreq to an aligned
5918 structure in all cases.
5920 Also, in some cases, the returned ifreq's were being copied to a 'struct
5921 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5922 means the last couple of bytes of an IPv6 address could be chopped if the
5923 ifreq contained only a normal sockaddr (14 bytes storage).
5925 18. Named domain lists were not supported in the hosts_treat_as_local option.
5926 An entry such as +xxxx was not recognized, and was treated as a literal
5929 19. Ensure that header lines added by a DATA ACL are included in the reject log
5930 if the ACL subsequently rejects the message.
5932 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5933 MD5 (which is deprecated).
5935 21. When testing a filter file using -bf, Exim was writing a message when it
5936 took the sender from a "From " line in the message, but it was not doing so
5937 when it took $return_path from a Return-Path: header line. It now does.
5939 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5940 with a valid header line field name (a series of printing characters
5941 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5943 23. Changed "disc" in the source to "disk" to conform to the documentation and
5944 the book and for uniformity.
5946 24. Ignore Sendmail's -Ooption=value command line item.
5948 25. When execve() failed while trying to run a command in a pipe transport,
5949 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5950 could be confused with a return value of 69 from the command itself. This
5951 has been changed to 127, the value the shell returns if it is asked to run
5952 a non-existent command. The wording for the related log line suggests a
5953 non-existent command as the problem.
5955 26. If received_header_text expands to an empty string, do not add a Received:
5956 header line to the message. (Well, it adds a token one on the spool, but
5957 marks it "old" so that it doesn't get used or transmitted.)
5959 27. Installed eximstats 1.28 (addition of -nt option).
5961 28. There was no check for failure on the call to getsockname() in the daemon
5962 code. This can fail if there is a shortage of resources on the system, with
5963 ENOMEM, for example. A temporary error is now given on failure.
5965 29. Contrary to the C standard, it seems that in some environments, the
5966 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5967 program. Exim now does this explicitly; it affects the formatting of
5968 timestamps using strftime().
5970 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5971 complaints. I've now initialized all the variables, to avoid this.
5973 32. Header lines added by a system filter were not being "seen" during
5974 transport-time rewrites.
5976 33. The info_callback() function passed to OpenSSL is set up with type void
5977 (*)(SSL *, int, int), as described somewhere. However, when calling the
5978 function (actually a macro) that sets it up, the type void(*)() is
5979 expected. I've put in a cast to prevent warnings from picky compilers.
5981 34. If a DNS black list lookup found a CNAME record, but there were no A
5982 records associated with the domain it pointed at, Exim crashed.
5984 35. If a DNS black list lookup returned more than one A record, Exim ignored
5985 all but the first. It now scans all returned addresses if a particular IP
5986 value is being sought. In this situation, the contents of the
5987 $dnslist_value variable are a list of all the addresses, separated by a
5990 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5991 to accept a host name and all its aliases if the forward lookup for any of
5992 them yielded the IP address of the incoming connection. Now it accepts only
5993 those names whose forward lookup yields the correct IP address. Any other
5994 names are discarded. This closes a loophole whereby a rogue DNS
5995 administrator could create reverse DNS records to break through a
5996 wildcarded host restriction in an ACL.
5998 37. If a user filter or a system filter that ran in a subprocess used any of
5999 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6000 the wrong values were passed to the pipe command ($thisaddress had the
6001 value of $0, $0 had the value of $1, etc). This bug was introduced by
6002 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6004 38. Improved the line breaking for long SMTP error messages from ACLs.
6005 Previously, if there was no break point between 40 and 75 characters, Exim
6006 left the rest of the message alone. Two changes have been made: (a) I've
6007 reduced the minimum length to 35 characters; (b) if it can't find a break
6008 point between 35 and 75 characters, it looks ahead and uses the first one
6009 that it finds. This may give the occasional overlong line, but at least the
6010 remaining text gets split now.
6012 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6013 file descriptors might be low, and that setting 1000 would always raise it.
6014 It turns out that in some environments, the limit is already over 1000 and
6015 that lowering it causes trouble. So now Exim takes care not to decrease it.
6017 40. When delivering a message, the value of $return_path is set to $sender_
6018 address at the start of routing (routers may change the value). By an
6019 oversight, this default was not being set up when an address was tested by
6020 -bt or -bv, which affected the outcome if any router or filter referred to
6023 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6024 log only when "message" or "log_message" are set. However, if one of the
6025 conditions was an address verification, or a call to a nested ACL, the
6026 messages generated by the underlying test were being passed through. This
6027 no longer happens. The underlying message is available in $acl_verify_
6028 message for both "message" and "log_message" expansions, so it can be
6029 passed through if needed.
6031 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6032 new expansion $bh_ to give the encoded byte string without charset
6033 translation. Translation happens only if iconv() is available; HAVE_ICONV
6034 indicates this at build time. HEADERS_CHARSET gives the charset to
6035 translate to; headers_charset can change it in the configuration, and
6036 "headers charset" can change it in an individual filter file.
6038 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6039 that creates RFC 2047 encoded "words" labels them as that charset instead
6040 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6041 expansion operator; (ii) when Exim creates a From: line for a local
6042 message; (iii) when a header line is rewritten to include a "phrase" part.
6044 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6045 buggy, causing it to skip the first lines of messages whose message ID
6046 ended in 'D'. This would not have bitten before Exim release 4.14, because
6047 message IDs were unlikely to end in 'D' before then. The effect was to have
6048 incorrect size information for certain domains.
6050 45. #include "config.h" was missing at the start of the crypt16.c module. This
6051 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6054 46. If there was a timeout during a "random" callout check, Exim treated it as
6055 a failure of the random address, and carried on sending RSET and the real
6056 address. If the delay was just some slowness somewhere, the response to the
6057 original RCPT would be taken as a response to RSET and so on, causing
6058 mayhem of various kinds.
6060 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6061 when I implemented it. It didn't allow for the fact that some option values
6062 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6063 the right test for positive values.
6065 48. Domain names in DNS records are case-independent. Exim always looks them up
6066 in lower case. Some resolvers return domain names in exactly the case they
6067 appear in the zone file, that is, they may contain uppercase letters. Not
6068 all resolvers do this - some return always lower case. Exim was treating a
6069 change of case by a resolver as a change of domain, similar to a widening
6070 of a domain abbreviation. This triggered its re-routing code and so it was
6071 trying to route what was effectively the same domain again. This normally
6072 caused routing to fail (because the router wouldn't handle the domain
6073 twice). Now Exim checks for this case specially, and just changes the
6074 casing of the domain that it ultimately uses when it transmits the message
6077 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6080 50. If a filter generated a file delivery with a non-absolute name (possible if
6081 no home directory exists for the router), the forbid_file option was not
6084 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6085 the existing equality matching.
6087 52. Exim was using ints instead of ino_t variables in some places where it was
6088 dealing with inode numbers.
6090 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6091 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6092 called TMPDIR, and if it finds it is different, it changes its value.
6094 54. The smtp_printf() function is now made available to local_scan() so
6095 additional output lines can be written before returning. There is also an
6096 smtp_fflush() function to enable the detection of a dropped connection.
6097 The variables smtp_input and smtp_batched_input are exported to
6100 55. Changed the default runtime configuration: the message "Unknown user"
6101 has been removed from the ACL, and instead placed on the localuser router,
6102 using the cannot_route_message feature. This means that any verification
6103 failures that generate their own messages won't get overridden. Similarly,
6104 the "Unrouteable address" message that was in the ACL for unverifiable
6105 relay addresses has also been removed.
6107 56. Added hosts_avoid_esmtp to the smtp transport.
6109 57. The exicyclog script was not checking for the esoteric option
6110 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6111 will work only if exicyclog is run under the appropriate euid.
6113 58. Following a discussion on the list, the rules by which Exim recognises line
6114 endings on incoming messages have been changed. The -dropcr and drop_cr
6115 options are now no-ops, retained only for backwards compatibility. The
6116 following line terminators are recognized: LF CRLF CR. However, special
6117 processing applies to CR:
6119 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6120 nor a local message in the state where . is a terminator.
6122 (ii) If a bare CR is encountered in a header line, an extra space is added
6123 after the line terminator so as not to end the header. The reasoning
6124 behind this is that bare CRs in header lines are most likely either
6125 to be mistakes, or people trying to play silly games.
6127 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6128 was being incorrectly given as 18 bytes larger than it should have been.
6129 This is a VOB (very old bug).
6131 60. This may never have affected anything current, but just in case it has:
6132 When the local host is found other than at the start of a list of hosts,
6133 the local host, those with the same MX, and any that follow, are discarded.
6134 When the list in question was part of a longer list of hosts, the following
6135 hosts (not currently being processed) were also being discarded. This no
6136 longer happens. I'm not sure if this situation could ever has previously
6139 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6142 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6143 header. This is something that is recommended in a new Internet Draft, and
6144 is something that is documented as being done by Sendmail. There are two
6145 possible values. For messages generated by the autoreply transport, Exim
6148 Auto-Submitted: auto-replied
6150 whereas for all other generated messages (e.g. bounces) it adds
6152 Auto-Submitted: auto-generated
6154 63. The "personal" condition in filters now includes a test for the
6155 Auto-Submitted: header. If it contains the string "auto-" the message it
6156 not considered personal.
6158 64. Added rcpt_include_affixes as a generic transport option.
6160 65. Added queue_only_override (default true).
6162 66. Added the syslog_duplication option.
6164 67. If what should have been the first header line of a message consisted of
6165 a space followed by a colon, Exim was mis-interpreting it as a header line.
6166 It isn't of course - it is syntactically invalid and should therefore be
6167 treated as the start of the message body. The misbehaviour could have
6168 caused a number of strange effects, including loss of data in subsequent
6169 header lines, and spool format errors.
6171 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6172 client host had authenticated. This control can now be exercised by an ACL
6173 for more flexibility.
6175 69. By default, callouts do not happen when testing with -bh. There is now a
6176 variant, -bhc, which does actually run the callout code, including
6177 consulting and updating the callout cache.
6179 70. Added support for saslauthd authentication, courtesy of Alexander
6182 71. If statvfs() failed on the spool or log directories while checking their
6183 size for availability, Exim confusingly gave the error "space shortage".
6184 Furthermore, in debugging mode it crashed with a floating point exception.
6185 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6186 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6187 serious problem, Exim now writes to the main and panic logs when this
6188 happens, with details of the failure. It then refuses to accept the
6189 incoming message, giving the message "spool directory problem" or "log
6190 directory problem" with a 421 code for SMTP messages.
6192 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6193 0, 1, and 2 exist, because some OS complain for execs without them (see
6194 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6195 descriptors. However, the code omitted to check that the open succeeded,
6196 causing mysterious errors if for some reason the permissions on /dev/null
6197 got screwed. Now Exim writes a message to the main and panic logs, and
6198 bombs out if it can't open /dev/null.
6200 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6201 interact so that it is all more flexible. It is supposed to remain
6202 backwards compatible. Also added extra_local_interfaces.
6204 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6205 to bomb out with an assertion failure - to the client this appears as a
6206 connection drop. This problem occurs in the part of the code that was taken
6207 from the Samba project. Fortunately, the assertion is in a very simple
6208 function, so I have fixed this by reproducing the function inline in the
6209 one place where it is called, and arranging for authentication to fail
6210 instead of killing the process with assert().
6212 75. The SPA client code was not working when the server requested OEM rather
6213 than Unicode encoding.
6215 76. Added code to make require_files with a specific uid setting more usable in
6216 the case where statting the file as root fails - usually a non-root-mounted
6217 NFS file system. When this happens and the failure is EACCES, Exim now
6218 forks a subprocess and does the per-uid checking as the relevant uid.
6220 77. Added process_log_path.
6222 78. If log_file_path was not explicitly set, a setting of check_log_space or
6223 check_log_inodes was ignored.
6225 79. If a space check for the spool or log partitions fails, the incident is now
6226 logged. Of course, in the latter case the data may get lost...
6228 80. Added the %p formatting code to string_format() so that it can be used to
6229 print addresses in debug_print(). Adjusted all the address printing in the
6230 debugging in store.c to use %p rather than %d.
6232 81. There was a concern that a line of code in smtp_in.c could overflow a
6233 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6234 initially expressed, the concern was not well-founded, because trailing
6235 spaces are removed early. However, if the trailing spaces were followed by
6236 a NULL, they did not get removed, so the overflow was possible. Two fixes
6239 (a) I re-wrote the offending code in a cleaner fashion.
6240 (b) If an incoming SMTP command contains a NULL character, it is rejected
6243 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6244 runs initgroups(), so that if the Exim user is in any additional groups,
6245 they will be used during message reception.
6251 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob