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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
71 TL/01 When running the test suite, the README says that variables such as
72 no_msglog_check are global and can be placed anywhere in a specific
73 test's script, however it was observed that placement needed to be near
74 the beginning for it to behave that way. Changed the runtest perl
75 script to read through the entire script once to detect and set these
76 variables, reset to the beginning of the script, and then run through
77 the script parsing/test process like normal.
79 TL/02 The BSD's have an arc4random API. One of the functions to induce
80 adding randomness was arc4random_stir(), but it has been removed in
81 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
82 function when detected.
84 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
85 cause callback expansion.
87 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
88 syntax errors in an expansion can be treated as a string instead of
89 logging or causing an error, due to the internal use of bool_lax
90 instead of bool when processing it.
92 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
93 server certificates when making smtp deliveries.
95 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
97 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
99 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
101 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
102 Merged patch from Sebastian Wiedenroth.
104 JH/05 Fix results-pipe from transport process. Several recipients, combined
105 with certificate use, exposed issues where response data items split
106 over buffer boundaries were not parsed properly. This eventually
107 resulted in duplicates being sent. This issue only became common enough
108 to notice due to the introduction of conection certificate information,
109 the item size being so much larger. Found and fixed by Wolfgang Breyha.
111 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
112 size buffer was used, resulting in syntax errors when an expansion
115 JH/07 Add support for directories of certificates when compiled with a GnuTLS
116 version 3.3.6 or later.
118 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
119 is EXPERIMENTAL_EVENT, the main-configuration and transport options
120 both become "event_action", the variables become $event_name, $event_data
121 and $event_defer_errno. There is a new variable $verify_mode, usable in
122 routers, transports and related events. The tls:cert event is now also
123 raised for inbound connections, if the main configuration event_action
126 TL/06 In test suite, disable OCSP for old versions of openssl which contained
127 early OCSP support, but no stapling (appears to be less than 1.0.0).
129 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
130 server certificate names available under the smtp transport option
131 "tls_verify_cert_hostname" now do not permit multi-component wildcard
134 JH/10 Time-related extraction expansions from certificates now use the main
135 option "timezone" setting for output formatting, and are consistent
136 between OpenSSL and GnuTLS compilations. Bug 1541.
138 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
139 encoded parameter in the incoming message. Bug 1558.
141 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
142 include certificate info, eximon was claiming there were spoolfile
145 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
147 JH/14 Log delivery-related information more consistently, using the sequence
148 "H=<name> [<ip>]" wherever possible.
150 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
151 are problematic for Debian distribution, omit them from the release
154 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
156 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
159 JH/17 Fix a null-indirection in certextract expansions when a nondefault
160 output list separator was used.
165 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
166 checkers that were complaining about end of non-void function with no
169 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
170 This was a regression intruduced in 4.83 by another bugfix.
172 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
174 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
175 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
181 TF/01 Correctly close the server side of TLS when forking for delivery.
183 When a message was received over SMTP with TLS, Exim failed to clear up
184 the incoming connection properly after forking off the child process to
185 deliver the message. In some situations the subsequent outgoing
186 delivery connection happened to have the same fd number as the incoming
187 connection previously had. Exim would try to use TLS and fail, logging
188 a "Bad file descriptor" error.
190 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
191 utilities have not been installed.
193 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
194 temporary space as the ACL may create new global variables.
196 TL/01 LDAP support uses per connection or global context settings, depending
197 upon the detected version of the libraries at build time.
199 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
200 to extract and use the src ip:port in logging and expansions as if it
201 were a direct connection from the outside internet. PPv2 support was
202 updated based on HAProxy spec change in May 2014.
204 JH/02 Add ${listextract {number}{list}{success}{fail}}.
206 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
207 Properly escape header and check for NULL return.
209 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
212 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
214 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
215 characters in header names, implemented as a verify condition.
216 Contributed by Michael Fischer v. Mollard.
218 TL/05 Rename SPF condition results err_perm and err_temp to standardized
219 results permerror and temperror. Previous values are deprecated but
220 still accepted. In a future release, err_perm and err_temp will be
221 completely removed, which will be a backward incompatibility if the
222 ACL tests for either of these two old results. Patch contributed by
223 user bes-internal on the mailing list.
225 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
227 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
228 selectors, in both main and reject logs.
230 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
233 JH/07 Add malware type "sock" for talking to simple daemon.
235 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
237 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
238 routers/transports under cutthrough routing.
240 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
241 numbers. Touch up "bool" conditional to keep the same definition.
243 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
245 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
247 JH/12 Expand items in router/transport headers_add or headers_remove lists
248 individually rather than the list as a whole. Bug 1452.
250 Required for reasonable handling of multiple headers_ options when
251 they may be empty; requires that headers_remove items with embedded
252 colons must have them doubled (or the list-separator changed).
254 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
255 view the policy declared in the DMARC record. Currently, $dmarc_status
256 is a combined value of both the record presence and the result of the
259 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
261 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
262 dnslookup router and the smtp transport (applying to the forward
265 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
266 of ldap servers used for a specific lookup. Patch provided by Heiko
269 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
270 New variable $lookup_dnssec_authenticated for observability.
272 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
273 Patch submitted by Lars Timman.
275 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
277 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
278 Requires trusted mode and valid format message id, aborts otherwise.
279 Patch contributed by Heiko Schlichting.
281 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
282 certextract with support for various fields. Bug 1358.
284 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
285 is requested by default, modifiable by smtp transport option
288 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
289 operate on certificate variables to give certificate fingerprints
290 Also new ${sha256:cert_variable}.
292 JH/23 The PRDR feature is moved from being Experimental into the mainline.
294 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
295 Christian Aistleitner.
297 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
299 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
300 file. Patch from Wolfgang Breyha.
302 JH/25 Expand the coverage of the delivery $host and $host_address to
303 client authenticators run in verify callout. Bug 1476.
305 JH/26 Port service names are now accepted for tls_on_connect_ports, to
306 align with daemon_smtp_ports. Bug 72.
308 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
309 support and error reporting did not work properly.
311 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
312 and is readable. Patch from Andrew Colin Kissa.
314 TL/14 Enhance documentation of ${run expansion and how it parses the
315 commandline after expansion, particularly in the case when an
316 unquoted variable expansion results in an empty value.
318 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
320 PP/02 Fix internal collision of T_APL on systems which support RFC3123
321 by renaming away from it. Addresses GH issue 15, reported by
324 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
326 TL/15 SECURITY: prevent double expansion in math comparison functions
327 (can expand unsanitized data). Not remotely exploitable.
334 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
336 PP/02 Make -n do something, by making it not do something.
337 When combined with -bP, the name of an option is not output.
339 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
342 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
343 $sender_host_name and config options to manage this, and basic check
346 PP/05 DSCP support for outbound connections and control modifier for inbound.
348 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
349 (Only plugin which currently uses this is kerberos4, which nobody should
350 be using, but we should make it available and other future plugins might
351 conceivably use it, even though it would break NAT; stuff *should* be
352 using channel bindings instead).
354 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
355 name; added for Sendmail compatibility; requires admin caller.
356 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
357 just ignore it); requires trusted caller.
358 Also parse but ignore: -Ac -Am -X<logfile>
361 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
363 TL/02 Add +smtp_confirmation as a default logging option.
365 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
366 Patch by Magnus Holmgren from 2007-02-20.
368 TL/04 Bugzilla 1281 - Spec typo.
369 Bugzilla 1283 - Spec typo.
370 Bugzilla 1290 - Spec grammar fixes.
372 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
374 TL/06 Add Experimental DMARC support using libopendmarc libraries.
376 TL/07 Fix an out of order global option causing a segfault. Reported to dev
377 mailing list by by Dmitry Isaikin.
379 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
381 JH/02 Support "G" suffix to numbers in ${if comparisons.
383 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
385 NM/01 Bugzilla 1197 - Spec typo
386 Bugzilla 1196 - Spec examples corrections
388 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
390 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
391 gnutls_enable_pkcs11, but renamed to more accurately indicate its
394 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
395 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
397 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
398 "acl {{name}{arg}...}", and optional args on acl condition
401 JH/05 Permit multiple router/transport headers_add/remove lines.
403 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
405 JH/07 Avoid using a waiting database for a single-message-only transport.
406 Performance patch from Paul Fisher. Bugzilla 1262.
408 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
411 JH/09 Add $headers_added variable, with content from use of ACL modifier
412 add_header (but not yet added to the message). Bugzilla 199.
414 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
415 Pulled from Bugzilla 817 by Wolfgang Breyha.
417 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
419 (nb: this is the same fix as in Exim 4.80.1)
421 JH/11 Add A= logging on delivery lines, and a client_set_id option on
424 JH/12 Add optional authenticated_sender logging to A= and a log_selector
427 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
429 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
430 advertise SMTP AUTH mechanism to us, instead of a generic
431 protocol violation error. Also, make Exim more robust to bad
432 data from the Dovecot auth socket.
434 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
436 When a queue runner is handling a message, Exim first routes the
437 recipient addresses, during which it prunes them based on the retry
438 hints database. After that it attempts to deliver the message to
439 any remaining recipients. It then updates the hints database using
442 So if a recipient address works intermittently, it can get repeatedly
443 deferred at routing time. The retry hints record remains fresh so the
444 address never reaches the final cutoff time.
446 This is a fairly common occurrence when a user is bumping up against
447 their storage quota. Exim had some logic in its local delivery code
448 to deal with this. However it did not apply to per-recipient defers
449 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
451 This change adds a proper retry rule check during routing so that the
452 final cutoff time is checked against the message's age. We only do
453 this check if there is an address retry record and there is not a
454 domain retry record; this implies that previous attempts to handle
455 the address had the retry_use_local_parts option turned on. We use
456 this as an approximation for the destination being like a local
457 delivery, as in LMTP.
459 I suspect this new check makes the old local delivery cutoff check
460 redundant, but I have not verified this so I left the code in place.
462 TF/02 Correct gecos expansion when From: is a prefix of the username.
464 Test 0254 submits a message to Exim with the header
468 When I ran the test suite under the user fanf2, Exim expanded
469 the header to contain my full name, whereas it should have added
470 a Resent-Sender: header. It erroneously treats any prefix of the
471 username as equal to the username.
473 This change corrects that bug.
475 GF/01 DCC debug and logging tidyup
476 Error conditions log to paniclog rather than rejectlog.
477 Debug lines prefixed by "DCC: " to remove any ambiguity.
479 TF/03 Avoid unnecessary rebuilds of lookup-related code.
481 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
482 Bug spotted by Jeremy Harris; was flawed since initial commit.
483 Would have resulted in OCSP responses post-SNI triggering an Exim
484 NULL dereference and crash.
486 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
488 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
489 Bug detection, analysis and fix by Samuel Thibault.
490 Bugzilla 1331, Debian bug #698092.
492 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
494 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
495 Server implementation by Todd Lyons, client by JH.
496 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
497 config variable "prdr_enable" controls whether the server
498 advertises the facility. If the client requests PRDR a new
499 acl_data_smtp_prdr ACL is called once for each recipient, after
500 the body content is received and before the acl_smtp_data ACL.
501 The client is controlled by bolth of: a hosts_try_prdr option
502 on the smtp transport, and the server advertisement.
503 Default client logging of deliveries and rejections involving
504 PRDR are flagged with the string "PRDR".
506 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
507 fclose(). Diagnosis by Todd Lyons.
509 PP/17 Update configure.default to handle IPv6 localhost better.
510 Patch by Alain Williams (plus minor tweaks).
513 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
514 This is now consistent with GnuTLS, and is now documented: the
515 previous undocumented portable approach to treating the option as
516 unset was to force an expansion failure. That still works, and
517 an empty string is now equivalent.
519 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
520 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
521 not performing validation itself.
523 PP/20 Added force_command boolean option to pipe transport.
524 Patch from Nick Koston, of cPanel Inc.
526 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
529 TF/04 Added udpsend ACL modifer and hexquote expansion operator
531 PP/21 Fix eximon continuous updating with timestamped log-files.
532 Broken in a format-string cleanup in 4.80, missed when I repaired the
533 other false fix of the same issue.
534 Report and fix from Heiko Schlichting.
537 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
538 Report from Prashanth Katuri.
540 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
541 It's SecureTransport, so affects any MacOS clients which use the
542 system-integrated TLS libraries, including email clients.
544 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
545 using a MIME ACL for non-SMTP local injection.
546 Report and assistance in diagnosis by Warren Baker.
548 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
550 JH/16 Fix comparisons for 64b. Bugzilla 1385.
552 TL/09 Add expansion variable $authenticated_fail_id to keep track of
553 last id that failed so it may be referenced in subsequent ACL's.
555 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
558 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
559 ldap library initialization, allowing self-signed CA's to be
560 used. Also properly sets require_cert option later in code by
561 using NULL (global ldap config) instead of ldap handle (per
562 session). Bug diagnosis and testing by alxgomz.
564 TL/12 Enhanced documentation in the ratelimit.pl script provided in
565 the src/util/ subdirectory.
567 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
568 renamed to Transport Post Delivery Action by Jeremy Harris, as
571 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
572 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
573 redis_servers = needs to be configured which will be used by the redis
574 lookup. Patch from Warren Baker, of The Packet Hub.
576 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
578 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
579 hostname or reverse DNS when processing a host list. Used suggestions
580 from multiple comments on this bug.
582 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
584 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
585 Missed a few lines, added it to make the runtest require no keyboard
588 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
589 contains upper case chars. Make router use caseful_local_part.
591 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
592 support when GnuTLS has been built with p11-kit.
598 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
600 This, or similar/improved, will also be change PP/11 of 4.82.
606 PP/01 Handle short writes when writing local log-files.
607 In practice, only affects FreeBSD (8 onwards).
608 Bugzilla 1053, with thanks to Dmitry Isaikin.
610 NM/01 Bugzilla 949 - Documentation tweak
612 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
615 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
617 PP/02 Implemented gsasl authenticator.
619 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
621 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
622 `pkg-config foo` for cflags/libs.
624 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
625 with rest of GSASL and with heimdal_gssapi.
627 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
628 `pkg-config foo` for cflags/libs for the TLS implementation.
630 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
631 properties get this fed in as external SSF. A number of robustness
632 and debugging improvements to the cyrus_sasl authenticator.
634 PP/08 cyrus_sasl server now expands the server_realm option.
636 PP/09 Bugzilla 1214 - Log authentication information in reject log.
637 Patch by Jeremy Harris.
639 PP/10 Added dbmjz lookup type.
641 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
643 PP/12 MAIL args handles TAB as well as SP, for better interop with
644 non-compliant senders.
645 Analysis and variant patch by Todd Lyons.
647 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
648 Bug report from Lars Müller <lars@samba.org> (via SUSE),
649 Patch from Dirk Mueller <dmueller@suse.com>
651 PP/13 tls_peerdn now print-escaped for spool files.
652 Observed some $tls_peerdn in wild which contained \n, which resulted
653 in spool file corruption.
655 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
656 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
657 or write after TLS renegotiation, which otherwise led to messages
660 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
661 as a tracking header (ie: a signed header comes before the signature).
662 Patch from Wolfgang Breyha.
664 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
665 comma-sep list; embedded commas doubled.
667 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
669 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
671 Report and patch from Dmitry Banschikov.
673 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
674 Removed SSL_clear() after SSL_new() which led to protocol negotiation
675 failures. We appear to now support TLS1.1+ with Exim.
677 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
678 lets Exim select keys and certificates based upon TLS SNI from client.
679 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
680 before an outbound SMTP session. New log_selector, +tls_sni.
682 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
683 NULL dereference. Report and patch from Alun Jones.
685 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
686 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
687 Not seeing resolver debug output on NetBSD, but suspect this is a
688 resolver implementation change.
690 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
691 Left warnings. Added "eximon gdb" invocation mode.
693 PP/21 Defaulting "accept_8bitmime" to true, not false.
695 PP/22 Added -bw for inetd wait mode support.
697 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
698 locate the relevant includes and libraries. Made this the default.
700 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
701 Bugzilla 1246, report and most of solution from Tomasz Kusy.
703 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
704 This may cause build issues on older platforms.
706 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
707 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
708 gnutls_require_mac & gnutls_require_protocols (no longer supported).
709 Added SNI support via GnuTLS too.
710 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
712 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
714 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
715 Applied second patch from Janne, implementing suggestion to default
716 multiple-strings-in-record handling to match SPF spec.
718 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
720 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
721 read-only, out of scope).
722 Patch from Wolfgang Breyha, report from Stuart Northfield.
724 PP/29 Fix three issues highlighted by clang analyser static analysis.
725 Only crash-plausible issue would require the Cambridge-specific
726 iplookup router and a misconfiguration.
727 Report from Marcin Mirosław.
729 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
731 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
732 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
733 As part of this, removing so much warning spew let me fix some minor
734 real issues in debug logging.
736 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
737 assignment on my part. Fixed.
739 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
740 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
741 Janne Snabb (who went above and beyond: thank you).
743 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
744 string otherwise requires a connection and a bunch more work and it's
745 relatively easy to get wrong. Should also expose TLS library linkage
748 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
749 64-bit ${eval} (JH/03).
751 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
752 GNU libc to support some of the 64-bit stuff, should not lead to
753 conflicts. Defined before os.h is pulled in, so if a given platform
754 needs to override this, it can.
756 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
757 protection layer was required, which is not implemented.
758 Bugzilla 1254, patch from Wolfgang Breyha.
760 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
761 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
762 tls_dhparam take prime identifiers. Also unbreak combination of
763 OpenSSL+DH_params+TLSSNI.
765 PP/39 Disable SSLv2 by default in OpenSSL support.
771 PP/01 Solaris build fix for Oracle's LDAP libraries.
772 Bugzilla 1109, patch from Stephen Usher.
774 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
776 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
779 TF/02 Fix a couple more cases where we did not log the error message
780 when unlink() failed. See also change 4.74-TF/03.
782 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
783 lock up or crash if it happened to be inside a call to libc when it
784 got a SIGUSR1 from exiwhat.
786 The SIGUSR1 handler appends the current process status to the process
787 log which is later printed by exiwhat. It used to use the general
788 purpose logging code to do this, but several functions it calls are
789 not safe for signals.
791 The new output code in the SIGUSR1 handler is specific to the process
792 log, and simple enough that it's easy to inspect for signal safety.
793 Removing some special cases also simplifies the general logging code.
794 Removing the spurious timestamps from the process log simplifies
797 TF/04 Improved ratelimit ACL condition.
799 The /noupdate option has been deprecated in favour of /readonly which
800 has clearer semantics. The /leaky, /strict, and /readonly update modes
801 are mutually exclusive. The update mode is no longer included in the
802 database key; it just determines when the database is updated. (This
803 means that when you upgrde Exim will forget old rate measurements.)
805 Exim now checks that the per_* options are used with an update mode that
806 makes sense for the current ACL. For example, when Exim is processing a
807 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
808 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
809 must specify per_mail/readonly. If you omit the update mode it defaults to
810 /leaky where that makes sense (as before) or /readonly where required.
812 The /noupdate option is now undocumented but still supported for
813 backwards compatibility. It is equivalent to /readonly except that in
814 ACLs where /readonly is required you may specify /leaky/noupdate or
815 /strict/noupdate which are treated the same as /readonly.
817 A useful new feature is the /count= option. This is a generalization
818 of the per_byte option, so that you can measure the throughput of other
819 aggregate values. For example, the per_byte option is now equivalent
820 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
822 The per_rcpt option has been generalized using the /count= mechanism
823 (though it's more complicated than the per_byte equivalence). When it is
824 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
825 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
826 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
827 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
828 Note that using per_rcpt with a non-readonly update mode in more than
829 one ACL will cause the recipients to be double-counted. (The per_mail
830 and per_byte options don't have this problem.)
832 The handling of very low rates has changed slightly. If the computed rate
833 is less than the event's count (usually one) then this event is the first
834 after a long gap. In this case the rate is set to the same as this event's
835 count, so that the first message of a spam run is counted properly.
837 The major new feature is a mechanism for counting the rate of unique
838 events. The new per_addr option counts the number of different
839 recipients that someone has sent messages to in the last time period. It
840 behaves like per_rcpt if all the recipient addresses are different, but
841 duplicate recipient addresses do not increase the measured rate. Like
842 the /count= option this is a general mechanism, so the per_addr option
843 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
844 example, measure the rate that a client uses different sender addresses
845 with the options per_mail/unique=$sender_address. There are further
846 details in the main documentation.
848 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
850 TF/06 Removed a few PCRE remnants.
852 TF/07 Automatically extract Exim's version number from tags in the git
853 repository when doing development or release builds.
855 PP/02 Raise smtp_cmd_buffer_size to 16kB.
856 Bugzilla 879. Patch from Paul Fisher.
858 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
859 Heavily based on revision 40f9a89a from Simon Arlott's tree.
862 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
864 PP/05 Variable $av_failed, true if the AV scanner deferred.
865 Bugzilla 1078. Patch from John Horne.
867 PP/06 Stop make process more reliably on build failure.
868 Bugzilla 1087. Patch from Heiko Schlittermann.
870 PP/07 Make maildir_use_size_file an _expandable_ boolean.
871 Bugzilla 1089. Patch from Heiko Schlittermann.
873 PP/08 Handle ${run} returning more data than OS pipe buffer size.
874 Bugzilla 1131. Patch from Holger Weiß.
876 PP/09 Handle IPv6 addresses with SPF.
877 Bugzilla 860. Patch from Wolfgang Breyha.
879 PP/10 GnuTLS: support TLS 1.2 & 1.1.
881 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
884 PP/11 match_* no longer expand right-hand-side by default.
885 New compile-time build option, EXPAND_LISTMATCH_RHS.
886 New expansion conditions, "inlist", "inlisti".
888 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
890 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
892 PP/14 fix log_write() format string regression from TF/03.
893 Bugzilla 1152. Patch from Dmitry Isaikin.
899 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
901 PP/02 Harmonised TLS library version reporting; only show if debugging.
902 Layout now matches that introduced for other libraries in 4.74 PP/03.
904 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
906 PP/04 New "dns_use_edns0" global option.
908 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
911 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
912 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
914 TK/01 Updated PolarSSL code to 0.14.2.
915 Bugzilla 1097. Patch from Andreas Metzler.
917 PP/07 Catch divide-by-zero in ${eval:...}.
920 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
923 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
924 format-string attack -- SECURITY: remote arbitrary code execution.
926 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
927 time unintentionally subject to list matching rules, letting the header
928 cause arbitrary Exim lookups (of items which can occur in lists, *not*
929 arbitrary string expansion). This allowed for information disclosure.
931 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
932 INT_MIN/-1 -- value coerced to INT_MAX.
938 NM/01 Workround for PCRE version dependency in version reporting
941 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
942 This fixes portability to compilers other than gcc, notably
943 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
945 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
946 makefiles for portability to HP-UX and POSIX correctness.
948 PP/01 Permit LOOKUP_foo enabling on the make command-line.
949 Also via indented variable definition in the Makefile.
950 (Debugging by Oliver Heesakkers).
952 PP/02 Restore caching of spamd results with expanded spamd_address.
953 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
955 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
956 Improves build reliability. Fix from: Frank Elsner
958 NM/02 Fix wide character breakage in the rfc2047 coding
959 Fixes bug 1064. Patch from Andrey N. Oktyabrski
961 NM/03 Allow underscore in dnslist lookups
962 Fixes bug 1026. Patch from Graeme Fowler
964 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
965 Code patches from Adam Ciarcinski of NetBSD.
967 NM/04 Fixed exiqgrep to cope with mailq missing size issue
970 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
971 is logged, to avoid truncation. Patch from John Horne.
973 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
974 Patch from Jakob Hirsch.
976 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
977 SQL string expansion failure details.
978 Patch from Andrey Oktyabrski.
980 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
981 Patch from Simon Arlott.
983 PP/09 New lookups functionality failed to compile on old gcc which rejects
984 extern declarations in function scope.
985 Patch from Oliver Fleischmann
987 PP/10 Use sig_atomic_t for flags set from signal handlers.
988 Check getgroups() return and improve debugging.
989 Fixed developed for diagnosis in bug 927 (which turned out to be
992 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
993 Patch from Mark Zealey.
995 PP/12 Bugzilla 1056: Improved spamd server selection.
996 Patch from Mark Zealey.
998 PP/13 Bugzilla 1086: Deal with maildir quota file races.
999 Based on patch from Heiko Schlittermann.
1001 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1002 Patch from Uwe Doering, sign-off by Michael Haardt.
1004 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1005 variable declaration deep within a block. Bug and patch from
1008 PP/15 lookups-Makefile IRIX compatibilty coercion.
1010 PP/16 Make DISABLE_DKIM build knob functional.
1012 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1013 Patch by Simon Arlott
1015 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1016 variable argument macros. Our copy now differs from upstream.
1022 TF/01 Failure to get a lock on a hints database can have serious
1023 consequences so log it to the panic log.
1025 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1026 controlled using the smtp_confirmation log selector.
1028 TF/03 Include the error message when we fail to unlink a spool file.
1030 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1031 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1032 for maintaining out-of-tree patches for some time.
1034 PP/01 Bugzilla 139: Documentation and portability issues.
1035 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1036 Handle per-OS dynamic-module compilation flags.
1038 PP/02 Let /dev/null have normal permissions.
1039 The 4.73 fixes were a little too stringent and complained about the
1040 permissions on /dev/null. Exempt it from some checks.
1041 Reported by Andreas M. Kirchwitz.
1043 PP/03 Report version information for many libraries, including
1044 Exim version information for dynamically loaded libraries. Created
1045 version.h, now support a version extension string for distributors
1046 who patch heavily. Dynamic module ABI change.
1048 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1049 privilege escalation vulnerability whereby the Exim run-time user
1050 can cause root to append content of the attacker's choosing to
1053 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1056 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1057 If dropping privileges for untrusted macros, we disabled normal logging
1058 on the basis that it would fail; for the Exim run-time user, this is not
1059 the case, and it resulted in successful deliveries going unlogged.
1060 Fixed. Reported by Andreas Metzler.
1066 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1067 only prepend for the Resent-* case. Fixes regression introduced in
1068 Exim 4.70 by NM/22 for Bugzilla 607.
1070 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1071 increasing numbers of administrators be bitten by this.
1073 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1075 PP/03 Bugzilla 994: added openssl_options main configuration option.
1077 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1079 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1081 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1083 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1084 without a peer certificate, leading to a segfault because of an
1085 assumption that peers always have certificates. Be a little more
1086 paranoid. Problem reported by Martin Tscholak.
1088 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1089 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1090 NB: ClamAV planning to remove STREAM in "middle of 2010".
1091 CL also introduces -bmalware, various -d+acl logging additions and
1092 more caution in buffer sizes.
1094 PP/09 Implemented reverse_ip expansion operator.
1096 PP/10 Bugzilla 937: provide a "debug" ACL control.
1098 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1100 PP/12 Bugzilla 973: Implement --version.
1102 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1104 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1106 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1108 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1109 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1110 ignore trailing whitespace.
1112 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1114 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1117 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1118 Notification from Dr Andrew Aitchison.
1120 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1121 ExtendedDetectionInfo response format.
1122 Notification from John Horne.
1124 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1127 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1128 XSL and documented dependency on system catalogs, with examples of how
1131 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1134 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1135 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1136 configuration file which is writeable by the Exim user or group.
1138 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1139 of configuration files to cover files specified with the -C option if
1140 they are going to be used with root privileges, not just the default
1143 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1144 option (effectively making it always true).
1146 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1147 files to be used while preserving root privileges.
1149 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1150 that rogue child processes cannot use them.
1152 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1153 run-time user, instead of root.
1155 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1156 Exim run-time user without dropping privileges.
1158 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1159 result string, instead of calling string_vformat() twice with the same
1162 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1163 for other users. Others should always drop root privileges if they use
1164 -C on the command line, even for a whitelisted configure file.
1166 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1168 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1174 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1175 $data_path, and $header_path variables; fixed documentation bugs and
1178 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1179 exipick to access non-standard spools, including the "frozen" queue
1182 NM/01 Bugzilla 965: Support mysql stored procedures.
1183 Patch from Alain Williams
1185 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1187 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1188 Patch from Andreas Metzler
1190 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1191 Patch from Kirill Miazine
1193 NM/05 Bugzilla 671: Added umask to procmail example.
1195 JJ/03 installed exipick 20100323.0, fixing doc bug
1197 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1198 directory. Notification and patch from Dan Rosenberg.
1200 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1202 TK/02 Improve log output when DKIM signing operation fails.
1204 MH/01 Treat the transport option dkim_domain as a colon separated
1205 list, not as a single string, and sign the message with each element,
1206 omitting multiple occurences of the same signer.
1208 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1209 Bugzilla 985, 986. Patch by Simon Arlott
1211 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1212 Patch by Simon Arlott
1214 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1215 MBX locking. Notification from Dan Rosenberg.
1221 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1223 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1225 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1227 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1229 NM/04 Bugzilla 915: Flag broken perl installation during build.
1235 TK/01 Added patch by Johannes Berg that expands the main option
1236 "spamd_address" if it starts with a dollar sign.
1238 TK/02 Write list of recipients to X-Envelope-Sender header when building
1239 the mbox-format spool file for content scanning (suggested by Jakob
1242 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1243 (http://www.dcc-servers.net/) support via dccifd. Activated by
1244 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1246 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1247 by Mark Daniel Reidel <mr@df.eu>.
1249 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1250 When building exim an external PCRE library is now needed -
1251 PCRE is a system library on the majority of modern systems.
1252 See entry on PCRE_LIBS in EDITME file.
1254 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1255 conversation. Added nologin parameter to request.
1256 Patch contributed by Kirill Miazine.
1258 TF/01 Do not log submission mode rewrites if they do not change the address.
1260 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1262 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1263 log files in place. Contributed by Roberto Lima.
1265 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1267 TF/03 Bugzilla 615: When checking the local_parts router precondition
1268 after a local_part_suffix or local_part_prefix option, Exim now
1269 does not use the address's named list lookup cache, since this
1270 contains cached lookups for the whole local part.
1272 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1273 Robert Millan. Documentation is in experimental-spec.txt.
1275 TF/04 Bugzilla 668: Fix parallel build (make -j).
1277 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1279 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1280 Patch provided by Jan Srzednicki.
1282 TF/05 Leading white space used to be stripped from $spam_report which
1283 wrecked the formatting. Now it is preserved.
1285 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1286 that they are available at delivery time.
1288 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1290 TF/08 TLS error reporting now respects the incoming_interface and
1291 incoming_port log selectors.
1293 TF/09 Produce a more useful error message if an SMTP transport's hosts
1294 setting expands to an empty string.
1296 NM/06 Bugzilla 744: EXPN did not work under TLS.
1297 Patch provided by Phil Pennock.
1299 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1300 Patch provided by Richard Godbee.
1302 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1303 acl_smtp_notquit, added index entry.
1305 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1306 Patch provided by Eugene Bujak.
1308 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1309 accept(). Patch provided by Maxim Dounin.
1311 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1312 Patch provided by Phil Pennock.
1314 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1316 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1317 Patch provided by Brad "anomie" Jorsch.
1319 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1320 Patch provided by Dean Brooks.
1322 TK/05 Add native DKIM support (does not depend on external libraries).
1324 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1325 Patch provided by Graeme Fowler.
1327 NM/16 Bugzilla 851: Documentation example syntax fix.
1329 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1331 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1334 NM/19 Bugzilla 745: TLS version reporting.
1335 Patch provided by Phil Pennock.
1337 NM/20 Bugzilla 167: bool: condition support.
1338 Patch provided by Phil Pennock.
1340 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1341 clients. Patch provided by Phil Pennock.
1343 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1344 Patch provided by Brad "anomie" Jorsch.
1346 NM/23 Bugzilla 687: Fix misparses in eximstats.
1347 Patch provided by Heiko Schlittermann.
1349 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1350 Patch provided by Heiko Schlittermann.
1352 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1353 plus update to original patch.
1355 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1357 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1358 Patch provided by David Brownlee.
1360 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1362 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1364 NM/30 Bugzilla 888: TLS documentation bugfixes.
1366 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1368 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1369 Unlike the original bugzilla I have changed all shell scripts in src tree.
1371 NM/33 Bugzilla 898: Transport filter timeout fix.
1372 Patch by Todd Rinaldo.
1374 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1375 Patch by Serge Demonchaux.
1377 NM/35 Bugzilla 39: Base64 decode bug fixes.
1378 Patch by Jakob Hirsch.
1380 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1382 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1384 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1386 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1392 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1393 ALT-N's libdkim that I have put here:
1394 http://duncanthrax.net/exim-experimental/
1396 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1397 were called 'true' and it seems that C99 defines that as a reserved
1398 keyword to be used with 'bool' variable types. That means you could
1399 not include C99-style headers which use bools without triggering
1400 build errors in sieve.c.
1402 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1403 as mailq or other aliases. Changed the --help handling significantly
1404 to do whats expected. exim_usage() emits usage/help information.
1406 SC/01 Added the -bylocaldomain option to eximstats.
1408 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1410 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1412 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1418 PH/01 Another patch from the Sieve maintainer.
1420 PH/02 When an IPv6 address is converted to a string for single-key lookup
1421 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1422 dots are used instead of colons so that keys in lsearch files need not
1423 contain colons. This was done some time before quoting was made available
1424 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1425 (notated using the quote facility) so as to distinguish them from IPv4
1426 keys. This meant that lookups for IP addresses in host lists did not work
1427 for iplsearch lookups.
1429 This has been fixed by arranging for IPv6 addresses to be expressed with
1430 colons if the lookup type is iplsearch. This is not incompatible, because
1431 previously such lookups could never work.
1433 The situation is now rather anomolous, since one *can* have colons in
1434 ordinary lsearch keys. However, making the change in all cases is
1435 incompatible and would probably break a number of configurations.
1437 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1440 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1441 conversion specification without a maximum field width, thereby enabling
1442 a rogue spamd server to cause a buffer overflow. While nobody in their
1443 right mind would setup Exim to query an untrusted spamd server, an
1444 attacker that gains access to a server running spamd could potentially
1445 exploit this vulnerability to run arbitrary code as the Exim user.
1447 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1448 $primary_hostname instead of what libspf2 thinks the hosts name is.
1450 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1451 a directory entry by the name of the lookup key. Previously, if a
1452 symlink pointed to a non-existing file or a file in a directory that
1453 Exim lacked permissions to read, a lookup for a key matching that
1454 symlink would fail. Now it is enough that a matching directory entry
1455 exists, symlink or not. (Bugzilla 503.)
1457 PH/03 The body_linecount and body_zerocount variables are now exported in the
1460 PH/04 Added the $dnslist_matched variable.
1462 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1463 This means they are set thereafter only if the connection becomes
1466 PH/06 Added the client_condition to authenticators so that some can be skipped
1467 by clients under certain conditions.
1469 PH/07 The error message for a badly-placed control=no_multiline_responses left
1470 "_responses" off the end of the name.
1472 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1474 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1475 (without spaces) instead of just copying the configuration text.
1477 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1479 PH/11 Added $max_received_linelength.
1481 PH/12 Added +ignore_defer and +include_defer to host lists.
1483 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1484 way in which PCRE > 7.0 is built.
1486 PH/14 Implemented queue_only_load_latch.
1488 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1489 MAIL command. The effect was to mangle the value on 64-bit systems.
1491 PH/16 Another patch from the Sieve maintainer.
1493 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1495 PH/18 If a system quota error occurred while trying to create the file for
1496 a maildir delivery, the message "Mailbox is full" was not appended to the
1497 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1498 only to a quota excession during the actual writing of the file.
1500 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1501 characters?) which causes problems in log lines. The DN values are now
1502 passed through string_printing() before being added to log lines.
1504 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1505 and InterBase are left for another time.)
1507 PH/21 Added message_body_newlines option.
1509 PH/22 Guard against possible overflow in moan_check_errorcopy().
1511 PH/23 POSIX allows open() to be a macro; guard against that.
1513 PH/24 If the recipient of an error message contained an @ in the local part
1514 (suitably quoted, of course), incorrect values were put in $domain and
1515 $local_part during the evaluation of errors_copy.
1521 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1522 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1525 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1526 issue a MAIL command.
1528 PH/02 In an ACL statement such as
1530 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1532 if a client was not listed at all, or was listed with a value other than
1533 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1534 the condition was not true (as it should be), so access was not denied.
1535 The bug was that the ! inversion was incorrectly passed on to the second
1536 item. This has been fixed.
1538 PH/03 Added additional dnslists conditions == and =& which are different from
1539 = and & when the dns lookup returns more than one IP address.
1541 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1542 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1544 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1545 FSYNC, which compiles an option called disable_fsync that allows for
1546 bypassing fsync(). The documentation is heavily laced with warnings.
1548 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1550 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1551 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1552 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1553 including adding "make clean"; (3) Added -fPIC when compiling the test
1554 dynamically loaded module, to get rid of a warning.
1556 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1557 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1558 = 0s. The bug is otherwise harmless.
1560 PH/07 There was a bug in the dovecot authenticator such that the value of
1561 $auth1 could be overwritten, and so not correctly preserved, after a
1562 successful authentication. This usually meant that the value preserved by
1563 the server_setid option was incorrect.
1565 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1567 PH/09 Installed PCRE release 7.0.
1569 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1570 run for batched SMTP input. It is now run at the start of every message
1571 in the batch. While fixing this I discovered that the process information
1572 (output by running exiwhat) was not always getting set for -bs and -bS
1573 input. This is fixed, and it now also says "batched" for BSMTP.
1575 PH/11 Added control=no_pipelining.
1577 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1578 patch, slightly modified), and move the expansion of helo_data till after
1579 the connection is made in the smtp transport (so it can use these
1582 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1584 PH/14 Added log_selector = +pid.
1586 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1588 PH/16 Add ${if forany and ${if forall.
1590 PH/17 Added dsn_from option to vary the From: line in DSNs.
1592 PH/18 Flush SMTP output before performing a callout, unless control =
1593 no_callout_flush is set.
1595 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1596 was true (the default) a successful delivery failed to delete the retry
1597 item, thus causing premature timeout of the address. The bug is now
1600 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1602 PH/21 Long custom messages for fakedefer and fakereject are now split up
1603 into multiline reponses in the same way that messages for "deny" and
1604 other ACL rejections are.
1606 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1607 with slight modification.
1609 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1610 draft, changing the syntax and factoring some duplicate code".
1612 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1613 for deliveries of the second and subsequent messages over the same SMTP
1616 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1617 ${reduce, with only minor "tidies".
1619 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1621 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1622 expansion side effects.
1624 PH/27 When a message times out after an over-quota error from an Exim-imposed
1625 quota, the bounce message says "mailbox is full". This message was not
1626 being given when it was a system quota that was exceeded. It now should
1629 MH/03 Made $recipients available in local_scan(). local_scan() already has
1630 better access to the recipient list through recipients_list[], but
1631 $recipients can be useful in postmaster-provided expansion strings.
1633 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1634 in the case of a MAIL command with additional options following the
1635 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1636 were accidentally chopped off.
1638 PH/29 SMTP synchronization checks are implemented when a command is read -
1639 there is a check that no more input is waiting when there shouldn't be
1640 any. However, for some commands, a delay in an ACL can mean that it is
1641 some time before the response is written. In this time, more input might
1642 arrive, invalidly. So now there are extra checks after an ACL has run for
1643 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1644 pipelining has not been advertised.
1646 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1648 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1649 returned in the SMTP response when smtp_return_error_details was false.
1650 This has been fixed.
1652 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1653 instead of the C I/O that was originally supplied, because problems were
1654 reported on Solaris.
1656 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1657 Exim which did not show up earlier: it was assuming that a call to
1658 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1659 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1660 was a macro that became an assignment, so it seemed to work. This has
1661 changed to a proper function call with a void return, hence the compile
1662 error. Exim's code has been fixed.
1664 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1667 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1669 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1671 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1672 because it turns out to be too big or there is a timeout) there is no
1673 "Completed" line in the log. When some messages of this type were
1674 selected by exigrep, they were listed as "not completed". Others were
1675 picked up by some special patterns. I have improved the selection
1676 criteria to be more general.
1678 PH/38 The host_find_failed option in the manualroute router can now be set
1679 to "ignore", to completely ignore a host whose IP address cannot be
1680 found. If all hosts are ignored, the behaviour is controlled by the new
1681 host_all_ignored option.
1683 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1684 homing or because of multiple MX records with /mx) generated more than
1685 one IP address, and the following item turned out to be the local host,
1686 all the secondary addresses of the first item were incorrectly removed
1687 from the list, along with the local host and any following hosts (which
1688 is what is supposed to happen).
1690 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1691 whoever called Exim into the -H file. In the case of the daemon it was
1692 behaving confusingly. When first started, it used values for whoever
1693 started the daemon, but after a SIGHUP it used the Exim user (because it
1694 calls itself on a restart). I have changed the code so that it now always
1697 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1698 message are rejected with the same error (e.g. no authentication or bad
1699 sender address), and a DATA command is nevertheless sent (as can happen
1700 with PIPELINING or a stupid MUA), the error message that was given to the
1701 RCPT commands is included in the rejection of the DATA command. This is
1702 intended to be helpful for MUAs that show only the final error to their
1705 PH/42 Another patch from the Sieve maintainer.
1707 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1708 Eximstats - Fixed some broken HTML links and added missing column headers
1710 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1711 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1713 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1715 PH/43 Yet another patch from the Sieve maintainer.
1717 PH/44 I found a way to check for a TCP/IP connection going away before sending
1718 the response to the final '.' that terminates a message, but only in the
1719 case where the client has not sent further data following the '.'
1720 (unfortunately, this is allowed). However, in many cases there won't be
1721 any further data because there won't be any more messages to send. A call
1722 to select() can be used: if it shows that the input is "ready", there is
1723 either input waiting, or the socket has been closed. An attempt to read
1724 the next input character can distinguish the two cases. Previously, Exim
1725 would have sent an OK response which the client would never have see.
1726 This could lead to message repetition. This fix should cure that, at
1727 least in a lot of common cases.
1729 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1730 advertised in response to EHLO.
1736 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1737 fixed by 4.65/MH/01 (is this a record?) are fixed:
1739 (i) An empty string was always treated as zero by the numeric comparison
1740 operators. This behaviour has been restored.
1742 (ii) It is documented that the numeric comparison operators always treat
1743 their arguments as decimal numbers. This was broken in that numbers
1744 starting with 0 were being interpreted as octal.
1746 While fixing these problems I realized that there was another issue that
1747 hadn't been noticed. Values of message_size_limit (both the global option
1748 and the transport option) were treated as octal if they started with 0.
1749 The documentation was vague. These values are now always treated as
1750 decimal, and I will make that clear in the documentation.
1756 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1757 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1760 MH/01 Don't check that the operands of numeric comparison operators are
1761 integers when their expansion is in "skipping" mode (fixes bug
1762 introduced by 4.64-PH/07).
1764 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1765 child addresses, Exim now panics and dies. Previously, because the count
1766 is held in a short int, deliveries were likely to be lost. As such a
1767 large number of recipients for a single message is ridiculous
1768 (performance will be very, very poor), I have chosen to impose a limit
1769 rather than extend the field.
1775 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1776 leftover -K file (the existence of which was triggered by #402).
1777 While we were at it, introduced process PID as part of the -K
1778 filename. This should rule out race conditions when creating
1781 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1782 processing considerably. Previous code took too long for large mails,
1783 triggering a timeout which in turn triggers #401.
1785 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1786 in the DK code in transports.c. sendfile() is not really portable,
1787 hence the _LINUX specificness.
1789 TF/01 In the add_headers option to the mail command in an Exim filter,
1790 there was a bug that Exim would claim a syntax error in any
1791 header after the first one which had an odd number of characters
1794 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1795 callout verification, Exim cached a "reject" for the entire domain. This
1796 is correct for most verifications, but it is not correct for a recipient
1797 verification with use_sender or use_postmaster set, because in that case
1798 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1799 case of MAIL FROM:<> rejection from other early rejections (e.g.
1800 rejection of HELO). When verifying a recipient using a non-null MAIL
1801 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1802 Whatever the result of the callout, the value of the domain cache is
1803 left unchanged (for any other kind of callout, getting as far as trying
1804 RCPT means that the domain itself is ok).
1806 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1809 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1810 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1811 session, the daemon ignores ECONNECT errors and logs others; it now
1812 ignores EPIPE as well.
1814 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1815 (quoted-printable decoding).
1817 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1818 later the small subsequent patch to fix an introduced bug.
1820 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1822 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1824 PH/08 An error is now given if message_size_limit is specified negative.
1826 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1827 to be given (somewhat) arbitrary names.
1829 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1832 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1833 miscellaneous code fixes
1835 PH/10 Added the log_reject_target ACL modifier to specify where to log
1838 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1839 hostname. This is wrong, because it relates to the incoming message (and
1840 probably the interface on which it is arriving) and not to the outgoing
1841 callout (which could be using a different interface). This has been
1842 changed to use the value of the helo_data option from the smtp transport
1843 instead - this is what is used when a message is actually being sent. If
1844 there is no remote transport (possible with a router that sets up host
1845 addresses), $smtp_active_hostname is used.
1847 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1848 tweaks were necessary in order to get it to work (see also 21 below):
1849 (a) The code assumed that strncpy() returns a negative number on buffer
1850 overflow, which isn't the case. Replaced with Exim's string_format()
1852 (b) There were several signed/unsigned issues. I just did the minimum
1853 hacking in of casts. There is scope for a larger refactoring.
1854 (c) The code used strcasecmp() which is not a standard C function.
1855 Replaced with Exim's strcmpic() function.
1856 (d) The code set only $1; it now sets $auth1 as well.
1857 (e) A simple test gave the error "authentication client didn't specify
1858 service in request". It would seem that Dovecot has changed its
1859 interface. Fortunately there's a specification; I followed it and
1860 changed what the client sends and it appears to be working now.
1862 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1865 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1866 address A is aliased to B and C, where B exists and C does not. Without
1867 -v the output is "A verified" because verification stops after a
1868 successful redirection if more than one address is generated. However,
1869 with -v the child addresses are also verified. Exim was outputting "A
1870 failed to verify" and then showing the successful verification for C,
1871 with its parentage. It now outputs "B failed to verify", showing B's
1872 parentage before showing the successful verification of C.
1874 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1875 look up a TXT record in a specific list after matching in a combined
1878 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1879 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1880 they consult the DNS. I had assumed they would set it the way they
1881 wanted; and indeed my experiments on Linux seem to show that in some
1882 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1883 To be on the safe side, however, I have now made the interface to
1884 host_find_byname() similar to host_find_bydns(), with an argument
1885 containing the DNS resolver options. The host_find_byname() function now
1886 sets these options at its start, just as host_find_bydns() does. The smtp
1887 transport options dns_qualify_single and dns_search_parents are passed to
1888 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1889 of host_find_byname() use the default settings of RES_DEFNAMES
1890 (qualify_single) but not RES_DNSRCH (search_parents).
1892 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1893 spool_read_header() do less string testing, by means of a preliminary
1894 switch on the second character of optional "-foo" lines. (This is
1895 overdue, caused by the large number of possibilities that now exist.
1896 Originally there were few.) While I was there, I also converted the
1897 str(n)cmp tests so they don't re-test the leading "-" and the first
1898 character, in the hope this might squeeze out yet more improvement.
1900 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1901 flag allowing group syntax was set by the header_syntax check but not
1902 turned off, possible causing trouble later; (2) The flag was not being
1903 set at all for the header_verify test, causing "group"-style headers to
1904 be rejected. I have now set it in this case, and also caused header_
1905 verify to ignore an empty address taken from a group. While doing this, I
1906 came across some other cases where the code for allowing group syntax
1907 while scanning a header line wasn't quite right (mostly, not resetting
1908 the flag correctly in the right place). These bugs could have caused
1909 trouble for malformed header lines. I hope it is now all correct.
1911 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1912 with the "reply" argument non-NULL. The code, however (which originally
1913 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1914 but it didn't always do it. This confused somebody who was copying the
1915 code for some other use. I have removed all the tests.
1917 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1918 feature that was used to support insecure browsers during the U.S. crypto
1919 embargo. It requires special client support, and Exim is probably the
1920 only MTA that supported it -- and would never use it because real RSA is
1921 always available. This code has been removed, because it had the bad
1922 effect of slowing Exim down by computing (never used) parameters for the
1923 RSA_EXPORT functionality.
1925 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1926 authenticator to fail if there's a tab character in the incoming data
1927 (there should never be unless someone is messing about, as it's supposed
1928 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1929 if the connection is using TLS or if the remote IP is the same as the
1930 local IP, and the "valid-client-cert option" if a client certificate has
1933 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1934 authenticators. This can be used for authorization after authentication
1935 succeeds. (In the case of plaintext, it servers for both authentication
1938 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1939 if any retry times were supplied.
1941 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1942 connection, where there is no client IP address to check. In this
1943 situation, the verify now always succeeds.
1945 PH/25 Applied John Jetmore's -Mset patch.
1947 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1949 PH/27 In a string expansion for a processed (not raw) header when multiple
1950 headers of the same name were present, leading whitespace was being
1951 removed from all of them, but trailing whitespace was being removed only
1952 from the last one. Now trailing whitespace is removed from each header
1953 before concatenation. Completely empty headers in a concatenation (as
1954 before) are ignored.
1956 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1957 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1959 PH/29 [Removed. This was a change that I later backed out, and forgot to
1960 correct the ChangeLog entry (that I had efficiently created) before
1961 committing the later change.]
1963 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1964 address errors (4xx response to RCPT) over the same connection as other
1965 messages routed to the same hosts. Such deliveries are always "forced",
1966 so retry times are not inspected. This resulted in far too many retries
1967 for the affected addresses. The effect occurred only when there were more
1968 hosts than the hosts_max_try setting in the smtp transport when it had
1969 the 4xx errors. Those hosts that it had tried were not added to the list
1970 of hosts for which the message was waiting, so if all were tried, there
1971 was no problem. Two fixes have been applied:
1973 (i) If there are any address or message errors in an SMTP delivery, none
1974 of the hosts (tried or untried) are now added to the list of hosts
1975 for which the message is waiting, so the message should not be a
1976 candidate for sending over the same connection that was used for a
1977 successful delivery of some other message. This seems entirely
1978 reasonable: after all the message is NOT "waiting for some host".
1979 This is so "obvious" that I'm not sure why it wasn't done
1980 previously. Hope I haven't missed anything, but it can't do any
1981 harm, as the worst effect is to miss an optimization.
1983 (ii) If, despite (i), such a delivery is accidentally attempted, the
1984 routing retry time is respected, so at least it doesn't keep
1985 hammering the server.
1987 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1988 in ${readsocket because some servers need this prod.
1990 PH/32 Added some extra debug output when updating a wait-xxx database.
1992 PH/33 The hint "could be header name not terminated by colon", which has been
1993 given for certain expansion errors for a long time, was not being given
1994 for the ${if def:h_colon_omitted{... case.
1996 PH/34 The spec says: "With one important exception, whenever a domain list is
1997 being scanned, $domain contains the subject domain." There was at least
1998 one case where this was not true.
2000 PH/35 The error "getsockname() failed: connection reset by peer" was being
2001 written to the panic log as well as the main log, but it isn't really
2002 panic-worthy as it just means the connection died rather early on. I have
2003 removed the panic log writing for the ECONNRESET error when getsockname()
2006 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2007 runs only) independently of the message's sender address. This meant
2008 that, if the 4xx error was in fact related to the sender, a different
2009 message to the same recipient with a different sender could confuse
2010 things. In particualar, this can happen when sending to a greylisting
2011 server, but other circumstances could also provoke similar problems.
2012 I have changed the default so that the retry time for these errors is now
2013 based a combination of the sender and recipient addresses. This change
2014 can be overridden by setting address_retry_include_sender=false in the
2017 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2018 remote server are returned as part of bounce messages. This was not
2019 happening for LMTP over a pipe (the lmtp transport), but now it is the
2020 same for both kinds of LMTP.
2022 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2023 in header lines that were in fact CNAMEs. This is no longer the case.
2025 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2026 and queue runs started by the daemon processed all messages. This has
2027 been fixed so that -R and -S can now usefully be given with -q<time>.
2029 PH/40 Import PCRE release 6.7 (fixes some bugs).
2031 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2033 PH/42 Give an error if -q is specified more than once.
2035 PH/43 Renamed the variables $interface_address and $interface_port as
2036 $received_ip_address and $received_port, to make it clear that these
2037 values apply to message reception, and not to the outgoing interface when
2038 a message is delivered. (The old names remain recognized, of course.)
2040 PH/44 There was no timeout on the connect() call when using a Unix domain
2041 socket in the ${readsocket expansion. There now is.
2043 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2044 be meaningful with "accept".
2046 SC/01 Eximstats V1.43
2047 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2049 SC/02 Eximstats V1.44
2050 Use a glob alias rather than an array ref in the generated
2051 parser. This improves both readability and performance.
2053 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2054 Collect SpamAssassin and rejection statistics.
2055 Don't display local sender or destination tables unless
2056 there is data to show.
2057 Added average volumes into the top table text output.
2059 SC/04 Eximstats V1.46
2060 Collect data on the number of addresses (recipients)
2061 as well as the number of messages.
2063 SC/05 Eximstats V1.47
2064 Added 'Message too big' to the list of mail rejection
2065 reasons (thanks to Marco Gaiarin).
2067 SC/06 Eximstats V1.48
2068 Mainlog lines which have GMT offsets and are too short to
2069 have a flag are now skipped.
2071 SC/07 Eximstats V1.49 (Alain Williams)
2072 Added the -emptyok flag.
2074 SC/08 Eximstats V1.50
2075 Fixes for obtaining the IP address from reject messages.
2077 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2078 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2079 whitesspace changes from 4.64-PH/27
2081 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2084 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2085 are found, allow negative numbers in numeric criteria)
2087 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2089 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2090 to match changes made in 4.64-PH/43
2092 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2094 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2095 "domain" is actually an IP address, and force a failure. This locks out
2096 those revolvers/nameservers that support "A-for-A" lookups, in
2097 contravention of the specifications.
2099 PH/48 When a host name was looked up from an IP address, and the subsequent
2100 forward lookup of the name timed out, the host name was left in
2101 $sender_host_name, contrary to the specification.
2103 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2104 restricted to single-key lookups, Exim was not diagnosing an error if
2105 * or *@ was used with a query-style lookup.
2107 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2109 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2110 long ago, but noone interested enough thought of it. Let's just say that
2111 the "1.1" means that there are some new functions that weren't there at
2112 some point in the past.
2114 PH/51 Error processing for expansion failure of helo_data from an smtp
2115 transport during callout processing was broken.
2117 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2118 tested/used via the -bh/-bhc/-bs options.
2120 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2121 bug, fixed in subsequent PCRE releases).
2123 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2124 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2126 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2132 SC/01 Use a glob alias rather than an array ref in eximstats generated
2133 parser. This improves both readability and performance.
2135 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2136 Don't display local sender or destination tables in eximstats unless
2137 there is data to show.
2138 Added average volumes into the eximstats top table text output.
2140 SC/03 Collect data on the number of addresses (recipients) as well
2141 as the number of messages in eximstats.
2143 TF/01 Correct an error in the documentation for the redirect router. Exim
2144 does (usually) call initgroups() when daemonizing.
2146 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2147 with consistent privilege compared to when running as a daemon.
2149 TF/03 Note in the spec that $authenticated_id is not set for local
2150 submissions from trusted users.
2152 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2153 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2155 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2156 by adding some example configuration directives to the default
2157 configuration file. A little bit of work is required to uncomment the
2158 directives and define how usernames and passwords are checked, but
2159 there is now a framework to start from.
2161 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2162 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2163 without this. I don't know how relevant this is to other LDAP libraries.
2165 PH/02 Add the verb name to the "unknown ACL verb" error.
2167 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2169 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2171 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2172 directory not expanded when it should be if an expanded home directory
2173 was set for the address (which is overridden by the transport).
2175 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2178 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2179 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2180 because it is too late at that time, and has no effect.
2182 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2183 security issue with \' (bugzilla #107). I could not use the
2184 PQescapeStringConn() function, because it needs a PGconn value as one of
2187 PH/08 When testing addresses using -bt, indicate those final addresses that
2188 are duplicates that would not cause an additional delivery. At least one
2189 person was confused, thinking that -bt output corresponded to deliveries.
2190 (Suppressing duplicates isn't a good idea as you lose the information
2191 about possibly different redirections that led to the duplicates.)
2193 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2194 systems where poll() doesn't work, in particular OS X.
2196 PH/10 Added more information to debugging output for retry time not reached.
2198 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2199 operations in malware.c.
2201 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2204 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2205 syslog_duplication set false, log lines that would normally be written
2206 both the the main log and to the reject log were not written to syslog at
2209 PH/14 In the default configuration, change the use of "message" in ACL warn
2210 statements to "add_header".
2212 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2213 not followed by a command (e.g. "seen endif").
2215 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2216 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2219 PH/17 Added extra conditions to the default value of delay_warning_condition
2223 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2224 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2225 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2228 The Auto-Submitted: and various List- headers are standardised, whereas I
2229 don't think Precedence: ever was.
2231 PH/18 Refactored debugging code in route_finduser() to show more information,
2232 in particular, the error code if getpwnam() issues one.
2234 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2235 This is apparently needed in addition to the PH/07 change above to avoid
2236 any possible encoding problems.
2238 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2239 but not after initializing Perl.
2241 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2242 output them only if debugging. By default they are written stderr,
2243 apparently, which is not desirable.
2245 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2248 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2251 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2253 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2254 authenticated or an ident call has been made. Suppress the default
2255 values for $authenticated_id and $authenticated_sender (but permit -oMai
2256 and -oMas) when testing with -bh.
2258 PH/24 Re-jigged the order of the tests in the default configuration so that the
2259 tests for valid domains and recipients precede the DNS black list and CSA
2260 tests, on the grounds that those ones are more expensive.
2262 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2263 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2264 This bug exists in every version of Exim that I still have, right back to
2267 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2268 However, an attempt to turn on case-sensitivity in a regex key by
2269 including (?-i) didn't work because the subject string was already
2270 lowercased, and the effects were non-intuitive. It turns out that a
2271 one-line patch can be used to allow (?-i) to work as expected.
2277 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2278 other effects) broke the use of negated acl sub-conditions.
2280 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2283 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2284 "Deny" causes Exim to reject the incoming connection with a 554 error.
2285 Unfortunately, if there is a major crisis, such as a disk failure,
2286 tcp-wrappers gives "deny", whereas what one would like would be some
2287 kind of temporary error. A kludge has been added to help with this.
2288 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2289 554 error is used if errno is still zero or contains ENOENT (which occurs
2290 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2293 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2295 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2296 errors. Otherwise a message that provokes a temporary error (when other
2297 messages do not) can cause a whole host to time out.
2299 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2300 addresses were routed directly to files or pipes from a redirect router.
2301 File deliveries just didn't batch; pipe deliveries might have suffered
2304 PH/06 A failure to get a lock for a hints database would erroneously always say
2305 "Failed to get write lock", even when it was really a read lock.
2307 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2308 of 0600. This has been changed to use the value of the lockfile_mode
2309 option (which defaults to 0600).
2311 PH/08 Applied small patch from the Sieve maintainer.
2313 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2314 folder from quota calculations, a direct delivery into this folder messed
2315 up the contents of the maildirsize file. This was because the regex was
2316 used only to exclude .Trash (or whatever) when the size of the mailbox
2317 was calculated. There was no check that a delivery was happening into an
2318 excluded directory. This bug has been fixed by ignoring all quota
2319 processing for deliveries into excluded directories.
2321 PH/10 Added the maildirfolder_create_regex option to appendfile.
2327 PH/01 The code for finding all the local interface addresses on a FreeBSD
2328 system running IPv6 was broken. This may well have applied to all BSD
2329 systems, as well as to others that have similar system calls. The broken
2330 code found IPv4 interfaces correctly, but gave incorrect values for the
2331 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2332 that it would not match correctly against @[] and not recognize the IPv6
2335 PH/02 The ipliteral router was not recognizing addresses of the form user@
2336 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2338 PH/03 Added disable_ipv6.
2340 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2341 decoded form, because it is most often used to construct To: headers
2342 lines in autoreplies, and the decoded form may well be syntactically
2343 invalid. However, $reply_address has leading white space removed, and all
2344 newlines turned into spaces so that the autoreply transport does not
2347 PH/05 If group was specified without a user on a router, and no group or user
2348 was specified on a transport, the group from the router was ignored.
2350 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2351 for visible compile-time settings that can be used to change these
2352 numbers, for those that want even more. Backwards compatibility with old
2353 spool files has been maintained. However, going back to a previous Exim
2354 release will lost any variables that are in spool files.
2356 PH/07 Two small changes when running in the test harness: increase delay when
2357 passing a TCP/IP connection to a new process, in case the original
2358 process has to generate a bounce, and remove special handling of
2359 127.0.0.2 (sic), which is no longer necessary.
2361 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2362 be the same on different OS.
2364 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2367 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2368 whether --show-vars was specified or not
2370 JJ/02 exipick: Added support for new ACL variable spool format introduced
2373 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2374 syntactically invalid From: or Reply-to: line, and a filter used this to
2375 generate an autoreply, and therefore failed to obtain an address for the
2376 autoreply, Exim could try to deliver to a non-existent relative file
2377 name, causing unrelated and misleading errors. What now happens is that
2378 it logs this as a hard delivery error, but does not attempt to create a
2381 PH/11 The exinext utility has a -C option for testing purposes, but although
2382 the given file was scanned by exinext itself; it wasn't being passed on
2383 when Exim was called.
2385 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2386 an end-of-file indication when reading a command response.
2388 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2389 compiled. In many other places in Exim, IPv6 addresses are always
2390 recognized, so I have changed this. It also means that IPv4 domain
2391 literals of the form [IPV4:n.n.n.n] are now always recognized.
2393 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2394 used if the router is not running as root, for example, when verifying at
2395 ACL time, or when using -bh. The debugging output from this situation was
2396 non-existent - all you got was a failure to exec. I have made two
2399 (a) Failures to set uid/gid, the current directory, or a process leader
2400 in a subprocess such as that created by queryprogram now generate
2401 suitable debugging ouput when -d is set.
2403 (b) The queryprogram router detects when it is not running as root,
2404 outputs suitable debugging information if -d is set, and then runs
2405 the subprocess without attempting to change uid/gid.
2407 PH/15 Minor change to Makefile for building test_host (undocumented testing
2410 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2411 additional section of a DNS packet that returns MX or SRV records.
2412 Instead, it always explicitly searches for A/AAAA records. This avoids
2413 major problems that occur when a DNS server includes only records of one
2414 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2415 fixed another bug: if SRV records were looked up and the corresponding
2416 address records were *not* found in the additional section, the port
2417 values from the SRV records were lost.
2419 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2420 using the correct key (the original address) when searching the retry
2421 rules in order to find which one to use for generating the retry hint.
2423 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2424 adding the default one. Similarly, if it contains a Reply-To: header, the
2425 errors_reply_to option, if set, is not used.
2427 PH/19 When calculating a retry time, Exim used to measure the "time since
2428 failure" by looking at the "first failed" field in the retry record. Now
2429 it does not use this if it is later than than the arrival time of the
2430 message. Instead it uses the arrival time. This makes for better
2431 behaviour in cases where some deliveries succeed, thus re-setting the
2432 "first failed" field. An example is a quota failure for a huge message
2433 when small messages continue to be delivered. Without this change, the
2434 "time since failure" will always be short, possible causing more frequent
2435 delivery attempts for the huge message than are intended.
2436 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2438 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2439 $1, $2, $3) because the numerical variables can be reset during some
2440 expansion items (e.g. "match"), thereby losing the authentication data.
2442 PH/21 Make -bV show the size of off_t variables so that the test suite can
2443 decide whether to run tests for quotas > 2G.
2445 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2446 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2447 filecount value is greater than 2G or if a quota value is greater than 2G
2448 on a system where the size of off_t is not greater than 4, a panic error
2451 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2452 never match. The debug and -bh output now contains an explicit error
2453 message indicating a malformed IPv4 address or mask.
2455 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2456 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2457 PH/23 above applies.
2459 PH/25 Do not write to syslog when running in the test harness. The only
2460 occasion when this arises is a failure to open the main or panic logs
2461 (for which there is an explicit test).
2463 PH/26 Added the /no_tell option to "control=freeze".
2465 PH/27 If a host name lookup failed very early in a connection, for example, if
2466 the IP address matched host_lookup and the reverse lookup yielded a name
2467 that did not have a forward lookup, an error message of the form "no IP
2468 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2469 could be logged. Now it outputs the IP address instead of "NULL".
2471 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2472 allows the sender and the authenticated sender to be set when
2473 submitting a message from within Exim. Since child_open_exim() is
2474 documented for local_scan(), the new function should be too.
2476 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2477 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2478 results in an empty string is now treated as unset.
2480 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2482 PH/31 Added sender_verify_fail logging option.
2484 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2485 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2486 tidied the source and removed it altogether.
2488 PH/33 When a queue run was abandoned because the load average was too high, a
2489 log line was always written; now it is written only if the queue_run log
2490 selector is set. In addition, the log line for abandonment now contains
2491 information about the queue run such as the pid. This is always present
2492 in "start" and "stop" lines but was omitted from the "abandon" line.
2494 PH/34 Omit spaces between a header name and the colon in the error message that
2495 is given when verify = headers_syntax fails (if there are lots of them,
2496 the message gets confusing).
2498 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2499 names, as there are now some PTR records that contain slashes. This check
2500 is only to protect against broken name servers that fall over on strange
2501 characters, so the fact that it applies to all lookups doesn't matter.
2503 PH/36 Now that the new test suite is complete, we can remove some of the
2504 special code in Exim that was needed for the old test suite. For example,
2505 sorting DNS records because real resolvers return them in an arbitrary
2506 order. The new test suite's fake resolver always returns records in the
2509 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2510 bounces) except when queue_only is set, to avoid logging races between
2511 the different processes.
2513 PH/38 Panic-die if .include specifies a non-absolute path.
2515 PH/39 A tweak to the "H" retry rule from its user.
2517 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2518 a label. They prevented compilation on older perls.
2520 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2521 a warning to be raised on newish perls.
2523 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2524 on queue. Changes to match documented behaviour of showing count of
2525 messages matching specified criteria.
2527 PH/40 Changed the default ident timeout from 30s to 5s.
2529 PH/41 Added support for the use of login_cap features, on those BSD systems
2530 that have them, for controlling the resources used by pipe deliveries.
2532 PH/42 The content-scanning code uses fopen() to create files in which to put
2533 message data. Previously it was not paying any attention to the mode of
2534 the files. Exim runs with umask(0) because the rest of the code creates
2535 files with open(), and sets the required mode explicitly. Thus, these
2536 files were ending up world-writeable. This was not a big issue, because,
2537 being within the spool directory, they were not world-accessible. I have
2538 created a function called modefopen, which takes an additional mode
2539 argument. It sets umask(777), creates the file, chmods it to the required
2540 mode, then resets the umask. All the relevant calls to fopen() in the
2541 content scanning code have been changed to use this function.
2543 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2544 to 24 hours. This avoids potential overflow problems when processing G
2545 and H retry rules. I suspect nobody ever tinkers with this value.
2547 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2549 PH/45 When the plaintext authenticator is running as a client, the server's
2550 challenges are checked to ensure they are valid base64 strings. By
2551 default, the authentication attempt is cancelled if an invalid string is
2552 received. Setting client_ignore_invalid_base64 true ignores these errors.
2553 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2554 they are received. Thus, the responses can be made to depend on the
2555 challenges. If an invalid string is ignored, an empty string is placed in
2558 PH/46 Messages that are created by the autoreply transport now contains a
2559 References: header, in accordance with RFCs 2822 and 3834.
2561 PH/47 Added authenticated_sender_force to the smtp transport.
2563 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2565 PH/49 Installed latest patch from the Sieve maintainer.
2567 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2568 was also set, the appendfile transport was unnecessarily scanning a
2569 directory of message files (e.g. for maildir delivery) to find the count
2570 of files (along with the size), even though it did not need this
2571 information. It now does the scan only if it needs to find either the
2572 size of the count of files.
2574 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2576 PH/52 Two bugs concerned with error handling when the smtp transport is
2579 (i) Exim was not creating retry information for temporary errors given
2580 for individual recipients after the DATA command when the smtp transport
2581 was used in LMTP mode. This meant that they could be retried too
2582 frequently, and not timed out correctly.
2584 (ii) Exim was setting the flag that allows error details to be returned
2585 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2586 recipients that were returned after the DATA command.
2588 PH/53 This is related to PH/52, but is more general: for any failing address,
2589 when detailed error information was permitted to be returned to the
2590 sender, but the error was temporary, then after the final timeout, only
2591 "retry timeout exceeded" was returned. Now it returns the full error as
2592 well as "retry timeout exceeded".
2594 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2595 do this, and (what is worse) MTAs that accept it.
2597 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2598 will now be deprecated.
2600 PH/56 New os.c-cygwin from the Cygwin maintainer.
2602 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2603 formats (previously only available in exim formats via -bpr, -bpru,
2604 and -bpra. Now also available in native and exiqgrep formats)
2606 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2607 with very large, slow to parse queues
2609 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2611 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2613 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2614 responses to authentication challenges, though it was showing the
2615 challenges; (ii) I've removed the CR characters from the debug output for
2618 PH/58 Allow for the insertion of a newline as well as a space when a string
2619 is turned into more than one encoded-word during RFC 2047 encoding. The
2620 Sieve code now uses this.
2622 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2623 data_4xx, lost_connection, tls_required.
2625 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2626 message was being sent as an SMTP response.
2628 PH/61 Add -l and -k options to exicyclog.
2630 PH/62 When verifying, if an address was redirected to one new address, so that
2631 verification continued, and the new address failed or deferred after
2632 having set something in $address_data, the value of $address_data was not
2633 passed back to the ACL. This was different to the case when no
2634 redirection occurred. The value is now passed back in both cases.
2636 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2637 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2638 use login_cap.h, so on its own it isn't the distinguishing feature. The
2639 new name refers directly to the setclassresources() function.
2641 PH/65 Added configuration files for NetBSD3.
2643 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2645 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2646 is preferred over IPv4.
2648 PH/68 The bounce_return_message and bounce_return_body options were not being
2649 honoured for bounces generated during the reception of non-SMTP messages.
2650 In particular, this applied to messages rejected by the ACL. This bug has
2651 been fixed. However, if bounce_return_message is true and bounce_return_
2652 body is false, the headers that are returned for a non-SMTP message
2653 include only those that have been read before the error was detected.
2654 (In the case of an ACL rejection, they have all been read.)
2656 PH/69 The HTML version of the specification is now built in a directory called
2657 spec_html instead of spec.html, because the latter looks like a path with
2658 a MIME-type, and this confuses some software.
2660 PH/70 Catch two compiler warnings in sieve.c.
2662 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2663 function verify_get_ident() calls ip_connect() to connect a socket, but
2664 if the "connect()" function timed out, ip_connect() used to close the
2665 socket. However, verify_get_ident() also closes the socket later, and in
2666 between Exim writes to the log, which may get opened at this point. When
2667 the socket was closed in ip_connect(), the log could get the same file
2668 descriptor number as the socket. This naturally causes chaos. The fix is
2669 not to close the socket in ip_connect(); the socket should be closed by
2670 the function that creates it. There was only one place in the code where
2671 this was missing, in the iplookup router, which I don't think anybody now
2672 uses, but I've fixed it anyway.
2674 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2675 well as to direct DNS lookups. Otherwise the handling of names in host
2676 lists is inconsistent and therefore confusing.
2682 PH/01 Two changes to the default runtime configuration:
2684 (1) Move the checks for relay_from_hosts and authenticated clients from
2685 after to before the (commented out) DNS black list checks.
2687 (2) Add control=submission to the relay_from_hosts and authenticated
2688 clients checks, on the grounds that messages accepted by these
2689 statements are most likely to be submissions.
2691 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2693 (1) Generate an error if the third argument for the ${prvs expansion is
2696 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2699 (3) Reset the variables that are obtained from the first argument of
2700 ${prvscheck and used in the second argument before leaving the code,
2701 because their memory is reclaimed, so using them afterwards may do
2704 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2705 one (it's much easier than Tom thought :-).
2707 (5) Because of (4), we can now allow for the use of $prvscheck_result
2708 inside the third argument.
2710 PH/03 For some reason, the default setting of PATH when running a command from
2711 a pipe transport was just "/usr/bin". I have changed it to
2714 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2715 anything to be listed in the output from -bV.
2717 PH/05 When a filter generated an autoreply, the entire To: header line was
2718 quoted in the delivery log line, like this:
2720 => >A.N.Other <ano@some.domain> <original@ddress> ...
2722 This has been changed so that it extracts the operative address. There
2723 may be more than one such address. If so, they are comma-separated, like
2726 => >ano@some.domain,ona@other.domain <original@ddress> ...
2728 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2729 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2730 not being looked up in the rDNS to get a host name, Exim was showing the
2731 IP address twice in Received: lines, even though the IP addresses were
2732 identical. For example:
2734 Received: from [1.2.3.4] (helo=[1.2.3.4])
2736 However, if the real host name was known, it was omitting the HELO data
2737 if it matched the actual IP address. This has been tidied up so that it
2738 doesn't show the same IP address twice.
2740 PH/07 When both +timestamp and +memory debugging was on, the value given by
2741 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2742 called by the debug printing, thereby overwriting the timestamp buffer.
2743 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2745 PH/08 When the original message was included in an autoreply transport, it
2746 always said "this is a copy of the message, including all the headers",
2747 even if body_only or headers_only was set. It now gives an appropriate
2750 PH/09 Applied a patch from the Sieve maintainer which:
2752 o fixes some comments
2753 o adds the (disabled) notify extension core
2754 o adds some debug output for the result of if/elsif tests
2755 o points to the current vacation draft in the documentation
2756 and documents the missing references header update
2760 o fixes a bug in processing the envelope test (when testing
2761 multiple envelope elements, the last element determinted the
2764 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2765 Electronic Mail") by including:
2767 Auto-submitted: auto-generated
2769 in the messages that it generates (bounce messages and others, such as
2770 warnings). In the case of bounce messages for non-SMTP mesages, there was
2771 also a typo: it was using "Auto_submitted" (underscore instead of
2772 hyphen). Since every message generated by Exim is necessarily in response
2773 to another message, thes have all been changed to:
2775 Auto-Submitted: auto-replied
2777 in accordance with these statements in the RFC:
2779 The auto-replied keyword:
2781 - SHOULD be used on messages sent in direct response to another
2782 message by an automatic process,
2784 - MUST NOT be used on manually-generated messages,
2786 - MAY be used on Delivery Status Notifications (DSNs) and Message
2787 Disposition Notifications (MDNs),
2789 - MUST NOT be used on messages generated by automatic or periodic
2790 processes, except for messages which are automatic responses to
2793 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2794 to the default Received: header definition.
2796 PH/12 Added log selector acl_warn_skipped (default on).
2798 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2799 variables because (a) they are in the wrong storage pool and (b) even if
2800 they were copied, it wouldn't work properly because of the caching.
2802 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2803 checking when decoding. Apparently there are clients that generate
2804 overlong encoded strings. Why am I not surprised?
2806 PH/15 If the first argument of "${if match_address" was not empty, but did not
2807 contain an "@" character, Exim crashed. Now it writes a panic log message
2808 and treats the condition as false.
2810 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2812 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2813 extension "envelope-auth". The code is finished and in agreement with
2814 other implementations, but there is no documentation so far and in fact,
2815 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2816 not changing the active code.
2818 Print executed "if" and "elsif" statements when debugging is used. This
2819 helps a great deal to understand what a filter does.
2821 Document more things not specified clearly in RFC3028. I had all this
2822 sorted out, when out of a sudden new issues came to my mind. Oops."
2824 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2827 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2828 canonical form (as documented). However, after a host name lookup from
2829 the IP address, check_host() was doing a simple string comparison with
2830 addresses acquired from the DNS when checking that the found name did
2831 have the original IP as one of its addresses. Since any found IPv6
2832 addresses are likely to be in abbreviated form, the comparison could
2833 fail. Luckily, there already exists a function for doing the comparison
2834 by converting both addresses to binary, so now that is used instead of
2835 the text comparison.
2837 PH/20 There was another similar case to PH/19, when a complete host name was
2838 given in a host list; looking up its IP address could give an abbreviated
2839 form, whereas the current host's name might or might not be abbreviated.
2840 The same fix has been applied.
2846 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2847 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2850 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2851 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2853 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2855 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2856 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2857 Also, for "Auto-Submitted", treat anything other than "no" as
2858 non-personal, in accordance with RFC 3834. (Previously it treated
2859 anything starting "auto-" as non-personal.)
2861 TF/01 The control=submission/name=... option had a problem with syntax
2862 errors if the name included a slash character. The /name= option
2863 now slurps the rest of the string, so it can include any characters
2864 but it must come last in the list of options (after /sender_retain
2867 PH/05 Some modifications to the interface to the fake nameserver for the new
2875 TK/01 Added the "success_on_redirect" address verification option. See
2876 NewStuff for rationale and an example.
2878 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2880 PH/02 Patch to exigrep to allow it to work on syslog lines.
2882 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2883 fread() to skip over the body file's header line, because in Cygwin the
2884 header line is locked and is inaccessible.
2886 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2887 co-exist for some time) to make it clear that it is the Exim ID that is
2888 referenced, not the Message-ID: header line.
2890 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2891 string_format() function, because snprintf() does not exist on all
2894 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2895 ${stat: expansion item.
2897 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2898 protocol synchronization error", to keep the pedants happy.
2900 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2901 well as for IRIX systems, when gcc is being used. See the host.c source
2904 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2906 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2909 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2910 transport and to the smtp transport in LMTP mode.
2912 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2914 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2915 run a filter in a subprocess. This could lead to confusion in subsequent
2916 lookups in the parent process. There should also be a search_tidyup() at
2917 the end of the subprocess.
2919 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2920 only if the host matched helo_try_verify_hosts, which caused the
2921 verification to occur when the EHLO/HELO command was issued. The ACL just
2922 tested the remembered result. Now, if a previous verification attempt has
2923 not happened, "verify = helo" does it there and then.
2925 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2927 TK/03 Fix log output including CR from clamd.
2929 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2930 exist provoked a memory error which could cause a segfault.
2932 PH/15 Installed PCRE 6.2
2934 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2936 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2937 of the problem. Specifically, suggested +O2 rather than +O1 for the
2940 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2942 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2943 record, the port was not correctly logged when the outgoing_port log
2944 selector was set (it logged the transort's default port).
2946 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2947 fallback_hosts, and "hosts" in the smtp transport.
2949 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2950 host errors such as "Connection refused".
2952 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2953 authentication with radiusclient 0.4.9:
2955 - Error returned from rc_read_config was caught wrongly
2956 - Username/password not passed on to radius server due to wrong length.
2958 The presumption is that some radiusclient API changes for 4.51/PH/17
2959 were not taken care of correctly. The code is still untested by me (my
2960 Linux distribution still has 0.3.2 of radiusclient), but it was
2961 contributed by a Radius user.
2963 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2964 expanding the "port" option of the smtp transport.
2966 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2967 while reading a MIME header. Thanks to Tom Hughes for a patch.
2969 PH/24 Include config.h inside local_scan.h so that configuration settings are
2972 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2973 that in an ACL for RCPT (for example), you can examine exactly what was
2976 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2977 commands, but it was not correctly comparing the address with the actual
2978 client host address. Thus, it would show the EHLO address in Received:
2979 header lines when this was not necessary.
2981 PH/27 Added the % operator to ${eval:}.
2983 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2984 it should be ignoring failures (because with -C, for example, it has lost
2985 privilege). It wasn't ignoring creation failures other than "already
2988 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2991 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2992 because an input line was too long, either on its own, or by virtue of
2993 too many continuations, the temporary file was not being removed, and the
2994 return code was incorrect.
2996 PH/31 Missing "BOOL" in function definition in filtertest.c.
2998 PH/32 Applied Sieve patches from the maintainer.
3000 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3002 PH/33 Added "verify = not_blind".
3004 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3005 Local/Makefile (with some defaults set). These are used in built scripts
3006 such as exicyclog, but they have never been used in the exim_install
3007 script (though there are many overriding facilities there). I have
3008 arranged that the exim_install script now takes note of these two
3011 PH/35 Installed configuration files for Dragonfly.
3013 PH/36 When a locally submitted message by a trusted user did not contain a
3014 From: header, and the sender address was obtained from -f or from an SMTP
3015 MAIL command, and the trusted user did not use -F to supply a sender
3016 name, $originator_name was incorrectly used when constructing a From:
3017 header. Furthermore, $originator_name was used for submission mode
3018 messages from external hosts without From: headers in a similar way,
3019 which is clearly wrong.
3021 PH/37 Added control=suppress_local_fixups.
3023 PH/38 When log_selector = +received_sender was set, and the addition of the
3024 sender made the log line's construction buffer exactly full, or one byte
3025 less than full, an overflow happened when the terminating "\n" was
3028 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3029 when the result of a list match is failure because a DNS lookup failed.
3031 PH/40 RM_COMMAND is now used in the building process.
3033 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3034 the "build-* directories that it finds.
3036 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3037 address in a domain literal was a prefix of an interface address.
3039 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3040 when verifying a sender address, unless rewrite_headers is false.
3042 PH/44 Wrote a long comment about why errors_to addresses are verified as
3043 recipients, not senders.
3045 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3046 the ratelimit ACL was added.
3048 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3050 PH/46 Added extra information about PostgreSQL errors to the error string.
3052 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3053 avoiding the need to install special zones in a real server. This is
3054 backwards compatible; if it can't find the fake resolver, it drops back.
3055 Thus, both old and new test suites can be run.
3057 TF/02 Added util/ratelimit.pl
3059 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3060 clock is set back in time.
3062 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3063 Candler <B.Candler@pobox.com>.
3065 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3066 OK for addresses that are the result of redirections.
3068 PH/48 A number of further additions for the benefit of the new test suite,
3069 including a fake gethostbyname() that interfaces to the fake DNS resolver
3072 TF/06 The fix for widen_domains has also been applied to qualify_single and
3073 search_parents which are the other dnslookup options that can cause
3076 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3079 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3081 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3082 service" if it encountered an MX record with an empty target hostname.
3083 The message is now "an MX or SRV record indicated no SMTP service".
3085 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3086 if the DNS of the sending site is misconfigured. This is quite a
3087 common situation. This change restores the behaviour of treating a
3088 helo verification defer as a failure.
3090 PH/51 If self=fail was set on a router, the bounce message did not include the
3091 actual error message.
3097 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3099 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3100 command itself ended in error, the underlying message about the transport
3101 filter timeout was being overwritten with the pipe command error. Now the
3102 underlying error message should be appended to the second error message.
3104 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3106 PH/02 Reduce the amount of output that "make" produces by default. Full output
3107 can still be requested.
3109 PH/03 The warning log line about a condition test deferring for a "warn" verb
3110 was being output only once per connection, rather than after each
3111 occurrence (because it was using the same function as for successful
3112 "warn" verbs). This seems wrong, so I have changed it.
3114 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3115 it should not have, which might have caused a crash in the right
3116 circumstances, but probably never did.
3118 PH/04 Installed a modified version of Tony Finch's patch to make submission
3119 mode fix the return path as well as the Sender: header line, and to
3120 add a /name= option so that you can make the user's friendly name appear
3123 TF/03 Added the control = fakedefer ACL modifier.
3125 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3126 Mark Lowes for thorough testing.
3128 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3130 TK/03 Merged latest SRS patch from Miles Wilton.
3132 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3133 with the definition in sysexits.h (which is #included earlier).
3134 Fortunately, Exim does not actually use EX_OK. The code used to try to
3135 preserve the sysexits.h value, by assumimg that macro definitions were
3136 scanned for macro replacements. I have been disabused of this notion,
3137 so now the code just undefines EX_OK before #including unistd.h.
3139 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3140 in the smtp transport. When a block could not be written in a single
3141 write() function, the timeout was being re-applied to each part-write.
3142 This seems wrong - if the receiver was accepting one byte at a time it
3143 would take for ever. The timeout is now adjusted when this happens. It
3144 doesn't have to be particularly precise.
3146 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3147 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3149 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3150 without a domain if the check to <postmaster@domain> fails.
3152 SC/01 Eximstats: added -xls and the ability to specify output files
3153 (patch written by Frank Heydlauf).
3155 SC/02 Eximstats: use FileHandles for outputing results.
3157 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3159 SC/04 Eximstats: fixed display of large numbers with -nvr option
3161 SC/05 Eximstats: fixed merging of reports with empty tables.
3163 SC/06 Eximstats: added the -include_original_destination flag
3165 SC/07 Eximstats: removed tabs and trailing whitespace.
3167 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3169 TK/06 MBOX spool code: Add real "From " MBOX separator line
3170 so the .eml file is really in mbox format (even though
3171 most programs do not really care). Patch from Alex Miller.
3173 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3174 The latter is generated from $received_to and is only set if the
3175 message has one envelope recipient. SA can use these headers,
3176 obviously out-of-the-box. Patch from Alex Miller.
3178 PH/08 The ${def test on a variable was returning false if the variable's
3179 value was "0", contrary to what the specification has always said!
3180 The result should be true unless the variable is empty.
3182 PH/09 The syntax error of a character other than { following "${if
3183 def:variable_name" (after optional whitespace) was not being diagnosed.
3184 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3185 accidental colon was present, for example, could give incorrect results.
3187 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3188 result is used (not including appendfile, where other changes are about
3191 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3192 This involved changing a lot of size variables from int to off_t. It
3193 should work with maildirs and everything.
3195 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3196 spamd dying while we are connected to it.
3198 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3201 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3202 function for BDB 4.3.
3204 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3206 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3207 the sources for PCRE have been re-arranged and more files are now
3210 PH/15 The code I had for printing potentially long long variables in PH/11
3211 above was not the best (it lost precision). The length of off_t variables
3212 is now inspected at build time, and an appropriate printing format (%ld
3213 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3214 to be "long long int" or "long int". This is needed for the internal
3215 formatting function string_vformat().
3217 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3218 the configuration file to be ":syslog", then the script "guesses" where
3219 the logs files are, rather than using the compiled in default. In our
3220 case the guess is not the same as the compiled default, so the script
3221 suddenly stopped working when I started to use syslog. The patch checks
3222 to see if log_file_path is "". If so, it attempts to read it from exim
3223 with no configuration file to get the compiled in version, before it
3224 falls back to the previous guessing code."
3226 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3227 implementing BATV in an Exim configuration. See NewStuff for the gory
3230 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3231 Makefile that are specific to HP-UX.
3233 PH/18 If the "use_postmaster" option was set for a recipient callout together
3234 with the "random" option, the postmaster address was used as the MAIL
3235 FROM address for the random test, but not for the subsequent recipient
3236 test. It is now used for both.
3238 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3239 patch removes a few documentation additions to RFC 3028, because the
3240 latest draft now contains them. It adds the new en;ascii-case comparator
3241 and a new error check for 8bit text in MIME parts. Comparator and
3242 require names are now matched exactly. I enabled the subaddress
3243 extension, but it is not well tested yet (read: it works for me)."
3245 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3246 rework some of the code of TK/09 above to avoid the hardwired use of
3247 "%lld" and "long long". Replaced the call to snprintf() with a call to
3250 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3251 records point to non-existent hosts", "retry timeout exceeded", and
3252 "retry time not reached for any host after a long failure period".
3254 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3255 experimental DomainKeys support:
3257 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3258 (2) On an error such as an illegally used "control", the wrong name for
3259 the control was given.
3261 These problems did NOT occur unless DomainKeys support was compiled.
3263 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3265 PH/24 Added ${if match_ip condition.
3267 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3268 databases so that it will be absolutely obvious if a crash occurs in the
3269 DB library. This is a regular occurrence (often caused by mis-matched
3272 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3273 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3274 picked up on a user's system that detects such things. There doesn't seem
3275 to be a gcc warning option for this - only an attribute that has to be
3276 put on the function's prototype. It seems that in Fedora Core 4 they have
3277 set this on a number of new functions. No doubt there will be more in due
3280 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3281 specify a transport. However, if an address that was verified by such a
3282 router was the subject of a callout, Exim crashed because it tried to
3283 read the rcpt_include_affixes from the non-existent transport. Now it
3284 just assumes that the setting of that option is false. This bug was
3285 introduced by 4.51/PH/31.
3287 PH/28 Changed -d+all to exclude +memory, because that information is very
3288 rarely of interest, but it makes the output a lot bigger. People tend to
3289 do -d+all out of habit.
3291 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3292 code in os-type was giving problems when libc.so lives in lib64, like on
3295 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3296 aren't the modern standard, and it seems that some systems' include files
3297 don't always have them. Exim was already checking for some of the newer
3298 ones like T_AAAA, and defining it itself. I've added checks for all the
3299 record types that Exim uses.
3301 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3302 not automatically generating a new one, as it is supposed to. This
3303 prevented TLS from working. If the file did exist, but contained invalid
3304 data, a new version was generated, as expected. It was only the case of a
3305 non-existent file that was broken.
3307 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3308 with a change in libdomainkeys > 0.64.
3310 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3311 from DNS. If the selector record carries the flag, it now has
3312 precedence over the domain-wide flag.
3314 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3316 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3317 the use of a port name that isn't defined in /etc/services) occurred, the
3318 message was deferred as in a normal delivery, and thus remained on the
3319 spool, instead of being failed because of the mua_wrapper setting. This
3320 is now fixed, and I tidied up some of the mua_wrapper messages at the
3323 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3324 the messages in a hash of arrays rather than using individual hashes.
3325 This is a bit cleaner and results in dramatic memory savings, albeit
3326 at a slight CPU cost.
3328 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3329 as requested by Marc Sherman.
3331 SC/10 Eximstats: added histograms for user specified patterns as requested
3334 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3336 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3337 fopen() in the content-scanning modules that did not already have it.
3343 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3344 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3346 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3348 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3350 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3351 to test Sieve filters that use "vacation".
3353 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3354 that changes the way the GnuTLS parameters are stored in the cache file.
3355 The new format can be generated externally. For backward compatibility,
3356 if the data in the cache doesn't make sense, Exim assumes it has read an
3357 old-format file, and it generates new data and writes a new file. This
3358 means that you can't go back to an older release without removing the
3361 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3362 work if there are any delivery delays because "one_time" forces the
3363 parent to be marked "delivered", so its unseen clone is never tried
3364 again. For this reason, Exim now forbids the simultaneous setting of
3367 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3368 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3369 entry if you want to know the details. The fix, however, neglected to
3370 consider the case when local delivery batching is involved. The test for
3371 "previously delivered" was not happening when checking to see if an
3372 address could be batched with a previous (undelivered) one; under
3373 certain circumstances this could lead to multiple deliveries to the same
3376 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3377 in its include files, and this causes problems building Exim.
3379 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3380 header_syntax) but Exim was just ignoring anything given after a slash.
3381 In particular, this caused confusion with an attempt to use "verify =
3382 reverse_host_lookup/defer_ok". An error is now given when options are
3383 supplied for verify items that do not have them. (Maybe reverse_host_
3384 lookup should have a defer_ok option, but that's a different point.)
3386 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3387 defined by RFC 821) to 2048, because there were problems with some AUTH
3388 commands, and RFC 1869 says the size should be increased for extended
3389 SMTP commands that take arguments.
3391 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3394 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3395 "unknown" error; now it says that the functionality isn't in the binary.
3397 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3398 an address' error message when a string expansion fails (syntax or
3399 whatever). Otherwise the password may appear in the log. Following change
3400 PH/42 below, there is no longer a chance of it appearing in a bounce
3403 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3405 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3406 of its addresses was ever tried. (Bugzilla bug #2.)
3408 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3409 the result incorrectly in the debug output. (It correctly added a newline
3410 to what was transported.)
3412 TF/01 Added $received_time.
3414 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3415 commented out examples of how to interface to a virus scanner and to
3416 SpamAssassin. Also added commented examples of av_scanner and
3417 spamd_address settings.
3419 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3420 and controls are allowed in which ACLs. There were a couple of minor
3421 errors. Some of the entries in the conditions table (which is a table of
3422 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3423 negation of where the condition IS allowed.
3425 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3427 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3428 header file does not have a version number, so I've had to invent a new
3429 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3430 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3431 radiusclient), but it was contributed by a Radius user.
3433 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3434 files or directories, for OpenSSL.
3436 PH/19 When an Exim process that is running as root has to create an Exim log
3437 file, it does so in a subprocess that runs as exim:exim so as to get the
3438 ownership right at creation (otherwise, other Exim processes might see
3439 the file with the wrong ownership). There was no test for failure of this
3440 fork() call, which would lead to the process getting stuck as it waited
3441 for a non-existent subprocess. Forks do occasionally fail when resources
3442 run out. I reviewed all the other calls to fork(); they all seem to check
3445 PH/20 When checking for unexpected SMTP input at connect time (before writing
3446 the banner), Exim was not dealing correctly with a non-positive return
3447 from the read() function. If the client had disconnected by this time,
3448 the result was a log entry for a synchronization error with an empty
3449 string after "input=" when read() returned zero. If read() returned -1
3450 (an event I could not check), uninitialized data bytes were printed.
3451 There were reports of junk text (parts of files, etc) appearing after
3454 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3456 PH/22 Added support for macro redefinition, and (re)definition in between
3457 driver and ACL definitions.
3459 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3460 forgetting to use the resulting value; it was using the unexpanded value.
3462 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3463 hadn't been configured. The fix is from Juergen Kreileder, who
3464 understands it better than I do:
3466 "Here's what I see happening with three configured cyrus_sasl
3467 authenticators configured (plain, login, cram-md5):
3469 On startup auth_cyrus_sasl_init() gets called for each of these.
3470 This means three calls to sasl_listmech() without a specified mech_list.
3471 => SASL tests which mechs of all available mechs actually work
3472 => three warnings about OTP not working
3473 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3475 With the patch, sasl_listmech() also gets called three times. But now
3476 SASL's mech_list option is set to the server_mech specified in the the
3477 authenticator. Or in other words, the answer from sasl_listmech()
3478 gets limited to just the mech you're testing for (which is different
3480 => the return list contains just 'plain' or 'login', 'cram-md5' or
3481 nothing depending on the value of ob->server_mech.
3483 I've just tested the patch: Authentication still works fine,
3484 unavailable mechs specified in the exim configuration are still
3485 caught, and the auth.log warnings about OTP are gone."
3487 PH/25 When debugging is enabled, the contents of the command line are added
3488 to the debugging output, even when log_selector=+arguments is not
3491 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3492 answer is "GNU", and only if the return is "GNU/something" is the answer
3495 PH/27 $acl_verify_message is now set immediately after the failure of a
3496 verification in an ACL, and so is available in subsequent modifiers. In
3497 particular, the message can be preserved by coding like this:
3499 warn !verify = sender
3500 set acl_m0 = $acl_verify_message
3502 Previously, $acl_verify_message was set only while expanding "message"
3503 and "log_message" when a very denied access.
3505 PH/28 Modified OS/os.c-Linux with
3507 -#ifndef OS_LOAD_AVERAGE
3508 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3510 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3511 nomenclature these days.)
3513 PH/29 Installed patch from the Sieve maintainer that adds the options
3514 sieve_useraddress and sieve_subaddress to the redirect router.
3516 PH/30 In these circumstances:
3517 . Two addresses routed to the same list of hosts;
3518 . First host does not offer TLS;
3519 . First host accepts first address;
3520 . First host gives temporary error to second address;
3521 . Second host offers TLS and a TLS session is established;
3522 . Second host accepts second address.
3523 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3524 and peerdn, if requested) that were in fact used only for the second
3527 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3528 attention to any local part prefix or suffix that was matched by the
3529 router that accepted the address. It now behaves in the same way as it
3530 does for delivery: the affixes are removed from the local part unless
3531 rcpt_include_affixes is set on the transport.
3533 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3534 timeout during the DATA phase of an incoming message.
3536 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3537 applied a patch sanctioned by the Sieve maintainer.
3539 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3540 the uid or gid is negative. A case of a negative gid caused this to be
3541 noticed. The fix allows for either to be negative.
3543 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3544 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3547 PH/36 The change PH/12 above was broken. Fixed it.
3549 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3550 the grounds that routing the same address twice would always produce the
3551 same answer. This might have been true once, but it is certainly no
3552 longer true now. Routing a child address may depend on the previous
3553 routing that produced that child. Some complicated redirection strategies
3554 went wrong when messages had multiple recipients, and made Exim's
3555 behaviour dependent on the order in which the addresses were given.
3557 I have moved the duplicate checking until after the routing is complete.
3558 Exim scans the addresses that are assigned to local and remote
3559 transports, and removes any duplicates. This means that more work will be
3560 done, as duplicates will always all be routed, but duplicates are
3561 presumably rare, so I don't expect this is of any significance.
3563 For deliveries to pipes, files, and autoreplies, the duplicate checking
3564 still happens during the routing process, since they are not going to be
3567 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3568 It corrects a timeout issue with spamd. This is Ian's comment: "The
3569 background is that sometimes spamd either never reads data from a
3570 connection it has accepted, or it never writes response data. The exiscan
3571 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3572 blindly assumes that writes won't block so it may never time out."
3574 PH/39 Allow G after quota size as well as K and M.
3576 PH/40 The value set for $authenticated_id in an authenticator may not contain
3577 binary zeroes or newlines because the value is written to log lines and
3578 to spool files. There was no check on this. Now the value is run through
3579 the string_printing() function so that such characters are converted to
3580 printable escape sequences.
3582 PH/41 $message_linecount is a new variable that contains the total number of
3583 lines in the message. Compare $body_linecount, which is the count for the
3586 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3587 bounce and delay warning messages, except in certain special cases, which
3590 (a) An SMTP error message from a remote host;
3591 (b) A message specified in a :fail: redirection;
3592 (c) A message specified in a "fail" command in a system filter;
3593 (d) A message specified in a FAIL return from the queryprogram router;
3594 (e) A message specified by the cannot_route_message router option.
3596 In these cases only, Exim does include the error details in bounce and
3597 warning messages. There are also a few cases where bland messages such
3598 as "unrouteable address" or "local delivery error" are given.
3600 PH/43 $value is now also set for the "else" part of a ${run expansion.
3602 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3603 being worked on, but at least Exim now implements the latest version to
3606 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3607 process to complete was treated as a delivery failure, a timeout while
3608 writing the message to the pipe was logged, but erroneously treated as a
3609 successful delivery. Such timeouts include transport filter timeouts. For
3610 consistency with the overall process timeout, these timeouts are now
3611 treated as errors, giving rise to delivery failures by default. However,
3612 there is now a new Boolean option for the pipe transport called
3613 timeout_defer, which, if set TRUE, converts the failures into defers for
3614 both kinds of timeout. A transport filter timeout is now identified in
3617 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3618 systems where "make" and "gmake" are different, calling "gmake" at top
3619 level broke things. I've arranged for the value of $(MAKE) to be passed
3620 from the Makefile to this script so that it can call the same version of
3624 A note about Exim versions 4.44 and 4.50
3625 ----------------------------------------
3627 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3628 changes of various kinds. As a consequence, a big documentation update was
3629 needed. This delayed the release for rather longer than seemed good, especially
3630 in the light of a couple of (minor) security issues. Therefore, the changes
3631 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3632 release. So 4.44 and 4.50 are in effect two different branches that both start
3635 I have left the 4.50 change log unchanged; it contains all the changes since
3636 4.43. The change log for 4.44 is below; many of its items are identical to
3637 those for 4.50. This seems to be the most sensible way to preserve the
3638 historical information.
3644 1. Minor wording change to the doc/README.SIEVE file.
3646 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3647 computation of the current number of files was incorrect.
3649 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3650 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3653 4. Give more explanation in the error message when the command for a transport
3654 filter fails to execute.
3656 5. There are several places where Exim runs a non-Exim command in a
3657 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3658 was being done only for the command run by the queryprogram router. It is
3659 now done for all such subprocesses. The other cases are: ${run, transport
3660 filters, and the commands run by the lmtp and pipe transports.
3662 6. Added CONFIGURE_GROUP build-time option.
3664 7. Some older OS have a limit of 256 on the maximum number of file
3665 descriptors. Exim was using setrlimit() to set 1000 as a large value
3666 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3667 systems. I've change it so that if it can't get 1000, it tries for 256.
3669 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3670 was an oversight, and furthermore, ever since the addition of extra
3671 controls (e.g. 4.43/32), the checks on when to allow different forms of
3672 "control" were broken. There should now be diagnostics for all cases when a
3673 control that does not make sense is encountered.
3675 9. Added the /retain_sender option to "control=submission".
3677 10. $recipients is now available in the predata ACL (oversight).
3679 11. Tidy the search cache before the fork to do a delivery from a message
3680 received from the command line. Otherwise the child will trigger a lookup
3681 failure and thereby defer the delivery if it tries to use (for example) a
3682 cached ldap connection that the parent has called unbind on.
3684 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3685 of $address_data from the recipient verification was clobbered by the
3686 sender verification.
3688 13. The value of address_data from a sender verification is now available in
3689 $sender_address_data in subsequent conditions in the ACL statement.
3691 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3693 15. Added a new option "connect=<time>" to callout options, to set a different
3696 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3697 was its contents. (It was OK if the option was not defined at all.)
3699 17. A "Completed" log line is now written for messages that are removed from
3700 the spool by the -Mrm option.
3702 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3703 information about exactly what failed.
3705 19. Added -dd to debug only the daemon process.
3707 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3708 handles timeouts, both on the server side and network timeouts. Renamed the
3709 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3711 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3712 It is now set to "smtps".
3714 22. $host_address is now set to the target address during the checking of
3715 ignore_target_hosts.
3717 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3718 being passed; this would have caused $sender_host_name to have been used if
3719 matching the list had actually called for a host name (not very likely,
3720 since this list is usually IP addresses). A host name is now passed as
3723 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3724 code that specifies a non-restarting handler (typically sigaction() in
3725 modern systems) in an attempt to fix a rare and obscure crash bug.
3727 25. Narrowed the window for a race in the daemon that could cause it to ignore
3728 SIGCHLD signals. This is not a major problem, because they are used only to
3729 wake it up if nothing else does.
3731 26. A malformed maildirsize file could cause Exim to calculate negative values
3732 for the mailbox size or file count. Odd effects could occur as a result.
3733 The maildirsize information is now recalculated if the size or filecount
3736 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3737 support for a long time. Removed HAVE_SYS_VFS_H.
3739 28. Installed the latest version of exipick from John Jetmore.
3741 29. In an address list, if the pattern was not a regular expression, an empty
3742 subject address (from a bounce message) matched only if the pattern was an
3743 empty string. Non-empty patterns were not even tested. This was the wrong
3744 because it is perfectly reasonable to use an empty address as part of a
3745 database query. An empty address is now tested by patterns that are
3746 lookups. However, all the other forms of pattern expect the subject to
3747 contain a local part and a domain, and therefore, for them, an empty
3748 address still always fails if the pattern is not itself empty.
3750 30. Exim went into a mad DNS loop when attempting to do a callout where the
3751 host was specified on an smtp transport, and looking it up yielded more
3752 than one IP address.
3754 31. Re-factored the code for checking spool and log partition space into a
3755 function that finds that data and another that does the check. The former
3756 is then used to implement four new variables: $spool_space, $log_space,
3757 $spool_inodes, and $log_inodes.
3759 32. The RFC2047 encoding function was originally intended for short strings
3760 such as real names; it was not keeping to the 75-character limit for
3761 encoded words that the RFC imposes. It now respects the limit, and
3762 generates multiple encoded words if necessary. To be on the safe side, I
3763 have increased the buffer size for the ${rfc2047: expansion operator from
3766 33. It is now permitted to omit both strings after an "if" condition; if the
3767 condition is true, the result is "true". As before, when the second string
3768 is omitted, a false condition yields an empty string. This makes it less
3769 cumbersome to write custom ACL and router conditions.
3771 34. Failure to deliver a bounce message always caused it to be frozen, even if
3772 there was an errors_to setting on the router. The errors_to setting is now
3775 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3776 canonical form (fully expanded) before being placed in
3777 $sender_host_address.
3779 36. The table in the code that translates DNS record types into text (T_A to
3780 "A" for instance) was missing entries for NS and CNAME. It is just possible
3781 that this could have caused confusion if both these types were looked up
3782 for the same domain, because the text type is used as part of Exim's
3783 per-process caching. But the chance of anyone hitting this buglet seems
3786 37. The dnsdb lookup has been extended in a number of ways.
3788 (1) There is a new type, "zns", which walks up the domain tree until it
3789 finds some nameserver records. It should be used with care.
3791 (2) There is a new type, "mxh", which is like "mx" except that it returns
3792 just the host names, not the priorities.
3794 (3) It is now possible to give a list of domains (or IP addresses) to be
3795 looked up. The behaviour when one of the lookups defers can be
3796 controlled by a keyword.
3798 (4) It is now possible to specify the separator character for use when
3799 multiple records are returned.
3801 38. The dnslists ACL condition has been extended: it is now possible to supply
3802 a list of IP addresses and/or domains to be looked up in a particular DNS
3805 39. Added log_selector=+queue_time_overall.
3807 40. When running the queue in the test harness, wait just a tad after forking a
3808 delivery process, to get repeatability of debugging output.
3810 41. Include certificate and key file names in error message when GnuTLS fails
3811 to set them up, because the GnuTLS error message doesn't include the name
3812 of the failing file when there is a problem reading it.
3814 42. Allow both -bf and -bF in the same test run.
3816 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3818 44. The "Exiscan patch" is now merged into the mainline Exim source.
3820 45. Sometimes the final signoff response after QUIT could fail to get
3821 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3822 < 0 before doing a fflush(). This bug looks as though it goes back to the
3823 introduction of TLS in release 3.20, but "sometimes" must have been rare
3824 because the tests only now provoked it.
3826 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3827 (this can affect the format of dates).
3829 47. exim_tidydb, when checking for the continued existence of a message for
3830 which it has found a message-specific retry record, was not finding
3831 messages that were in split spool directories. Consequently, it was
3832 deleting retry records that should have stayed in existence.
3834 48. Steve fixed some bugs in eximstats.
3836 49. The SPA authentication driver was not abandoning authentication and moving
3837 on to the next authenticator when an expansion was forced to fail,
3838 contradicting the general specification for all authenticators. Instead it
3839 was generating a temporary error. It now behaves as specified.
3841 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3842 (the order specifies the preference for clients). The order is now AES256,
3843 AES128, 3DES, ARCFOUR128.
3845 51. Small patch to Sieve code - explicitly set From: when generating an
3848 52. Exim crashed if a remote delivery caused a very long error message to be
3849 recorded - for instance if somebody sent an entire SpamAssassin report back
3850 as a large number of 550 error lines. This bug was coincidentally fixed by
3851 increasing the size of one of Exim's internal buffers (big_buffer) that
3852 happened as part of the Exiscan merge. However, to be on the safe side, I
3853 have made the code more robust (and fixed the comments that describe what
3856 53. Now that there can be additional text after "Completed" in log lines (if
3857 the queue_time_overall log selector is set), a one-byte patch to exigrep
3858 was needed to allow it to recognize "Completed" as not the last thing in
3861 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3862 patch that reportedly fixes this has been added. I am not expert enough to
3863 create a test for it. This is what the patch creator wrote:
3865 "I found a little strange behaviour of ldap code when working with
3866 Windows 2003 AD Domain, where users was placed in more than one
3867 Organization Units. When I tried to give exim partial DN, the exit code
3868 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3869 But simultaneously result of request was absolutely normal ldap result,
3870 so I produce this patch..."
3872 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3873 REFERENCE, so I have modified the code to exclude the patch when that macro
3876 55. Some experimental protocols are using DNS PTR records for new purposes. The
3877 keys for these records are domain names, not reversed IP addresses. The
3878 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3879 leaves it alone. Component reversal etc. now happens only for IP addresses.
3882 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3884 57. Double the size of the debug message buffer (to 2048) so that more of very
3885 long debug lines gets shown.
3887 58. The exicyclog utility now does better if the number of log files to keep
3888 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3890 59. Two changes related to the smtp_active_hostname option:
3892 (1) $smtp_active_hostname is now available as a variable.
3893 (2) The default for smtp_banner uses $smtp_active_hostname instead
3894 of $primary_hostname.
3896 60. The host_aton() function is supposed to be passed a string that is known
3897 to be a valid IP address. However, in the case of IPv6 addresses, it was
3898 not checking this. This is a hostage to fortune. Exim now panics and dies
3899 if the condition is not met. A case was found where this could be provoked
3900 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3901 components; fortuitously, this particular loophole had already been fixed
3902 by change 4.50/55 above.
3904 If there are any other similar loopholes, the new check in host_aton()
3905 itself should stop them being exploited. The report I received stated that
3906 data on the command line could provoke the exploit when Exim was running as
3907 exim, but did not say which command line option was involved. All I could
3908 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3909 running as the user.
3912 61. There was a buffer overflow vulnerability in the SPA authentication code
3913 (which came originally from the Samba project). I have added a test to the
3914 spa_base64_to_bits() function which I hope fixes it.
3917 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3918 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3920 63. The daemon start-up calls getloadavg() while still root for those OS that
3921 need the first call to be done as root, but it missed one case: when
3922 deliver_queue_load_max is set with deliver_drop_privilege. This is
3923 necessary for the benefit of the queue runner, because there is no re-exec
3924 when deliver_drop_privilege is set.
3926 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3927 This has been fixed.
3929 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3930 was in use, was not putting the data itself into the right store pool;
3931 consequently, it could be overwritten for a subsequent message in the same
3932 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3935 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3937 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3938 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3939 the function were treating the return as a boolean value, which happened to
3940 work because 0=false and not-0=true, but is not correct code.
3942 68. The host_aton() function was not handling scoped IPv6 addresses (those
3943 with, for example, "%eth0" on the end) correctly.
3945 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3946 negated items (that is, ~something) in unsigned ints. Some compilers
3947 apparently mutter when there is no cast.
3949 70. If an address verification called from an ACL failed, and did not produce a
3950 user-specific message (i.e. there was only a "system" message), nothing was
3951 put in $acl_verify_message. In this situation, it now puts the system
3954 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3955 session; change 4.31/43 added the unwanted input to the log line - except
3956 that it did not do this in the start of session case. It now does.
3958 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3959 This is wrong and can cause the other end to generate a synchronization
3960 error if it is another Exim or anything else that does the synchronization
3961 check. A QUIT command is no longer sent after a timeout.
3963 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3964 during host lookups.
3966 74. The defer_ok option of callout verification was not working if it was used
3967 when verifying addresses in header lines, that is, for this case:
3969 verify = header_sender/callout=defer_ok
3971 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3972 those file descriptors could be used for SMTP connections. If anything
3973 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3974 could be sent to the SMTP client, causing chaos. The daemon now opens
3975 stdin, stdout, and stderr to /dev/null when it puts itself into the
3978 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3979 log by default. The user can override this with suitable Perl magic.
3981 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3982 the log message when discard triggers, was not working for the DATA ACL or
3983 for the non-SMTP ACL.
3985 78. Error message wording change in sieve.c.
3987 79. If smtp_accept_max_per_host was set, the number of connections could be
3988 restricted to fewer than expected, because the daemon was trying to set up
3989 a new connection before checking whether the processes handling previous
3990 connections had finished. The check for completed processes is now done
3991 earlier. On busy systems, this bug wouldn't be noticed because something
3992 else would have woken the daemon, and it would have reaped the completed
3995 80. If a message was submitted locally by a user whose login name contained one
3996 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3997 It caused a spool format error. I have fixed the spool reading code. A
3998 related problem was that the "from" clause in the Received: line became
3999 illegal because of the space(s). It is now covered by ${quote_local_part.
4001 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4004 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4006 83. Patch from Sieve maintainer to fix unterminated string problem in
4007 "vacation" handling.
4009 84. Some minor changes to the Linux configuration files to help with other
4010 OS variants using glibc.
4012 85. One more patch for Sieve to update vacation handling to latest spec.
4015 ----------------------------------------------------
4016 See the note above about the 4.44 and 4.50 releases.
4017 ----------------------------------------------------
4023 1. Change 4.43/35 introduced a bug that caused file counts to be
4024 incorrectly computed when quota_filecount was set in an appendfile
4027 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4028 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4031 3. Give more explanation in the error message when the command for a transport
4032 filter fails to execute.
4034 4. There are several places where Exim runs a non-Exim command in a
4035 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4036 was being done only for the command run by the queryprogram router. It is
4037 now done for all such subprocesses. The other cases are: ${run, transport
4038 filters, and the commands run by the lmtp and pipe transports.
4040 5. Some older OS have a limit of 256 on the maximum number of file
4041 descriptors. Exim was using setrlimit() to set 1000 as a large value
4042 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4043 systems. I've change it so that if it can't get 1000, it tries for 256.
4045 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4046 was an oversight, and furthermore, ever since the addition of extra
4047 controls (e.g. 4.43/32), the checks on when to allow different forms of
4048 "control" were broken. There should now be diagnostics for all cases when a
4049 control that does not make sense is encountered.
4051 7. $recipients is now available in the predata ACL (oversight).
4053 8. Tidy the search cache before the fork to do a delivery from a message
4054 received from the command line. Otherwise the child will trigger a lookup
4055 failure and thereby defer the delivery if it tries to use (for example) a
4056 cached ldap connection that the parent has called unbind on.
4058 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4059 of $address_data from the recipient verification was clobbered by the
4060 sender verification.
4062 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4063 was its contents. (It was OK if the option was not defined at all.)
4065 11. A "Completed" log line is now written for messages that are removed from
4066 the spool by the -Mrm option.
4068 12. $host_address is now set to the target address during the checking of
4069 ignore_target_hosts.
4071 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4072 being passed; this would have caused $sender_host_name to have been used if
4073 matching the list had actually called for a host name (not very likely,
4074 since this list is usually IP addresses). A host name is now passed as
4077 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4078 code that specifies a non-restarting handler (typically sigaction() in
4079 modern systems) in an attempt to fix a rare and obscure crash bug.
4081 15. Narrowed the window for a race in the daemon that could cause it to ignore
4082 SIGCHLD signals. This is not a major problem, because they are used only to
4083 wake it up if nothing else does.
4085 16. A malformed maildirsize file could cause Exim to calculate negative values
4086 for the mailbox size or file count. Odd effects could occur as a result.
4087 The maildirsize information is now recalculated if the size or filecount
4090 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4091 support for a long time. Removed HAVE_SYS_VFS_H.
4093 18. Updated exipick to current release from John Jetmore.
4095 19. Allow an empty sender to be matched against a lookup in an address list.
4096 Previously the only cases considered were a regular expression, or an
4099 20. Exim went into a mad DNS lookup loop when doing a callout where the
4100 host was specified on the transport, if the DNS lookup yielded more than
4103 21. The RFC2047 encoding function was originally intended for short strings
4104 such as real names; it was not keeping to the 75-character limit for
4105 encoded words that the RFC imposes. It now respects the limit, and
4106 generates multiple encoded words if necessary. To be on the safe side, I
4107 have increased the buffer size for the ${rfc2047: expansion operator from
4110 22. Failure to deliver a bounce message always caused it to be frozen, even if
4111 there was an errors_to setting on the router. The errors_to setting is now
4114 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4115 canonical form (fully expanded) before being placed in
4116 $sender_host_address.
4118 24. Updated eximstats to version 1.33
4120 25. Include certificate and key file names in error message when GnuTLS fails
4121 to set them up, because the GnuTLS error message doesn't include the name
4122 of the failing file when there is a problem reading it.
4124 26. Expand error message when OpenSSL has problems setting up cert/key files.
4127 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4128 (this can affect the format of dates).
4130 28. exim_tidydb, when checking for the continued existence of a message for
4131 which it has found a message-specific retry record, was not finding
4132 messages that were in split spool directories. Consequently, it was
4133 deleting retry records that should have stayed in existence.
4135 29. eximstats updated to version 1.35
4136 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4137 1.35 - bugfix such that pie charts by volume are generated correctly
4139 30. The SPA authentication driver was not abandoning authentication and moving
4140 on to the next authenticator when an expansion was forced to fail,
4141 contradicting the general specification for all authenticators. Instead it
4142 was generating a temporary error. It now behaves as specified.
4144 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4145 (the order specifies the preference for clients). The order is now AES256,
4146 AES128, 3DES, ARCFOUR128.
4148 31. Small patch to Sieve code - explicitly set From: when generating an
4151 32. Exim crashed if a remote delivery caused a very long error message to be
4152 recorded - for instance if somebody sent an entire SpamAssassin report back
4153 as a large number of 550 error lines. This bug was coincidentally fixed by
4154 increasing the size of one of Exim's internal buffers (big_buffer) that
4155 happened as part of the Exiscan merge. However, to be on the safe side, I
4156 have made the code more robust (and fixed the comments that describe what
4159 33. Some experimental protocols are using DNS PTR records for new purposes. The
4160 keys for these records are domain names, not reversed IP addresses. The
4161 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4162 leaves it alone. Component reversal etc. now happens only for IP addresses.
4165 34. The host_aton() function is supposed to be passed a string that is known
4166 to be a valid IP address. However, in the case of IPv6 addresses, it was
4167 not checking this. This is a hostage to fortune. Exim now panics and dies
4168 if the condition is not met. A case was found where this could be provoked
4169 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4170 components; fortuitously, this particular loophole had already been fixed
4171 by change 4.50/55 or 4.44/33 above.
4173 If there are any other similar loopholes, the new check in host_aton()
4174 itself should stop them being exploited. The report I received stated that
4175 data on the command line could provoke the exploit when Exim was running as
4176 exim, but did not say which command line option was involved. All I could
4177 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4178 running as the user.
4181 35. There was a buffer overflow vulnerability in the SPA authentication code
4182 (which came originally from the Samba project). I have added a test to the
4183 spa_base64_to_bits() function which I hope fixes it.
4186 36. The daemon start-up calls getloadavg() while still root for those OS that
4187 need the first call to be done as root, but it missed one case: when
4188 deliver_queue_load_max is set with deliver_drop_privilege. This is
4189 necessary for the benefit of the queue runner, because there is no re-exec
4190 when deliver_drop_privilege is set.
4192 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4193 was in use, was not putting the data itself into the right store pool;
4194 consequently, it could be overwritten for a subsequent message in the same
4195 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4198 38. Sometimes the final signoff response after QUIT could fail to get
4199 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4200 < 0 before doing a fflush(). This bug looks as though it goes back to the
4201 introduction of TLS in release 3.20, but "sometimes" must have been rare
4202 because the tests only now provoked it.
4208 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4209 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4210 Now it returns an integer. A number of calls were still expecting a T/F
4211 return. Fortuitously, in all cases, the tests worked in OK situations,
4212 which is the norm. However, things would have gone wrong on any write
4213 failures on the smtp file descriptor. This function is used when sending
4214 messages over SMTP and also when doing verify callouts.
4216 2. When Exim is called to do synchronous delivery of a locally submitted
4217 message (the -odf or -odi options), it no longer closes stderr before doing
4220 3. Implemented the mua_wrapper option.
4222 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4224 5. Implemented the functions header_remove(), header_testname(),
4225 header_add_at_position(), and receive_remove_recipient(), and exported them
4228 6. If an ACL "warn" statement specified the addition of headers, Exim already
4229 inserted X-ACL-Warn: at the start if there was no header name. However, it
4230 was not making this test for the second and subsequent header lines if
4231 there were newlines in the string. This meant that an invalid header could
4232 be inserted if Exim was badly configured.
4234 7. Allow an ACL "warn" statement to add header lines at the start or after all
4235 the Received: headers, as well as at the end.
4237 8. Added the rcpt_4xx retry error code.
4239 9. Added postmaster_mailfrom=xxx to callout verification option.
4241 10. Added mailfrom=xxxx to the callout verification option, for verify=
4244 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4245 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4247 12. Inserted some casts to stop certain compilers warning when using pointer
4248 differences as field lengths or precisions in printf-type calls (mostly
4249 affecting debugging statements).
4251 13. Added optional readline() support for -be (dynamically loaded).
4253 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4254 same clock tick as a message's arrival, so that its received time was the
4255 same as the "first fail" time on the retry record, and that message
4256 remained on the queue past the ultimate address timeout, every queue runner
4257 would try a delivery (because it was past the ultimate address timeout) but
4258 after another failure, the ultimate address timeout, which should have then
4259 bounced the address, did not kick in. This was a "< instead of <=" error;
4260 in most cases the first failure would have been in the next clock tick
4261 after the received time, and all would be well.
4263 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4264 being recognized when the domain list was tested by the match_domain
4265 condition in an expansion string.
4267 16. Added the ${str2b64: operator.
4269 17. Exim was always calling setrlimit() to set a large limit for the number of
4270 processes, without checking whether the existing limit was already
4271 adequate. (It did check for the limit on file descriptors.) Furthermore,
4272 errors from getrlimit() and setrlimit() were being ignored. Now they are
4273 logged to the main and panic logs, but Exim does carry on, to try to do its
4274 job under whatever limits there are.
4276 18. Imported PCRE 5.0.
4278 19. Trivial typo in log message " temporarily refused connection" (the leading
4281 20. If the log selector return_path_on_delivery was set and an address was
4282 redirected to /dev/null, the delivery process crashed because it assumed
4283 that a return path would always be set for a "successful" delivery. In this
4284 case, the whole delivery is bypassed as an optimization, and therefore no
4287 21. Internal re-arrangement: the function for sending a challenge and reading
4288 a response while authentication was assuming a zero-terminated challenge
4289 string. It's now changed to take a pointer and a length, to allow for
4290 binary data in such strings.
4292 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4294 23. Exim was not respecting finduser_retries when seeking the login of the
4295 uid under which it was called; it was always trying 10 times. (The default
4296 setting of finduser_retries is zero.) Also, it was sleeping after the final
4297 failure, which is pointless.
4299 24. Implemented tls_on_connect_ports.
4301 25. Implemented acl_smtp_predata.
4303 26. If the domain in control=submission is set empty, Exim assumes that the
4304 authenticated id is a complete email address when it generates From: or
4305 Sender: header lines.
4307 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4308 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4309 chown and chgrp in /bin and hostname in /usr/bin.
4311 28. Exim was keeping the "process log" file open after each use, just as it
4312 does for the main log. This opens the possibility of it remaining open for
4313 long periods when the USR1 signal hits a daemon. Occasional processlog
4314 errors were reported, that could have been caused by this. Anyway, it seems
4315 much more sensible not to leave this file open at all, so that is what now
4318 29. The long-running daemon process does not normally write to the log once it
4319 has entered its main loop, and it closes the log before doing so. This is
4320 so that log files can straightforwardly be renamed and moved. However,
4321 there are a couple of unusual error situations where the daemon does write
4322 log entries, and I had neglected to close the log afterwards.
4324 30. The text of an SMTP error response that was received during a remote
4325 delivery was being truncated at 512 bytes. This is too short for some of
4326 the long messages that one sometimes sees. I've increased the limit to
4329 31. It is now possible to make retry rules that apply only when a message has a
4330 specific sender, in particular, an empty sender.
4332 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4333 it possible to be selective about when SMTP synchronization is enforced.
4335 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4337 32. Implemented hosts_connection_nolog.
4339 33. Added an ACL for QUIT.
4341 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4344 35. Added mailbox_size and mailbox_filecount to appendfile.
4346 36. Added control = no_multiline_responses to ACLs.
4348 37. There was a bug in the logic of the code that waits for the clock to tick
4349 in the case where the clock went backwards by a substantial amount such
4350 that the microsecond fraction of "now" was more than the microsecond
4351 fraction of "then" (but the whole seconds number was less).
4353 38. Added support for the libradius Radius client library this is found on
4354 FreeBSD (previously only the radiusclient library was supported).
4360 1. When certain lookups returned multiple values in the form name=value, the
4361 quoting of the values was not always being done properly. Specifically:
4362 (a) If the value started with a double quote, but contained no whitespace,
4364 (b) If the value contained whitespace other than a space character (i.e.
4365 tabs or newlines or carriage returns) it was not quoted.
4366 This fix has been applied to the mysql and pgsql lookups by writing a
4367 separate quoting function and calling it from the lookup code. The fix
4368 should probably also be applied to nisplus, ibase and oracle lookups, but
4369 since I cannot test any of those, I have not disturbed their existing code.
4371 2. A hit in the callout cache for a specific address caused a log line with no
4372 reason for rejecting RCPT. Now it says "Previous (cached) callout
4373 verification failure".
4375 3. There was an off-by-one bug in the queryprogram router. An over-long
4376 return line was truncated at 256 instead of 255 characters, thereby
4377 overflowing its buffer with the terminating zero. As well as fixing this, I
4378 have increased the buffer size to 1024 (and made a note to document this).
4380 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4381 when Exim is waiting for an SMTP response from a remote server, Exim
4382 restarts its select() call on the socket, thereby resetting its timeout.
4383 This is not a problem when such interrupts are rare. Somebody set up a cron
4384 job to run exiwhat every 2 minutes, which is less than the normal select()
4385 timeout (5 or 10 minutes). This meant that the select() timeout never
4386 kicked in because it was always reset. I have fixed this by comparing the
4387 time when an interrupt arrives with the time at the start of the first call
4388 to select(). If more time than the timeout has elapsed, the interrupt is
4389 treated as a timeout.
4391 5. Some internal re-factoring in preparation for the addition of Sieve
4392 extensions (by MH). In particular, the "personal" test is moved to a
4393 separate function, and given an option for scanning Cc: and Bcc: (which is
4394 not set for Exim filters).
4396 6. When Exim created an email address using the login of the caller as the
4397 local part (e.g. when creating a From: or Sender: header line), it was not
4398 quoting the local part when it contained special characters such as @.
4400 7. Installed new OpenBSD configuration files.
4402 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4403 try to make them clearer.
4405 9. Callout options, other than the timeout value, were being ignored when
4406 verifying sender addresses in header lines. For example, when using
4408 verify = header_sender/callout=no_cache
4410 the cache was (incorrectly) being used.
4412 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4413 only the Cygwin environment.
4415 11. When return_path_on_delivery was set as a log selector, if different remote
4416 addresses in the same message used different return paths and parallel
4417 remote delivery occurred, the wrong values would sometimes be logged.
4418 (Whenever a remote delivery process finished, the return path value from
4419 the most recently started remote delivery process was logged.)
4421 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4422 lines when AUTH and/or TLS are in use. This is the "received protocol"
4423 field. Exim used to use "asmtp" for authenticated SMTP, without any
4424 indication (in the protocol name) for TLS use. Now it follows the RFC and
4425 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4426 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4427 names appear in log lines as well as in Received: header lines.
4429 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4430 extensions, and comparison tests, and to fix some bugs.
4432 14. Changes to the "personal" filter test:
4434 (1) The test was buggy in that it was just doing the equivalent of
4435 "contains" tests on header lines. For example, if a user's address was
4436 anne@some.where, the "personal" test would incorrectly be true for
4438 To: susanne@some.where
4440 This test is now done by extracting each address from the header in turn,
4441 and checking the entire address. Other tests that are part of "personal"
4442 are now done using regular expressions (for example, to check local parts
4443 of addresses in From: header lines).
4445 (2) The list of non-personal local parts in From: addresses has been
4446 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4447 taken from the Sieve specification recommendations.
4449 (3) If the message contains any header line starting with "List-" it is
4450 treated as non-personal.
4452 (4) The test for "circular" in the Subject: header line has been removed
4453 because it now seems ill-conceived.
4455 15. Minor typos in src/EDITME comments corrected.
4457 16. Installed latest exipick from John Jetmore.
4459 17. If headers_add on a router specified a text string that was too long for
4460 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4461 of string_sprintf() is now avoided.
4463 18. $message_body_size was not set (it was always zero) when running the DATA
4464 ACL and the local_scan() function.
4466 19. For the "mail" command in an Exim filter, no default was being set for
4467 the once_repeat time, causing a random time value to be used if "once" was
4468 specified. (If the value happened to be <= 0, no repeat happened.) The
4469 default is now 0s, meaning "never repeat". The "vacation" command was OK
4470 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4471 (I found it when inspecting the code).
4473 20. There is now an overall timeout for performing a callout verification. It
4474 defaults to 4 times the callout timeout, which applies to individual SMTP
4475 commands during the callout. The overall timeout applies when there is more
4476 than one host that can be tried. The timeout is checked before trying the
4477 next host. This prevents very long delays if there are a large number of
4478 hosts and all are timing out (e.g. when the network connections are timing
4479 out). The value of the overall timeout can be changed by specifying an
4480 additional sub-option for "callout", called "maxwait". For example:
4482 verify = sender/callout=5s,maxwait=20s
4484 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4485 to the end before writing, but this should make it even safer).
4487 22. Exim was forgetting that it had advertised PIPELINING for the second and
4488 subsequent messages on an SMTP connection. It was also not resetting its
4489 memory on STARTTLS and an internal HELO.
4491 23. When Exim logs an SMTP synchronization error within a session, it now
4492 records whether PIPELINING has been advertised or not.
4494 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4495 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4496 rather than long int.
4498 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4500 26. Added the never_mail option to autoreply.
4506 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4507 crash if the getsockname() call failed; this can happen if a connection is
4508 closed very soon after it is established. The problem was simply in the
4509 order in which certain operations were done, causing Exim to try to write
4510 to the SMTP stream before it had set up the file descriptor. The bug has
4511 been fixed by making things happen in the correct order.
4517 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4518 before the connection was closed, thus losing the rejection response.
4520 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4521 some early Solaris releases, but causes trouble in current releases where
4522 socklen_t is defined.
4524 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4527 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4530 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4531 that releasing the top of it at the end releases what was used for sub-
4532 expansions (unless the block got too big). However, discard this block if
4533 the first thing is a variable or header, so that we can use its block when
4534 it is dynamic (useful for very large $message_headers, for example).
4536 6. Lookups now cache *every* query, not just the most recent. A new, separate
4537 store pool is used for this. It can be recovered when all lookup caches are
4538 flushed. Lookups now release memory at the end of their result strings.
4539 This has involved some general refactoring of the lookup sources.
4541 7. Some code has been added to the store_xxx() functions to reduce the amount
4542 of flapping under certain conditions.
4544 8. log_incoming_interface used to affect only the <= reception log lines. Now
4545 it causes the local interface and port to be added to several more SMTP log
4546 lines, for example "SMTP connection from", and rejection lines.
4548 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4550 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4552 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4553 could be overwritten at the end of the current message (or the start of a
4554 new message if it was set in a HELO ACL). The value is now preserved for
4555 the duration of the SMTP connection.
4557 12. If a transport had a headers_rewrite setting, and a matching header line
4558 contained an unqualified address, that address was qualified, even if it
4559 did not match any rewriting rules. The underlying bug was that the values
4560 of the flags that permit the existence of unqualified sender and recipient
4561 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4562 non-local messages, and by -bnq for local messages) were not being
4563 preserved with the message after it was received.
4565 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4566 "next input=" as part of the text comprising the host identity instead of
4567 the correct text. The code was using the same buffer for two different
4568 strings. However, depending on which order the printing function evaluated
4569 its arguments, the bug did not always show up. Under Linux, for example, my
4570 test suite worked just fine.
4572 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4573 doesn't work with some older versions of Perl. It has been changed to "my",
4574 which in any case is probably the better facility to use.
4576 15. A really picky compiler found some instances of statements for creating
4577 error messages that either had too many or two few arguments for the format
4580 16. The size of the buffer for calls to the DNS resolver has been increased
4581 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4582 for addresses that have a lot of PTR records. This alleviates a problem; it
4583 does not fully solve it.
4585 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4586 buffer now truncates the list and logs the incident, which is the same
4587 action as happens when Exim is looking up a host name and its aliases.
4588 Previously in this situation something unpredictable would happen;
4589 sometimes it was "internal error: store_reset failed".
4591 18. If a server dropped the connection unexpectedly when an Exim client was
4592 using GnuTLS and trying to read a response, the client delivery process
4593 crashed while trying to generate an error log message.
4595 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4596 string, for example:
4598 warn message = H1: something\nH2: something
4600 the text was added as a single header line from Exim's point of view
4601 though it ended up OK in the delivered message. However, searching for the
4602 second and subsequent header lines using $h_h2: did not work. This has been
4603 fixed. Similarly, if a system filter added multiple headers in this way,
4604 the routers could not see them.
4606 20. Expanded the error message when iplsearch is called with an invalid key to
4607 suggest using net-iplsearch in a host list.
4609 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4610 ACLs are no longer actually imposed (and a message to that effect is
4613 22. If a "gecos" field in a passwd entry contained escaped characters, in
4614 particular, if it contained a \" sequence, Exim got it wrong when building
4615 a From: or a Sender: header from that name. A second bug also caused
4616 incorrect handling when an unquoted " was present following a character
4617 that needed quoting.
4619 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4620 was not being matched caselessly.
4622 24. Arranged for all hyphens in the exim.8 source to be escaped with
4625 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4626 properly. Recipient callout cache records were still being keyed to include
4627 the sender, even when use_sender was set false. This led to far more
4628 callouts that were necessary. The sender is no longer included in the key
4629 when use_sender is false.
4631 26. Added "control = submission" modifier to ACLs.
4633 27. Added the ${base62d: operator to decode base 62 numbers.
4635 28. dnsdb lookups can now access SRV records.
4637 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4638 the configuration file.
4640 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4641 (-v) mode. This makes the output for a verbose queue run more intelligible.
4643 31. Added a use_postmaster feature to recipient callouts.
4645 32. Added the $body_zerocount variable, containing the number of binary zero
4646 bytes in the message body.
4648 33. The time of last modification of the "new" subdirectory is now used as the
4649 "mailbox time last read" when there is a quota error for a maildir
4652 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4654 35. Added +ignore_unknown as a special item in host lists.
4656 36. Code for decoding IPv6 addresses in host lists is now included, even if
4657 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4658 address was recognized as an IP address, but was then not correctly decoded
4659 into binary, causing unexpected and incorrect effects when compared with
4666 1. Very minor rewording of debugging text in manualroute to say "list of
4667 hosts" instead of "hostlist".
4669 2. If verify=header_syntax was set, and a header line with an unqualified
4670 address (no domain) and a large number of spaces between the end of the
4671 name and the colon was received, the reception process suffered a buffer
4672 overflow, and (when I tested it) crashed. This was caused by some obsolete
4673 code that should have been removed. The fix is to remove it!
4675 3. When running in the test harness, delay a bit after writing a bounce
4676 message to get a bit more predictability in the log output.
4678 4. Added a call to search_tidyup() just before forking a reception process. In
4679 theory, someone could use a lookup in the expansion of smtp_accept_max_
4680 per_host which, without the tidyup, could leave open a database connection.
4682 5. Added the variables $recipient_data and $sender_data which get set from a
4683 lookup success in an ACL "recipients" or "senders" condition, or a router
4684 "senders" option, similar to $domain_data and $local_part_data.
4686 6. Moved the writing of debug_print from before to after the "senders" test
4689 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4690 problems for message scanning, either using a data ACL, or using
4691 local_scan() because the Received: header was not generated till after they
4692 were called (in order to set the time as the time of reception completion).
4693 I have revised the way this works. The header is now generated after the
4694 body is received, but before the ACL or local_scan() are called. After they
4695 are run, the timestamp in the header is updated.
4701 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4702 before starting a queue runner without re-exec. This happened only when
4703 deliver_drop_privilege was set or when the Exim user was set to root. The
4704 effect of the bug was that timeouts during subsequent deliveries caused
4705 crashes instead of being properly handled. The handler is now left at its
4706 default (and expected) setting.
4708 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4709 message, again when deliver_drop_privilege is set or Exim is run as root.
4710 The bug described in (1) was not present in this case, but the tidying up
4711 of the other signals was missing. I have made the two cases consistent.
4713 3. The ignore_target_hosts setting on a manualroute router was being ignored
4714 for hosts that were looked up using the /MX notation.
4716 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4719 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4720 operated on the sender address. After changing the $sender_address to <>
4721 for the sender address verify, Exim was re-instated it as the original
4722 (before rewriting) address, but remembering that it had rewritten it, so it
4723 wasn't rewriting it again. This bug also had the effect of breaking the
4724 sender address verification caching when the sender address was rewritten.
4726 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4727 This has been changed so that if the ip literal address matches
4728 ignore_target_hosts, the router declines.
4730 7. Added expansion conditions match_domain, match_address, and match_local_
4731 part (NOT match_host).
4733 8. The placeholder for the Received: header didn't have a length field set.
4735 9. Added code to Exim itself and to exim_lock to test for a specific race
4736 condition that could lead to file corruption when using MBX delivery. The
4737 issue is with the lockfile that is created in /tmp. If this file is removed
4738 after a process has opened it but before that process has acquired a lock,
4739 there is the potential for a second process to recreate the file and also
4740 acquire a lock. This could lead to two Exim processes writing to the file
4741 at the same time. The added code performs the same test as UW imapd; it
4742 checks after acquiring the lock that its file descriptor still refers to
4743 the same named file.
4745 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4746 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4749 11. Added the smtp_active_hostname option. If used, this will typically be made
4750 to depend on the incoming interface address. Because $interface_address is
4751 not set up until the daemon has forked a reception process, error responses
4752 that can happen earlier (such as "too many connections") no longer contain
4755 12. If an expansion in a condition on a "warn" statement fails because a lookup
4756 defers, the "warn" statement is abandoned, and the next ACL statement is
4757 processed. Previously this caused the whole ACL to be aborted.
4759 13. Added the iplsearch lookup type.
4761 14. Added ident_timeout as a log selector.
4763 15. Added tls_certificate_verified as a log selector.
4765 16. Added a global option tls_require_ciphers (compare the smtp transport
4766 option of the same name). This controls incoming TLS connections.
4768 17. I finally figured out how to make tls_require_ciphers do a similar thing
4769 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4770 before starting the TLS session.
4772 18. Tabs are now shown as \t in -bP output.
4774 19. If the log selector return_path_on_delivery was set, Exim crashed when
4775 bouncing a message because it had too many Received: header lines.
4777 20. If two routers both had headers_remove settings, and the first one included
4778 a superfluous trailing colon, the final name in the first list and the
4779 first name in the second list were incorrectly joined into one item (with a
4780 colon in the middle).
4786 1. Added -C and -D options to the exinext utility, mainly to make it easier
4787 to include in the automated testing, but these could be helpful when
4788 multiple configurations are in use.
4790 2. The exinext utility was not formatting the output nicely when there was
4791 an alternate port involved in the retry record key, nor when there was a
4792 message id as well (for retries that were specific to a specific message
4793 and a specific host). It was also confused by IPv6 addresses, because of
4794 the additional colons they contain. I have fixed the IPv4 problem, and
4795 patched it up to do a reasonable job for IPv6.
4797 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4798 delivery, the log line now contains "pipelined" if PIPELINING was used.
4800 4. An SMTP transport process used to panic and die if the bind() call to set
4801 an explicit outgoing interface failed. This has been changed; it is now
4802 treated in the same way as a connect() failure.
4804 5. A reference to $sender_host_name in the part of a conditional expansion
4805 that was being skipped was still causing a DNS lookup. This no longer
4808 6. The def: expansion condition was not recognizing references to header lines
4809 that used bh_ and bheader_.
4811 7. Added the _cache feature to named lists.
4813 8. The code for checking quota_filecount in the appendfile transport was
4814 allowing one more file than it should have been.
4816 9. For compatibility with Sendmail, the command line option
4824 and sets the incoming protocol and host name (for trusted callers). The
4825 host name and its colon can be omitted when only the protocol is to be set.
4826 Note the Exim already has two private options, -pd and -ps, that refer to
4827 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4828 "s", but I don't think that's a major issue.
4830 10. A number of refactoring changes to the code, none of which should affect
4833 (a) The number of logging options was getting close to filling up the
4834 32-bit word that was used as a bit map. I have split them into two classes:
4835 those that are passed in the argument to log_write(), and those that are
4836 only ever tested independently outside of that function. These are now in
4837 separate 32-bit words, so there is plenty of room for expansion again.
4838 There is no change in the user interface or the logging behaviour.
4840 (b) When building, for example, log lines, the code previously used a
4841 macro that called string_cat() twice, in order to add two strings. This is
4842 not really sufficiently general. Furthermore, there was one instance where
4843 it was actually wrong because one of the argument was used twice, and in
4844 one call a function was used. (As it happened, calling the function twice
4845 did not affect the overall behaviour.) The macro has been replaced by a
4846 function that can join an arbitrary number of extra strings onto a growing
4849 (c) The code for expansion conditions now uses a table and a binary chop
4850 instead of a serial search (which was left over from when there were very
4851 few conditions). Also, it now recognizes conditions like "pam" even when
4852 the relevant support is not compiled in: a suitably worded error message is
4853 given if an attempt is made to use such a condition.
4855 11. Added ${time_interval:xxxxx}.
4857 12. A bug was causing one of the ddress fields not to be passed back correctly
4858 from remote delivery subprocesses. The field in question was not being
4859 subsequently used, so this caused to problems in practice.
4861 13. Added new log selectors queue_time and deliver_time.
4863 14. Might have fixed a bug in maildirsizefile handling that threw up
4864 "unexpected character" debug warnings, and recalculated the data
4865 unnecessarily. In any case, I expanded the warning message to give more
4868 15. Added the message "Restricted characters in address" to the statements in
4869 the default ACL that block characters like @ and % in local parts.
4871 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4872 Three changes have been made:
4874 (a) There was a serious bug; a negative response to MAIL caused the whole
4875 recipient domain to be cached as invalid, thereby blocking all messages
4876 to all local parts at the same domain, from all senders. This bug has
4877 been fixed. The domain is no longer cached after a negative response to
4878 MAIL if the sender used is not empty.
4880 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4883 (c) A new callout option, "use_sender" has been added for people who want
4884 the modified behaviour.
4890 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4893 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4894 indeed breaks things for older releases.
4896 3. Added additional logging to the case where there is a problem reading data
4897 from a filter that is running in a subprocess using a pipe, in order to
4898 try to track down a specific problem.
4900 4. Testing facility fudge: when running in the test harness and attempting
4901 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4902 getting "No route to host". Convert this to a timeout.
4904 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4907 6. Some OS don't have socklen_t but use size_t instead. This affects the
4908 fifth argument of getsockopt() amongst other things. This is now
4909 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4910 can be set for individual OS. I have set it for SunOS5, OSF1, and
4911 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4912 some earlier ones do not.
4914 7. Change 4.30/15 was not doing the test caselessly.
4916 8. The standard form for an IPv6 address literal was being rejected by address
4917 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4918 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4919 this, as well as the form without the "IPv6" on the front (but only when
4920 address literals are enabled, of course).
4922 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4924 10. Exim crashed if a message with an empty sender address specified by -f
4925 encountered a router with an errors_to setting. This could be provoked only
4926 by a command such as
4930 where an empty string was supplied; "<>" did not hit this bug.
4932 11. Installed PCRE release 4.5.
4934 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4935 remained set. It is now erased.
4937 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4938 times from message ids (which are base 36 rather than the normal 62).
4940 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4941 were being counted as actual protocol errors, and logged if the log
4942 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4943 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4944 and DATA following a set of rejected RCPTs do not count as protocol errors.
4945 In other words, Exim assumes they were pipelined, though this may not
4946 actually be the case. Of course, in all cases the client gets an
4947 appropriate error code.
4949 15. If a lookup fails in an ACL condition, a message about the failure may
4950 be available; it is used if testing the ACL cannot continue, because most
4951 such messages specify what the cause of the deferral is. However, some
4952 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4953 that caused an old message to be retained and used if a later statement
4954 caused a defer, replacing the real cause of the deferral.
4956 16. If an IP address had so many PTR records that the DNS lookup buffer
4957 was not large enough to hold them, Exim could crash while trying to process
4958 the truncated data. It now detects and logs this case.
4960 17. Further to 4.21/58, another change has been made: if (and only if) the
4961 first line of a message (the first header line) ends with CRLF, a bare LF
4962 in a subsequent header line has a space inserted after it, so as not to
4963 terminate the header.
4965 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4966 unnecessarily, used atoi() instead of strtol(), and didn't check the
4967 termination when getting file sizes from file names by regex.
4969 19. Completely re-implemented the support for maildirsize files, in the light
4970 of a number of problems with the previous contributed implementation
4971 (4.30/29). In particular:
4973 . If the quota is zero, the maildirsize file is maintained, but no quota is
4976 . If the maildir directory does not exist, it is created before any attempt
4977 to write a maildirsize file.
4979 . The quota value in the file is just a cache; if the quota is changed in
4980 the transport, the new value overrides.
4982 . A regular expression is available for excluding directories from the
4985 20. The autoreply transport checks the characters in options that define the
4986 message's headers; it allows continued headers, but it was checking with
4987 isspace() after an embedded newline instead of explicitly looking for a
4990 21. If all the "regular" hosts to which an address was routed had passed their
4991 expiry times, and had not reached their retry times, the address was
4992 bounced, even if fallback hosts were defined. Now Exim should go on to try
4995 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4996 equivalent code in the SMTP transport. Some hosts send humungous responses
4997 to HELO/EHLO, more than 1024 it seems.
4999 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5000 gives compiler warnings in some environments. I've now done it "properly",
5003 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5004 (because of problems with the built-in one) was declared to return uschar *
5005 instead of char *, causing compiler failure.
5007 25. Fixed a file descriptor leak when processing alias/forward files.
5009 26. Fixed a minor format string issue in dbfn.c.
5011 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5013 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5014 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5017 29. When a sender address is verified, it is cached, to save repeating the test
5018 when there is more than one recipient in a message. However, when the
5019 verification involves a callout, it is possible for different callout
5020 options to be set for different recipients. It is too complicated to keep
5021 track of this in the cache, so now Exim always runs a verification when a
5022 callout is required, relying on the callout cache for the optimization.
5023 The overhead is duplication of the address routing, but this should not be
5026 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5027 to be verified with callout=postmaster, and the main callout worked but the
5028 postmaster check failed, the verification correctly failed. However, if a
5029 subsequent RCPT command asked for sender verification *without* the
5030 postmaster check, incorrect caching caused this verification also to fail,
5033 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5034 it was not caching the DNS options (qualify_single, search_parents) that
5035 were used when the lookup failed. A subsequent lookup with different
5036 options therefore always gave the same answer, though there were cases
5037 where it should not have. (Example: a "domains = !$mx_any" option on a
5038 dnslookup router: the "domains" option is always processed without any
5039 widening, but the router might have qualify_single set.) Now Exim uses the
5040 cached value only when the same options are set.
5042 32. Added John Jetmore's "exipick" utility to the distribution.
5044 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5045 than a timeout (e.g. a certificate is required, and is not provided), an
5046 Exim server now closes the connection immediately. Previously it waited for
5047 the client to close - but if the client is SSL, it seems that they each
5048 wait for each other, leading to a delay before one of them times out.
5050 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5051 maintained 0.8.x compatibility because I don't think many are using it, and
5052 it is clearly obsolete.
5054 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5057 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5058 client certificate was expired. A simple patch fixes this, though I don't
5059 understand the full logic of why the verify callback is called multiple
5062 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5063 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5064 which causes problems with some clients (such as the Certicom SSL Plus
5065 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5066 disables the coutermeasure allowing Eudora to connect."
5068 38. Exim was not checking that a write() to a log file succeeded. This could
5069 lead to Bad Things if a log got too big, in particular if it hit a file
5070 size limit. Exim now panics and dies if it cannot write to a log file, just
5071 as it does if it cannot open a log file.
5073 39. Modified OS/Makefile-Linux so that it now contains
5075 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5077 The two -D definitions ensure that Exim is compiled with large file
5078 support, which makes it possible to handle log files that are bigger than
5081 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5082 instance) a domain was checked against a named list that involved a lookup,
5083 causing $domain_data to be set, then another domain was checked against the
5084 same list, then the first domain was re-checked, the value of $domain_data
5085 after the final check could be wrong. In particular, if the second check
5086 failed, it could be set empty. This bug probably also applied to
5089 41. The strip_trailing_dot option was not being applied to the address given
5090 with the -f command-line option.
5092 42. The code for reading a message's header from the spool was incrementing
5093 $received_count, but never initializing it. This meant that the value was
5094 incorrect (doubled) while delivering a message in the same process in which
5095 it was received. In the most common configuration of Exim, this never
5096 happens - a fresh exec is done - but it can happen when
5097 deliver_drop_privilege is set.
5099 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5100 it now includes up to 150 characters of the unexpected data in the log
5103 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5104 and building data strings. The size of both of these buffers was 10 000
5105 bytes - far larger than anybody would *ever* want, thought I. Needless to
5106 say, somebody hit the limit. I have increased the maximum line length to
5107 20 000 and the maximum data length of concatenated lines to 100 000. I have
5108 also fixed two bugs, because there was no checking on these buffers. Tsk,
5109 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5110 buffer is too small.
5112 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5113 lsearch lookups. Now it does.
5115 46. When parsing a route_list item in a manualroute router, a fixed-length
5116 buffer was used for the list of hosts. I made this 1024 bytes long,
5117 thinking that nobody would ever have a list of hosts that long. Wrong.
5118 Somebody had a whole pile of complicated expansion conditions, and the
5119 string was silently truncated, leading to an expansion error. It turns out
5120 that it is easier to change to an unlimited length (owing to other changes
5121 that have happened since this code was originally written) than to build
5122 structure for giving a limitation error. The length of the item that
5123 expands into the list of hosts is now unlimited.
5125 47. The lsearch lookup could not handle data where the length of text line was
5126 more than 4095 characters. Such lines were truncated, leading to shortened
5127 data being returned. It should now handle lines of any length.
5129 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5130 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5133 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5134 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5135 no longer happens in comment lines. A list of replacements is now placed
5136 at the head of all of the source files, except those whose only change is
5137 to replace PERL_COMMAND in the very first #! line.
5139 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5140 messages on the queue, with a bottom-up merge sort, using code contributed
5141 by Michael Haardt. This should make operations like -bp somewhat faster on
5142 large queues. It won't affect queue runners, except when queue_run_in_order
5145 51. Installed eximstats 1.31 in the distribution.
5147 52. Added support for SRV lookups to the dnslookup router.
5149 53. If an ACL referred to $message_body or $message_body_end, the value was not
5150 reset for any messages that followed in the same SMTP session.
5152 54. The store-handling optimization for building very long strings was not
5153 differentiating between the different store pools. I don't think this
5154 actually made any difference in practice, but I've tidied it.
5156 55. While running the routers to verify a sender address, $sender_address
5157 was still set to the sender address. This is wrong, because when routing to
5158 send a bounce to the sender, it would be empty. Therefore, I have changed
5159 it so that, while verifying a sender address, $sender_address is set to <>.
5160 (There is no change to what happens when verifying a recipient address.)
5162 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5163 target A or AAAA records (if not already returned) without resetting the
5164 qualify_single or search_parents options of the DNS resolver. These are
5165 inappropriate in this case because the targets of MX and SRV records must
5166 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5167 target that, when qualified, matched something in the local domain. These
5168 two options are now turned off when doing these lookups.
5170 57. It seems that at least some releases of Reiserfs (which does not have the
5171 concept of a fixed number of inodes) returns zero and not -1 for the
5172 number of available inodes. This interacted badly with check_spool_inodes,
5173 which assumed that -1 was the "no such thing" setting. What I have done is
5174 to check that the total number of inodes is greater than zero before doing
5175 the test of how many are available.
5177 58. When a "warn" ACL statement has a log_message modifier, the message is
5178 remembered, and not repeated. This is to avoid a lot of repetition when a
5179 message has many recipients that cause the same warning to be written.
5180 Howewer, Exim was preserving the list of already written lines for an
5181 entire SMTP session, which doesn't seem right. The memory is now reset if a
5182 new message is started.
5184 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5185 debugging output unless log_rewrite was also set.
5187 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5188 of (int)(handle) when we know that handle contains (void *)(-1).
5190 61. The Exim daemon panic-logs an error return when it closes the incoming
5191 connection. However "connection reset by peer" seems to be common, and
5192 isn't really an error worthy of noting specially, so that particular error
5195 62. When Exim is trying to find all the local interfaces, it used to panic and
5196 die if the ioctl to get the interface flags failed. However, it seems that
5197 on at least one OS (Solaris 9) it is possible to have an interface that is
5198 included in the list of interfaces, but for which you get a failure error
5199 for this call. This happens when the interface is not "plumbed" into a
5200 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5201 failure of the "get flags" call assumes that the interface is down.
5203 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5204 makes life easier for people who are doing arithmetic on fields extracted
5205 from dates, where you often get leading zeros that should not be
5206 interpreted as octal.
5208 64. Added qualify_domain to the redirect router, to override the global
5211 65. If a pathologically long header line contained very many addresses (the
5212 report of this problem mentioned 10 000) and each of them was rewritten,
5213 Exim could use up a very large amount of memory. (It kept on making new
5214 copies of the header line as it rewrote, and never released the old ones.)
5215 At the expense of a bit more processing, the header rewriting function has
5216 been changed so that it no longer eats memory in this way.
5218 66. The generation of the Received: header has been moved from the time that a
5219 message starts to be received, to the time that it finishes. The timestamp
5220 in the Received: header should now be very close to that of the <= log
5221 line. There are two side-effects of this change:
5223 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5224 logged header lines no longer include the local Received: line, because
5225 it has not yet been created. The same applies to a copy of the message
5226 that is returned to a non-SMTP sender when a message is rejected.
5228 (b) When a filter file is tested using -bf, no additional Received: header
5229 is added to the test message. After some thought, I decided that this
5232 This change does not affect the value of $received_for. It is still set
5233 after address rewriting, but before local_scan() is called.
5235 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5237 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5238 gave an unhelpful panic error message, and a defer error. I have managed to
5239 change this behaviour so that it now rejects any supplied certificate,
5240 which seems right, as the list of acceptable certificates is empty.
5242 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5243 gave an unhelpful defer error. I have not managed to make this reject any
5244 supplied certificates, but the error message it gives is "no certificate
5245 supplied", which is not helpful.
5247 70. exigrep's output now also includes lines that are not associated with any
5248 message, but which match the given pattern. Implemented by a patch from
5249 Martin Sluka, which also tidied up the Perl a bit.
5251 71. Recipient callout verification, like sender verification, was using <> in
5252 the MAIL FROM command. This isn't really the right thing, since the actual
5253 sender may affect whether the remote host accepts the recipient or not. I
5254 have changed it to use the actual sender in the callout; this means that
5255 the cache record is now keyed on a recipient/sender pair, not just the
5256 recipient address. There doesn't seem to be a real danger of callout loops,
5257 since a callout by the remote host to check the sender would use <>.
5258 [SEE ABOVE: changed after hitting problems.]
5260 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5261 temporary errors. However, in the case of such a code being given after
5262 the end of a data transmission (i.e. after ".") Exim was failing to write
5263 a retry record for the message. (Yes, there was some broken host that was
5264 actually sending 8xx at this point.)
5266 73. An unknown lookup type in a host list could cause Exim to panic-die when
5267 the list was checked. (An example that provoked this was putting <; in the
5268 middle of a list instead of at the start.) If this happened during a DATA
5269 ACL check, a -D file could be left lying around. This kind of configuration
5270 error no longer causes Exim to die; instead it causes a defer errror. The
5271 incident is still logged to the main and panic logs.
5273 74. Buglet left over from Exim 3 conversion. The message "too many messages
5274 in one connection" was written to the rejectlog but not the mainlog, except
5275 when address rewriting (yes!) was being logged.
5277 75. Added write_rejectlog option.
5279 76. When a system filter was run not as root (that is, when system_filter_user
5280 was set), the values of the $n variables were not being returned to the
5281 main process; thus, they were not subsequently available in the $sn
5284 77. Added +return_path_on_delivery log selector.
5286 78. A connection timeout was being treated differently from recipients deferred
5287 when testing hosts_max_try with a message that was older than the host's
5288 retry timeout. (The host should not be counted, thus allowing all hosts to
5289 be tried at least once before bouncing.) This may have been the cause of an
5290 occasionally reported bug whereby a message would remain on the queue
5291 longer than the retry timeout, but would be bounced if a delivery was
5292 forced. I say "may" because I never totally pinned down the problem;
5293 setting up timeout/retry tests is difficult. See also the next item.
5295 79. The ultimate address timeout was not being applied to errors that involved
5296 a combination of host plus message (for example, a timeout on a MAIL
5297 command). When an address resolved to a number of possible hosts, and they
5298 were not all tried for each delivery (e.g. because of hosts_max_try), a
5299 message could remain on the queue longer than the retry timeout.
5301 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5304 81. Fixed an obscure SMTP outgoing bug which required at least the following
5305 conditions: (a) there was another message waiting for the same server;
5306 (b) the server returned 5xx to all RCPT commands in the first message so
5307 that the message was not completed; (c) the server dropped the connection
5308 or gave a negative response to the RSET that Exim sends to abort the
5309 transaction. The observed case was a dropped connection after DATA that had
5310 been sent in pipelining mode. That is, the server had advertised PIPELINING
5311 but was not implementing it correctly. The effect of the bug was incorrect
5312 behaviour, such as trying another host, and this could lead to a crash.
5318 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5319 and daemon.c were passed as pointers to ints; they should have been
5320 pointers to socklen_t variables (which are typically unsigned ints).
5322 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5325 3. Fixed a really odd bug that affected only the testing scheme; patching a
5326 certain fixed string in the binary changed the value of another string that
5327 happened to be identical to the end of the original first string.
5329 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5330 name", it returns that address as the IP address. On some operating
5331 systems (e.g. Solaris), it also passes back the IP address string as the
5332 "host name". However, on others (e.g. Linux), it passes back an empty
5333 string. Exim wasn't checking for this, and was changing the host name to an
5334 empty string, assuming it had been canonicized.
5336 5. Although rare, it is permitted to have more than one PTR record for a given
5337 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5338 all the names associated with an address, because they do in Solaris.
5339 However, it seems that they do not in Linux for data that comes from the
5340 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5341 I found this out when I moved to a new Linux workstation and tried to run
5342 the Exim test suite.
5344 To get round this problem I have changed the code so that it now does its
5345 own call to the DNS to look up PTR records when searching for a host name.
5346 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5347 addresses that are only in /etc/hosts are still found.
5349 This behaviour is, however, controlled by an option called host_lookup_
5350 order, which defaults to "bydns:byaddr". If people want to use the other
5351 order, or indeed, just use one or the other means of lookup, they can
5352 specify it in this variable.
5354 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5355 some operating systems, this comes back from gethostbyaddr() as an empty
5356 string, and this is what Exim used to test for. However, it seems that in
5357 other systems, "." is yielded. Exim now tests for this case too.
5359 7. The values of check_spool_space and check_log_space are now held internally
5360 as a number of kilobytes instead of an absolute number of bytes. If a
5361 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5362 kilobyte. This means that much larger values can be stored.
5364 8. Exim monitor: an attempt to get the action menu when not actually pointing
5365 at a message produces an empty menu entitled "No message selected". This
5366 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5367 no entries in it ("Shell widget menu has zero width and/or height"). So I
5368 have added a single, blank menu entry in this case.
5370 9. Added ${quote_local_part.
5372 10. MIME decoding is now applied to the contents of Subject: header lines when
5375 11. Now that a reference to $sender_host_address automatically causes a reverse
5376 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5377 host lookup before query-style lookups in lists that might use this
5378 variable. This has therefore been abolished, and the "net-" prefix is no
5379 longer necessary for query-style lookups.
5381 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5382 have been a typo for LFLAGS, so it has been changed.
5384 13. The install script calls Exim with "-C /dev/null" in order to find the
5385 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5386 to be output. Howeve, since Exim outputs its version number before the
5387 error, it didn't break the script. It just looked ugly. I fixed this by
5388 always allowing "-C /dev/null" if the caller is root.
5390 14. Ignore overlarge ACL variable number when reading spool file - insurance
5391 against a later release with more variables having written the file.
5393 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5394 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5395 this, as well as the form without the "IPv6" on the front.
5397 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5398 OS/Makefile-Darwin file.
5400 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5401 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5403 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5404 variable. However, some OS use ints for inodes. I've added cast to long int
5405 to get rid of the compiler warning.
5407 19. I had forgotten to lock out "/../" in configuration file names when
5408 ALT_CONFIG_PREFIX was set.
5410 20. Routers used for verification do not need to specify transports. However,
5411 if such a router generated a host list, and callout was configured, Exim
5412 crashed, because it could not find a port number from the (non-existent)
5413 transport. It now assumes port 25 in this circumstance.
5415 21. Added the -t option to exigrep.
5417 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5418 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5419 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5420 them is set, LOOKUP_LSEARCH is forced.
5422 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5424 24. Added sender and host information to the "rejected by local_scan()" log
5425 line; previously there was no indication of these.
5427 25. Added .include_if_exists.
5429 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5430 Linux. It turns out that not all file systems support this. Apparently some
5431 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5432 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5433 not supported on the file descriptor, is now ignored when Exim is trying to
5434 sync a directory. This applies only to Linux.
5436 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5438 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5439 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5442 29. Added support for maildirsize files from supplied patch (modified a bit).
5444 30. The use of :fail: followed by an empty string could lead Exim to respond to
5445 sender verification failures with (e.g.):
5447 550 Verification failed for <xxx>
5448 550 Sender verify failed
5450 where the first response line was missing the '-' that indicates it is not
5451 the final line of the response.
5453 31. The loop for finding the name of the user that called Exim had a hardwired
5454 limit of 10; it now uses the value of finduser_retries, which is used for
5455 all other user lookups.
5457 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5460 33. Exim was neglecting to zero errno before one call of strtol() when
5461 expanding a string and expecting an integer value. On some systems this
5462 resulted in spurious "integer overflow" errors. Also, it was casting the
5463 result into an int without checking.
5465 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5466 did not work. The code looks as if it has *never* worked, though it appears
5467 to have been documented since at least releast 1.62. I have made it work.
5469 35. The "timeout_DNS" error in retry rules, also documented since at least
5470 1.62, also never worked. As it isn't clear exactly what this means, and
5471 clearly it isn't a major issue, I have abolished the feature by treating it
5472 as "timeout", and writing a warning to the main and panic logs.
5474 36. The display of retry rules for -brt wasn't always showing the error code
5477 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5478 timeout_connect_A, timeout_connect_MX.
5480 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5481 to the empty sender.
5483 39. The daemon was not analysing the content of -oX till after it had closed
5484 stderr and disconnected from the controlling terminal. This meant that any
5485 syntax errors were only noted on the panic log, and the return code from
5486 the command was 0. By re-arranging the code a little, I've made the
5487 decoding happen first, so such errors now appear on stderr, and the return
5488 code is 1. However, the actual setting up of the sockets still happens in
5489 the disconnected process, so errors there are still only recorded on the
5492 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5493 connections (as happens on some IP stacks) was logged at start up time as
5494 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5495 it from "IPv6 and IPv4", which means that two separate sockets are being
5498 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5499 says whether AF_INET or AF_INET6 was passed as an argument.
5501 42. Exiwhat output was messed up when time zones were included in log
5504 43. Exiwhat now gives more information about the daemon's listening ports,
5505 and whether -tls-on-connect was used.
5507 44. The "port" option of the smtp transport is now expanded.
5509 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5510 silently ignored. Now an error message is written to the main and panic
5513 46. There's a new ACL modifier called "logwrite" which writes to a log file
5514 as soon as it is encountered.
5516 47. Added $local_user_uid and $local_user_gid at routing time.
5518 48. Exim crashed when trying to verify a sender address that was being
5521 49. Exim was recognizing only a space character after ".include". It now also
5522 recognizes a tab character.
5524 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5525 extracting the relevant information from the specification. The man page no
5526 longer contains scrambled data for the -d option, and I've added a section
5527 at the front about calling Exim under different names.
5529 51. Added "extra_headers" argument to the "mail" command in filter files.
5531 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5534 53. Installed eximstats 1.29.
5536 54. Added transport_filter_timeout as a generic transport option.
5538 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5539 Cc: header lines. This was required by RFC 822, but it not required by RFC
5542 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5543 incoming messages that did not have them. Now it does so only if the
5544 message originates locally, that is, if there is no associated remote host
5545 address. When Resent- header lines are present, this applies to the Resent-
5546 lines rather than the non-Resent- lines.
5548 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5549 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5551 58. Messages for configuration errors now include the name of the main
5552 configuration file - useful now that there may be more than one file in a
5553 list (.included file names were always shown).
5555 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5556 for those rare installations that do not start the daemon as root or run it
5557 setuid root. I've cut out the call to initgroups() if the daemon is not
5560 60. The Exim user and group can now be bound into the binary as text strings
5561 that are looked up at the start of Exim's processing.
5563 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5565 62. Added $mailstore_basename variable.
5567 63. Installed patch to sieve.c from Michael Haardt.
5569 64. When Exim failed to open the panic log after failing to open the main log,
5570 the original message it was trying to log was written to stderr and debug
5571 output, but if they were not available (the usual case in production), it
5572 was lost. Now it is written to syslog before the two lines that record the
5573 failures to open the logs.
5575 65. Users' Exim filters run in subprocesses under the user's uid. It is
5576 possible for a "deliver" command or an alias in a "personal" command to
5577 provoke an address rewrite. If logging of address rewriting is configured,
5578 this fails because the process is not running as root or exim. There may be
5579 a better way of dealing with this, but for the moment (because 4.30 needs
5580 to be released), I have disabled address rewrite logging when running a
5581 filter in a non-root, non-exim process.
5587 1. The buildconfig auxiliary program wasn't quoting the value set for
5588 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5589 not defined. This bug was masked in 4.22 by the effect that was fixed in
5592 2. Some messages that were rejected after a message id was allocated were
5593 shown as "incomplete" by exigrep. It no longer does this for messages that
5594 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5596 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5597 have allow_domain_literals set, the ID did not get logged in the <= line.
5598 Domain literals are now always recognized in Message-ID: header lines.
5600 4. The first argument for a ${extract expansion item is the key name or field
5601 number. Leading and trailing spaces in this item were not being ignored,
5602 causing some misleading effects.
5604 5. When deliver_drop_privilege was set, single queue runner processes started
5605 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5606 same command in the process it spins off) were not dropping privilege.
5608 6. When the daemon running as "exim" started a queue runner, it always
5609 re-executed Exim in the spun-off process. This is a waste of effort when
5610 deliver_drop_privilege is set. The new process now just calls the
5611 queue-runner function directly.
5617 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5620 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5621 ignored. Though the use of -oP was forcing the writing of a pid file, it
5622 was always written to the default place.
5624 3. If the message "no IP address found for host xxxx" is generated during
5625 incoming verification, it is now followed by identification of the incoming
5626 connection (so you can more easily find what provoked it).
5628 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5630 5. Added some features to "harden" Exim a bit more against certain attacks:
5632 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5633 be put in Local/Makefile. This is like the never_users runtime option,
5634 but it cannot be overridden. The default setting is "root".
5636 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5637 prefix string with which any file named in a -C command line option
5640 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5641 is retained for -C and -D only if the caller of Exim is root. Without
5642 it, the exim user may also use -C and -D and retain privilege.
5644 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5645 command line option is disabled.
5647 6. Macro names set by the -D option must start with an upper case letter, just
5648 like macro names defined in the configuration file.
5650 7. Added "dereference=" facility to LDAP.
5652 8. Two instances of the typo "uknown" in the source files are fixed.
5654 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5655 the Configure-Makefile script screwed up while processing it.
5657 10. Incorporated PCRE 4.4.
5659 11. The SMTP synchronization check was not operating right at the start of an
5660 SMTP session. For example, it could not catch a HELO sent before the client
5661 waited for the greeting. There is now a check for outstanding input at the
5662 point when the greeting is written. Because of the duplex, asynchronous
5663 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5664 way, but not yet received, when the check is performed.
5666 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5667 on TCP/IP sockets, because this apparently causes some broken clients to
5670 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5671 unchanged) from the Cygwin maintainer.
5673 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5674 was supported instead of testing for mbx. Effectively a typo.
5676 15. The spa authenticator server code was not checking that the input it
5677 received was valid base64.
5679 16. The debug output line for the "set" modifier in ACLs was not showing the
5680 name of the variable that was being set.
5682 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5684 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5685 lookup on its own. Something else was needed to trigger the lookup. For
5686 example, a match in host_lookup or the need for a host name in a host list.
5687 Now, if $sender_host_name is referenced and the host name has not yet been
5688 looked up, a lookup is performed. If the lookup fails, the variable remains
5689 empty, and $host_lookup_failed is set to "1".
5691 19. Added "eqi" as a case-independent comparison operator.
5693 20. The saslauthd authentication condition could segfault if neither service
5694 nor realm was specified.
5696 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5697 error message that was logged was misleading, and incoming SMTP
5698 connections were dropped. The message is now more accurate, and temporary
5699 errors are given to SMTP connections.
5701 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5702 (except RSET) with a 421 error, until QUIT is received. However, it was
5703 failing to send a response to QUIT.
5705 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5706 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5707 ACL is now run after the helo_try_verify_hosts code.
5709 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5710 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5711 case-independent, so other case variants are also recognized). Apparently
5712 some systems use these upper case variants.
5714 25. If more than two messages were waiting for the same host, and a transport
5715 filter was specified for the transport, Exim sent two messages over the
5716 same TCP/IP connection, and then failed with "socket operation on non-
5717 socket" when it tried to send the third.
5719 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5721 27. The extern definition of crypt16() in expand.c was not being excluded when
5722 the OS had its own crypt16() function.
5724 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5725 as a preferred synonym for return_size_limit, both as an option and as an
5728 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5730 30. Changed the default configuration ACL to relax the local part checking rule
5731 for addresses that are not in any local domains. For these addresses,
5732 slashes and pipe symbols are allowed within local parts, but the sequence
5733 /../ is explicitly forbidden.
5735 31. SPA server authentication was not clearing the challenge buffer before
5738 32. log_message in a "warn" ACL statement was writing to the reject log as
5739 well as to the main log, which contradicts the documentation and doesn't
5740 seem right (because no rejection is happening). So I have stopped it.
5742 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5743 However, I am unable to do any testing of this.
5745 34. Fixed an infelicity in the appendfile transport. When checking directories
5746 for a mailbox, to see if any needed to be created, it was accidentally
5747 using path names with one or more superfluous leading slashes; tracing
5748 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5750 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5751 contents are added to the log line that is written for every discarded
5752 recipient. (Previously a log_message setting was ignored.)
5754 36. The ${quote: operator now quotes the string if it is empty.
5756 37. The install script runs exim in order to find its version number. If for
5757 some reason other than non-existence or emptiness, which it checks, it
5758 could not run './exim', it was installing it with an empty version number,
5759 i.e. as "exim-". This error state is now caught, and the installation is
5762 38. An argument was missing from the function that creates an error message
5763 when Exim fails to connect to the socket for saslauthd authentication.
5764 This could cause Exim to crash, or give a corrupted message.
5766 39. Added isip, isip4, and isip6 to ${if conditions.
5768 40. The ACL variables $acl_xx are now saved with the message, and can be
5769 accessed later in routers, transports, and filters.
5771 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5772 strings in the file are not string-expanded.
5774 42. If a MAIL command specified a SIZE value that was too large to fit into an
5775 int variable, the check against message_size_limit failed. Such values are
5776 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5777 day this will have to be increased, but I don't think I want to be around
5778 when emails are that large.
5785 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5786 iconv() is not standard in FreeBSD.
5788 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5789 IPv6 enabled. The observed symptom was a segmentation fault on return from
5790 the function os_common_find_running_interfaces() in src/os.c.
5792 3. In the check_special_case() function in daemon.c I had used "errno" as an
5793 argument name, which causes warnings on some systems. This was basically a
5794 typo, since it was named "eno" in the comments!
5796 4. The code that waits for the clock to tick (at a resolution of some fraction
5797 of a second) so as to ensure message-id uniqueness was always waiting for
5798 at least one whole tick, when it could have waited for less. [This is
5799 almost certainly not relevant at current processor speeds, where it is
5800 unlikely to ever wait at all. But we try to future-proof.]
5802 5. The function that sleeps for a time interval that includes fractions of a
5803 second contained a race. It did not block SIGALRM between setting the
5804 timer, and suspending (a couple of lines later). If the interval was short
5805 and the sigsuspend() was delayed until after it had expired, the suspension
5806 never ended. On busy systems this could lead to processes getting stuck for
5809 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5810 process, before it forks any delivery processes. The open lookup caching
5811 mechanism meant that the open file or database connection was passed into
5812 the delivery process. The problem was that delivery processes always tidy
5813 up cached lookup data. This could cause a problem for the next delivery
5814 process started by the queue runner, because the external queue runner
5815 process does not know about the closure. So the next delivery process
5816 still has data in the lookup cache. In the case of a file lookup, there was
5817 no problem because closing a file descriptor in a subprocess doesn't affect
5818 the parent. However, if the lookup was caching a connection to a database,
5819 the connection was closed, and the second delivery process was likely to
5820 see errors such as "PGSQL: query failed: server closed the connection
5821 unexpectedly". The problem has been fixed by closing all cached lookups
5822 in a queue runner before running a delivery process.
5824 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5825 seem to have the "const" qualifier which it has on other OS. I've
5828 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5829 *of the same type* (client or server) with the same public name that an
5830 error should be diagnosed.
5832 9. When Exim looked up a host name for an IP address, but failed to find the
5833 original IP address when looking up the host name (a safety check), it
5834 output the message "<ip address> does not match any IP for NULL", which was
5835 confusing, to say the least. The bug was that the host name should have
5836 appeared instead of "NULL".
5838 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5839 user that is built into the binary, and the -C or -D options is used, root
5840 privilege is dropped before the configuration file is read. In addition,
5841 logging is switched to stderr instead of the normal log files. If the
5842 configuration then re-defines the Exim user, the unprivileged environment
5843 is probably not what is expected, so Exim logs a panic warning message (but
5846 However, if deliver_drop_privilege is set, the unprivileged state may well
5847 be exactly what is intended, so the warning has been cut out in that case,
5848 and Exim is allowed to try to write to its normal log files.
5854 1. smtp_return_error_details was not giving details for temporary sender
5855 or receiver verification errors.
5857 2. Diagnose a configuration error if two authenticators have the same public
5860 3. Exim used not to create the message log file for a message until the first
5861 delivery attempt. This could be confusing when incoming messages were held
5862 for policy or load reasons. The message log file is now created at the time
5863 the message is received, and an initial "Received" line is written to it.
5865 4. The automatically generated man page for command line options had a minor
5866 bug that caused no ill effects; however, a more serious problem was that
5867 the procedure for building the man page automatically didn't always
5868 operate. Consequently, release 4.20 contains an out-of-date version. This
5869 shouldn't happen again.
5871 5. When building Exim with embedded Perl support, the script that builds the
5872 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5873 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5875 6. The freeze_tell option was not being used for messages that were frozen on
5876 arrival, either by an ACL or by local_scan().
5878 7. Added the smtp_incomplete_transaction log selector.
5880 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5881 was accepting AUTH without a new EHLO.
5883 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5884 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5887 10. Logging of TCP/IP connections (when configured) now happens in the main
5888 daemon process instead of the child process, so that the TCP/IP connection
5889 count is more accurate (but it can never be perfect).
5891 11. The use of "drop" in a nested ACL was not being handled correctly in the
5892 outer ACL. Now, if condition failure induced by the nested "drop" causes
5893 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5894 or "require"), the connection is dropped.
5896 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5897 that yield "discard" can now be used with an "accept" or a "discard" verb,
5898 but an error is generated for any others (because I can't see a useful way
5899 to define what should happen).
5901 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5902 are now processed in the same way as lines in the Exim configuration file.
5903 In particular, continuation lines are supported.
5905 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5907 15. Added -ti meaning -t -i.
5909 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5910 domains, and warn by logging if others are found.
5912 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5913 returned from GIFCONF when Exim is trying to find the list of interfaces on
5914 a host. The code in os.c has been modified to copy each ifreq to an aligned
5915 structure in all cases.
5917 Also, in some cases, the returned ifreq's were being copied to a 'struct
5918 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5919 means the last couple of bytes of an IPv6 address could be chopped if the
5920 ifreq contained only a normal sockaddr (14 bytes storage).
5922 18. Named domain lists were not supported in the hosts_treat_as_local option.
5923 An entry such as +xxxx was not recognized, and was treated as a literal
5926 19. Ensure that header lines added by a DATA ACL are included in the reject log
5927 if the ACL subsequently rejects the message.
5929 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5930 MD5 (which is deprecated).
5932 21. When testing a filter file using -bf, Exim was writing a message when it
5933 took the sender from a "From " line in the message, but it was not doing so
5934 when it took $return_path from a Return-Path: header line. It now does.
5936 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5937 with a valid header line field name (a series of printing characters
5938 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5940 23. Changed "disc" in the source to "disk" to conform to the documentation and
5941 the book and for uniformity.
5943 24. Ignore Sendmail's -Ooption=value command line item.
5945 25. When execve() failed while trying to run a command in a pipe transport,
5946 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5947 could be confused with a return value of 69 from the command itself. This
5948 has been changed to 127, the value the shell returns if it is asked to run
5949 a non-existent command. The wording for the related log line suggests a
5950 non-existent command as the problem.
5952 26. If received_header_text expands to an empty string, do not add a Received:
5953 header line to the message. (Well, it adds a token one on the spool, but
5954 marks it "old" so that it doesn't get used or transmitted.)
5956 27. Installed eximstats 1.28 (addition of -nt option).
5958 28. There was no check for failure on the call to getsockname() in the daemon
5959 code. This can fail if there is a shortage of resources on the system, with
5960 ENOMEM, for example. A temporary error is now given on failure.
5962 29. Contrary to the C standard, it seems that in some environments, the
5963 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5964 program. Exim now does this explicitly; it affects the formatting of
5965 timestamps using strftime().
5967 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5968 complaints. I've now initialized all the variables, to avoid this.
5970 32. Header lines added by a system filter were not being "seen" during
5971 transport-time rewrites.
5973 33. The info_callback() function passed to OpenSSL is set up with type void
5974 (*)(SSL *, int, int), as described somewhere. However, when calling the
5975 function (actually a macro) that sets it up, the type void(*)() is
5976 expected. I've put in a cast to prevent warnings from picky compilers.
5978 34. If a DNS black list lookup found a CNAME record, but there were no A
5979 records associated with the domain it pointed at, Exim crashed.
5981 35. If a DNS black list lookup returned more than one A record, Exim ignored
5982 all but the first. It now scans all returned addresses if a particular IP
5983 value is being sought. In this situation, the contents of the
5984 $dnslist_value variable are a list of all the addresses, separated by a
5987 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5988 to accept a host name and all its aliases if the forward lookup for any of
5989 them yielded the IP address of the incoming connection. Now it accepts only
5990 those names whose forward lookup yields the correct IP address. Any other
5991 names are discarded. This closes a loophole whereby a rogue DNS
5992 administrator could create reverse DNS records to break through a
5993 wildcarded host restriction in an ACL.
5995 37. If a user filter or a system filter that ran in a subprocess used any of
5996 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5997 the wrong values were passed to the pipe command ($thisaddress had the
5998 value of $0, $0 had the value of $1, etc). This bug was introduced by
5999 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6001 38. Improved the line breaking for long SMTP error messages from ACLs.
6002 Previously, if there was no break point between 40 and 75 characters, Exim
6003 left the rest of the message alone. Two changes have been made: (a) I've
6004 reduced the minimum length to 35 characters; (b) if it can't find a break
6005 point between 35 and 75 characters, it looks ahead and uses the first one
6006 that it finds. This may give the occasional overlong line, but at least the
6007 remaining text gets split now.
6009 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6010 file descriptors might be low, and that setting 1000 would always raise it.
6011 It turns out that in some environments, the limit is already over 1000 and
6012 that lowering it causes trouble. So now Exim takes care not to decrease it.
6014 40. When delivering a message, the value of $return_path is set to $sender_
6015 address at the start of routing (routers may change the value). By an
6016 oversight, this default was not being set up when an address was tested by
6017 -bt or -bv, which affected the outcome if any router or filter referred to
6020 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6021 log only when "message" or "log_message" are set. However, if one of the
6022 conditions was an address verification, or a call to a nested ACL, the
6023 messages generated by the underlying test were being passed through. This
6024 no longer happens. The underlying message is available in $acl_verify_
6025 message for both "message" and "log_message" expansions, so it can be
6026 passed through if needed.
6028 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6029 new expansion $bh_ to give the encoded byte string without charset
6030 translation. Translation happens only if iconv() is available; HAVE_ICONV
6031 indicates this at build time. HEADERS_CHARSET gives the charset to
6032 translate to; headers_charset can change it in the configuration, and
6033 "headers charset" can change it in an individual filter file.
6035 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6036 that creates RFC 2047 encoded "words" labels them as that charset instead
6037 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6038 expansion operator; (ii) when Exim creates a From: line for a local
6039 message; (iii) when a header line is rewritten to include a "phrase" part.
6041 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6042 buggy, causing it to skip the first lines of messages whose message ID
6043 ended in 'D'. This would not have bitten before Exim release 4.14, because
6044 message IDs were unlikely to end in 'D' before then. The effect was to have
6045 incorrect size information for certain domains.
6047 45. #include "config.h" was missing at the start of the crypt16.c module. This
6048 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6051 46. If there was a timeout during a "random" callout check, Exim treated it as
6052 a failure of the random address, and carried on sending RSET and the real
6053 address. If the delay was just some slowness somewhere, the response to the
6054 original RCPT would be taken as a response to RSET and so on, causing
6055 mayhem of various kinds.
6057 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6058 when I implemented it. It didn't allow for the fact that some option values
6059 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6060 the right test for positive values.
6062 48. Domain names in DNS records are case-independent. Exim always looks them up
6063 in lower case. Some resolvers return domain names in exactly the case they
6064 appear in the zone file, that is, they may contain uppercase letters. Not
6065 all resolvers do this - some return always lower case. Exim was treating a
6066 change of case by a resolver as a change of domain, similar to a widening
6067 of a domain abbreviation. This triggered its re-routing code and so it was
6068 trying to route what was effectively the same domain again. This normally
6069 caused routing to fail (because the router wouldn't handle the domain
6070 twice). Now Exim checks for this case specially, and just changes the
6071 casing of the domain that it ultimately uses when it transmits the message
6074 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6077 50. If a filter generated a file delivery with a non-absolute name (possible if
6078 no home directory exists for the router), the forbid_file option was not
6081 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6082 the existing equality matching.
6084 52. Exim was using ints instead of ino_t variables in some places where it was
6085 dealing with inode numbers.
6087 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6088 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6089 called TMPDIR, and if it finds it is different, it changes its value.
6091 54. The smtp_printf() function is now made available to local_scan() so
6092 additional output lines can be written before returning. There is also an
6093 smtp_fflush() function to enable the detection of a dropped connection.
6094 The variables smtp_input and smtp_batched_input are exported to
6097 55. Changed the default runtime configuration: the message "Unknown user"
6098 has been removed from the ACL, and instead placed on the localuser router,
6099 using the cannot_route_message feature. This means that any verification
6100 failures that generate their own messages won't get overridden. Similarly,
6101 the "Unrouteable address" message that was in the ACL for unverifiable
6102 relay addresses has also been removed.
6104 56. Added hosts_avoid_esmtp to the smtp transport.
6106 57. The exicyclog script was not checking for the esoteric option
6107 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6108 will work only if exicyclog is run under the appropriate euid.
6110 58. Following a discussion on the list, the rules by which Exim recognises line
6111 endings on incoming messages have been changed. The -dropcr and drop_cr
6112 options are now no-ops, retained only for backwards compatibility. The
6113 following line terminators are recognized: LF CRLF CR. However, special
6114 processing applies to CR:
6116 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6117 nor a local message in the state where . is a terminator.
6119 (ii) If a bare CR is encountered in a header line, an extra space is added
6120 after the line terminator so as not to end the header. The reasoning
6121 behind this is that bare CRs in header lines are most likely either
6122 to be mistakes, or people trying to play silly games.
6124 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6125 was being incorrectly given as 18 bytes larger than it should have been.
6126 This is a VOB (very old bug).
6128 60. This may never have affected anything current, but just in case it has:
6129 When the local host is found other than at the start of a list of hosts,
6130 the local host, those with the same MX, and any that follow, are discarded.
6131 When the list in question was part of a longer list of hosts, the following
6132 hosts (not currently being processed) were also being discarded. This no
6133 longer happens. I'm not sure if this situation could ever has previously
6136 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6139 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6140 header. This is something that is recommended in a new Internet Draft, and
6141 is something that is documented as being done by Sendmail. There are two
6142 possible values. For messages generated by the autoreply transport, Exim
6145 Auto-Submitted: auto-replied
6147 whereas for all other generated messages (e.g. bounces) it adds
6149 Auto-Submitted: auto-generated
6151 63. The "personal" condition in filters now includes a test for the
6152 Auto-Submitted: header. If it contains the string "auto-" the message it
6153 not considered personal.
6155 64. Added rcpt_include_affixes as a generic transport option.
6157 65. Added queue_only_override (default true).
6159 66. Added the syslog_duplication option.
6161 67. If what should have been the first header line of a message consisted of
6162 a space followed by a colon, Exim was mis-interpreting it as a header line.
6163 It isn't of course - it is syntactically invalid and should therefore be
6164 treated as the start of the message body. The misbehaviour could have
6165 caused a number of strange effects, including loss of data in subsequent
6166 header lines, and spool format errors.
6168 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6169 client host had authenticated. This control can now be exercised by an ACL
6170 for more flexibility.
6172 69. By default, callouts do not happen when testing with -bh. There is now a
6173 variant, -bhc, which does actually run the callout code, including
6174 consulting and updating the callout cache.
6176 70. Added support for saslauthd authentication, courtesy of Alexander
6179 71. If statvfs() failed on the spool or log directories while checking their
6180 size for availability, Exim confusingly gave the error "space shortage".
6181 Furthermore, in debugging mode it crashed with a floating point exception.
6182 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6183 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6184 serious problem, Exim now writes to the main and panic logs when this
6185 happens, with details of the failure. It then refuses to accept the
6186 incoming message, giving the message "spool directory problem" or "log
6187 directory problem" with a 421 code for SMTP messages.
6189 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6190 0, 1, and 2 exist, because some OS complain for execs without them (see
6191 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6192 descriptors. However, the code omitted to check that the open succeeded,
6193 causing mysterious errors if for some reason the permissions on /dev/null
6194 got screwed. Now Exim writes a message to the main and panic logs, and
6195 bombs out if it can't open /dev/null.
6197 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6198 interact so that it is all more flexible. It is supposed to remain
6199 backwards compatible. Also added extra_local_interfaces.
6201 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6202 to bomb out with an assertion failure - to the client this appears as a
6203 connection drop. This problem occurs in the part of the code that was taken
6204 from the Samba project. Fortunately, the assertion is in a very simple
6205 function, so I have fixed this by reproducing the function inline in the
6206 one place where it is called, and arranging for authentication to fail
6207 instead of killing the process with assert().
6209 75. The SPA client code was not working when the server requested OEM rather
6210 than Unicode encoding.
6212 76. Added code to make require_files with a specific uid setting more usable in
6213 the case where statting the file as root fails - usually a non-root-mounted
6214 NFS file system. When this happens and the failure is EACCES, Exim now
6215 forks a subprocess and does the per-uid checking as the relevant uid.
6217 77. Added process_log_path.
6219 78. If log_file_path was not explicitly set, a setting of check_log_space or
6220 check_log_inodes was ignored.
6222 79. If a space check for the spool or log partitions fails, the incident is now
6223 logged. Of course, in the latter case the data may get lost...
6225 80. Added the %p formatting code to string_format() so that it can be used to
6226 print addresses in debug_print(). Adjusted all the address printing in the
6227 debugging in store.c to use %p rather than %d.
6229 81. There was a concern that a line of code in smtp_in.c could overflow a
6230 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6231 initially expressed, the concern was not well-founded, because trailing
6232 spaces are removed early. However, if the trailing spaces were followed by
6233 a NULL, they did not get removed, so the overflow was possible. Two fixes
6236 (a) I re-wrote the offending code in a cleaner fashion.
6237 (b) If an incoming SMTP command contains a NULL character, it is rejected
6240 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6241 runs initgroups(), so that if the Exim user is in any additional groups,
6242 they will be used during message reception.
6248 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob