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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
48 TL/01 When running the test suite, the README says that variables such as
49 no_msglog_check are global and can be placed anywhere in a specific
50 test's script, however it was observed that placement needed to be near
51 the beginning for it to behave that way. Changed the runtest perl
52 script to read through the entire script once to detect and set these
53 variables, reset to the beginning of the script, and then run through
54 the script parsing/test process like normal.
56 TL/02 The BSD's have an arc4random API. One of the functions to induce
57 adding randomness was arc4random_stir(), but it has been removed in
58 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
59 function when detected.
61 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
62 cause callback expansion.
64 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
65 syntax errors in an expansion can be treated as a string instead of
66 logging or causing an error, due to the internal use of bool_lax
67 instead of bool when processing it.
69 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
70 server certificates when making smtp deliveries.
72 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
74 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
76 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
78 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
79 Merged patch from Sebastian Wiedenroth.
81 JH/05 Fix results-pipe from transport process. Several recipients, combined
82 with certificate use, exposed issues where response data items split
83 over buffer boundaries were not parsed properly. This eventually
84 resulted in duplicates being sent. This issue only became common enough
85 to notice due to the introduction of conection certificate information,
86 the item size being so much larger. Found and fixed by Wolfgang Breyha.
88 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
89 size buffer was used, resulting in syntax errors when an expansion
92 JH/07 Add support for directories of certificates when compiled with a GnuTLS
93 version 3.3.6 or later.
95 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
96 is EXPERIMENTAL_EVENT, the main-configuration and transport options
97 both become "event_action", the variables become $event_name, $event_data
98 and $event_defer_errno. There is a new variable $verify_mode, usable in
99 routers, transports and related events. The tls:cert event is now also
100 raised for inbound connections, if the main configuration event_action
103 TL/06 In test suite, disable OCSP for old versions of openssl which contained
104 early OCSP support, but no stapling (appears to be less than 1.0.0).
106 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
107 server certificate names available under the smtp transport option
108 "tls_verify_cert_hostname" now do not permit multi-component wildcard
111 JH/10 Time-related extraction expansions from certificates now use the main
112 option "timezone" setting for output formatting, and are consistent
113 between OpenSSL and GnuTLS compilations. Bug 1541.
115 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
118 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
119 include certificate info, eximon was claiming there were spoolfile
122 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
124 JH/14 Log delivery-related information more consistently, using the sequence
125 "H=<name> [<ip>]" wherever possible.
130 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
131 checkers that were complaining about end of non-void function with no
134 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
135 This was a regression intruduced in 4.83 by another bugfix.
137 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
139 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
140 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
146 TF/01 Correctly close the server side of TLS when forking for delivery.
148 When a message was received over SMTP with TLS, Exim failed to clear up
149 the incoming connection properly after forking off the child process to
150 deliver the message. In some situations the subsequent outgoing
151 delivery connection happened to have the same fd number as the incoming
152 connection previously had. Exim would try to use TLS and fail, logging
153 a "Bad file descriptor" error.
155 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
156 utilities have not been installed.
158 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
159 temporary space as the ACL may create new global variables.
161 TL/01 LDAP support uses per connection or global context settings, depending
162 upon the detected version of the libraries at build time.
164 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
165 to extract and use the src ip:port in logging and expansions as if it
166 were a direct connection from the outside internet. PPv2 support was
167 updated based on HAProxy spec change in May 2014.
169 JH/02 Add ${listextract {number}{list}{success}{fail}}.
171 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
172 Properly escape header and check for NULL return.
174 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
177 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
179 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
180 characters in header names, implemented as a verify condition.
181 Contributed by Michael Fischer v. Mollard.
183 TL/05 Rename SPF condition results err_perm and err_temp to standardized
184 results permerror and temperror. Previous values are deprecated but
185 still accepted. In a future release, err_perm and err_temp will be
186 completely removed, which will be a backward incompatibility if the
187 ACL tests for either of these two old results. Patch contributed by
188 user bes-internal on the mailing list.
190 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
192 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
193 selectors, in both main and reject logs.
195 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
198 JH/07 Add malware type "sock" for talking to simple daemon.
200 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
202 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
203 routers/transports under cutthrough routing.
205 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
206 numbers. Touch up "bool" conditional to keep the same definition.
208 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
210 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
212 JH/12 Expand items in router/transport headers_add or headers_remove lists
213 individually rather than the list as a whole. Bug 1452.
215 Required for reasonable handling of multiple headers_ options when
216 they may be empty; requires that headers_remove items with embedded
217 colons must have them doubled (or the list-separator changed).
219 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
220 view the policy declared in the DMARC record. Currently, $dmarc_status
221 is a combined value of both the record presence and the result of the
224 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
226 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
227 dnslookup router and the smtp transport (applying to the forward
230 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
231 of ldap servers used for a specific lookup. Patch provided by Heiko
234 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
235 New variable $lookup_dnssec_authenticated for observability.
237 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
238 Patch submitted by Lars Timman.
240 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
242 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
243 Requires trusted mode and valid format message id, aborts otherwise.
244 Patch contributed by Heiko Schlichting.
246 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
247 certextract with support for various fields. Bug 1358.
249 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
250 is requested by default, modifiable by smtp transport option
253 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
254 operate on certificate variables to give certificate fingerprints
255 Also new ${sha256:cert_variable}.
257 JH/23 The PRDR feature is moved from being Experimental into the mainline.
259 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
260 Christian Aistleitner.
262 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
264 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
265 file. Patch from Wolfgang Breyha.
267 JH/25 Expand the coverage of the delivery $host and $host_address to
268 client authenticators run in verify callout. Bug 1476.
270 JH/26 Port service names are now accepted for tls_on_connect_ports, to
271 align with daemon_smtp_ports. Bug 72.
273 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
274 support and error reporting did not work properly.
276 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
277 and is readable. Patch from Andrew Colin Kissa.
279 TL/14 Enhance documentation of ${run expansion and how it parses the
280 commandline after expansion, particularly in the case when an
281 unquoted variable expansion results in an empty value.
283 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
285 PP/02 Fix internal collision of T_APL on systems which support RFC3123
286 by renaming away from it. Addresses GH issue 15, reported by
289 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
291 TL/15 SECURITY: prevent double expansion in math comparison functions
292 (can expand unsanitized data). Not remotely exploitable.
299 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
301 PP/02 Make -n do something, by making it not do something.
302 When combined with -bP, the name of an option is not output.
304 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
307 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
308 $sender_host_name and config options to manage this, and basic check
311 PP/05 DSCP support for outbound connections and control modifier for inbound.
313 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
314 (Only plugin which currently uses this is kerberos4, which nobody should
315 be using, but we should make it available and other future plugins might
316 conceivably use it, even though it would break NAT; stuff *should* be
317 using channel bindings instead).
319 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
320 name; added for Sendmail compatibility; requires admin caller.
321 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
322 just ignore it); requires trusted caller.
323 Also parse but ignore: -Ac -Am -X<logfile>
326 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
328 TL/02 Add +smtp_confirmation as a default logging option.
330 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
331 Patch by Magnus Holmgren from 2007-02-20.
333 TL/04 Bugzilla 1281 - Spec typo.
334 Bugzilla 1283 - Spec typo.
335 Bugzilla 1290 - Spec grammar fixes.
337 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
339 TL/06 Add Experimental DMARC support using libopendmarc libraries.
341 TL/07 Fix an out of order global option causing a segfault. Reported to dev
342 mailing list by by Dmitry Isaikin.
344 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
346 JH/02 Support "G" suffix to numbers in ${if comparisons.
348 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
350 NM/01 Bugzilla 1197 - Spec typo
351 Bugzilla 1196 - Spec examples corrections
353 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
355 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
356 gnutls_enable_pkcs11, but renamed to more accurately indicate its
359 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
360 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
362 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
363 "acl {{name}{arg}...}", and optional args on acl condition
366 JH/05 Permit multiple router/transport headers_add/remove lines.
368 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
370 JH/07 Avoid using a waiting database for a single-message-only transport.
371 Performance patch from Paul Fisher. Bugzilla 1262.
373 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
376 JH/09 Add $headers_added variable, with content from use of ACL modifier
377 add_header (but not yet added to the message). Bugzilla 199.
379 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
380 Pulled from Bugzilla 817 by Wolfgang Breyha.
382 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
384 (nb: this is the same fix as in Exim 4.80.1)
386 JH/11 Add A= logging on delivery lines, and a client_set_id option on
389 JH/12 Add optional authenticated_sender logging to A= and a log_selector
392 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
394 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
395 advertise SMTP AUTH mechanism to us, instead of a generic
396 protocol violation error. Also, make Exim more robust to bad
397 data from the Dovecot auth socket.
399 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
401 When a queue runner is handling a message, Exim first routes the
402 recipient addresses, during which it prunes them based on the retry
403 hints database. After that it attempts to deliver the message to
404 any remaining recipients. It then updates the hints database using
407 So if a recipient address works intermittently, it can get repeatedly
408 deferred at routing time. The retry hints record remains fresh so the
409 address never reaches the final cutoff time.
411 This is a fairly common occurrence when a user is bumping up against
412 their storage quota. Exim had some logic in its local delivery code
413 to deal with this. However it did not apply to per-recipient defers
414 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
416 This change adds a proper retry rule check during routing so that the
417 final cutoff time is checked against the message's age. We only do
418 this check if there is an address retry record and there is not a
419 domain retry record; this implies that previous attempts to handle
420 the address had the retry_use_local_parts option turned on. We use
421 this as an approximation for the destination being like a local
422 delivery, as in LMTP.
424 I suspect this new check makes the old local delivery cutoff check
425 redundant, but I have not verified this so I left the code in place.
427 TF/02 Correct gecos expansion when From: is a prefix of the username.
429 Test 0254 submits a message to Exim with the header
433 When I ran the test suite under the user fanf2, Exim expanded
434 the header to contain my full name, whereas it should have added
435 a Resent-Sender: header. It erroneously treats any prefix of the
436 username as equal to the username.
438 This change corrects that bug.
440 GF/01 DCC debug and logging tidyup
441 Error conditions log to paniclog rather than rejectlog.
442 Debug lines prefixed by "DCC: " to remove any ambiguity.
444 TF/03 Avoid unnecessary rebuilds of lookup-related code.
446 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
447 Bug spotted by Jeremy Harris; was flawed since initial commit.
448 Would have resulted in OCSP responses post-SNI triggering an Exim
449 NULL dereference and crash.
451 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
453 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
454 Bug detection, analysis and fix by Samuel Thibault.
455 Bugzilla 1331, Debian bug #698092.
457 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
459 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
460 Server implementation by Todd Lyons, client by JH.
461 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
462 config variable "prdr_enable" controls whether the server
463 advertises the facility. If the client requests PRDR a new
464 acl_data_smtp_prdr ACL is called once for each recipient, after
465 the body content is received and before the acl_smtp_data ACL.
466 The client is controlled by bolth of: a hosts_try_prdr option
467 on the smtp transport, and the server advertisement.
468 Default client logging of deliveries and rejections involving
469 PRDR are flagged with the string "PRDR".
471 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
472 fclose(). Diagnosis by Todd Lyons.
474 PP/17 Update configure.default to handle IPv6 localhost better.
475 Patch by Alain Williams (plus minor tweaks).
478 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
479 This is now consistent with GnuTLS, and is now documented: the
480 previous undocumented portable approach to treating the option as
481 unset was to force an expansion failure. That still works, and
482 an empty string is now equivalent.
484 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
485 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
486 not performing validation itself.
488 PP/20 Added force_command boolean option to pipe transport.
489 Patch from Nick Koston, of cPanel Inc.
491 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
494 TF/04 Added udpsend ACL modifer and hexquote expansion operator
496 PP/21 Fix eximon continuous updating with timestamped log-files.
497 Broken in a format-string cleanup in 4.80, missed when I repaired the
498 other false fix of the same issue.
499 Report and fix from Heiko Schlichting.
502 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
503 Report from Prashanth Katuri.
505 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
506 It's SecureTransport, so affects any MacOS clients which use the
507 system-integrated TLS libraries, including email clients.
509 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
510 using a MIME ACL for non-SMTP local injection.
511 Report and assistance in diagnosis by Warren Baker.
513 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
515 JH/16 Fix comparisons for 64b. Bugzilla 1385.
517 TL/09 Add expansion variable $authenticated_fail_id to keep track of
518 last id that failed so it may be referenced in subsequent ACL's.
520 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
523 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
524 ldap library initialization, allowing self-signed CA's to be
525 used. Also properly sets require_cert option later in code by
526 using NULL (global ldap config) instead of ldap handle (per
527 session). Bug diagnosis and testing by alxgomz.
529 TL/12 Enhanced documentation in the ratelimit.pl script provided in
530 the src/util/ subdirectory.
532 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
533 renamed to Transport Post Delivery Action by Jeremy Harris, as
536 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
537 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
538 redis_servers = needs to be configured which will be used by the redis
539 lookup. Patch from Warren Baker, of The Packet Hub.
541 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
543 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
544 hostname or reverse DNS when processing a host list. Used suggestions
545 from multiple comments on this bug.
547 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
549 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
550 Missed a few lines, added it to make the runtest require no keyboard
553 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
554 contains upper case chars. Make router use caseful_local_part.
556 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
557 support when GnuTLS has been built with p11-kit.
563 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
565 This, or similar/improved, will also be change PP/11 of 4.82.
571 PP/01 Handle short writes when writing local log-files.
572 In practice, only affects FreeBSD (8 onwards).
573 Bugzilla 1053, with thanks to Dmitry Isaikin.
575 NM/01 Bugzilla 949 - Documentation tweak
577 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
580 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
582 PP/02 Implemented gsasl authenticator.
584 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
586 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
587 `pkg-config foo` for cflags/libs.
589 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
590 with rest of GSASL and with heimdal_gssapi.
592 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
593 `pkg-config foo` for cflags/libs for the TLS implementation.
595 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
596 properties get this fed in as external SSF. A number of robustness
597 and debugging improvements to the cyrus_sasl authenticator.
599 PP/08 cyrus_sasl server now expands the server_realm option.
601 PP/09 Bugzilla 1214 - Log authentication information in reject log.
602 Patch by Jeremy Harris.
604 PP/10 Added dbmjz lookup type.
606 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
608 PP/12 MAIL args handles TAB as well as SP, for better interop with
609 non-compliant senders.
610 Analysis and variant patch by Todd Lyons.
612 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
613 Bug report from Lars Müller <lars@samba.org> (via SUSE),
614 Patch from Dirk Mueller <dmueller@suse.com>
616 PP/13 tls_peerdn now print-escaped for spool files.
617 Observed some $tls_peerdn in wild which contained \n, which resulted
618 in spool file corruption.
620 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
621 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
622 or write after TLS renegotiation, which otherwise led to messages
625 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
626 as a tracking header (ie: a signed header comes before the signature).
627 Patch from Wolfgang Breyha.
629 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
630 comma-sep list; embedded commas doubled.
632 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
634 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
636 Report and patch from Dmitry Banschikov.
638 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
639 Removed SSL_clear() after SSL_new() which led to protocol negotiation
640 failures. We appear to now support TLS1.1+ with Exim.
642 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
643 lets Exim select keys and certificates based upon TLS SNI from client.
644 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
645 before an outbound SMTP session. New log_selector, +tls_sni.
647 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
648 NULL dereference. Report and patch from Alun Jones.
650 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
651 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
652 Not seeing resolver debug output on NetBSD, but suspect this is a
653 resolver implementation change.
655 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
656 Left warnings. Added "eximon gdb" invocation mode.
658 PP/21 Defaulting "accept_8bitmime" to true, not false.
660 PP/22 Added -bw for inetd wait mode support.
662 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
663 locate the relevant includes and libraries. Made this the default.
665 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
666 Bugzilla 1246, report and most of solution from Tomasz Kusy.
668 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
669 This may cause build issues on older platforms.
671 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
672 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
673 gnutls_require_mac & gnutls_require_protocols (no longer supported).
674 Added SNI support via GnuTLS too.
675 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
677 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
679 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
680 Applied second patch from Janne, implementing suggestion to default
681 multiple-strings-in-record handling to match SPF spec.
683 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
685 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
686 read-only, out of scope).
687 Patch from Wolfgang Breyha, report from Stuart Northfield.
689 PP/29 Fix three issues highlighted by clang analyser static analysis.
690 Only crash-plausible issue would require the Cambridge-specific
691 iplookup router and a misconfiguration.
692 Report from Marcin Mirosław.
694 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
696 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
697 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
698 As part of this, removing so much warning spew let me fix some minor
699 real issues in debug logging.
701 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
702 assignment on my part. Fixed.
704 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
705 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
706 Janne Snabb (who went above and beyond: thank you).
708 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
709 string otherwise requires a connection and a bunch more work and it's
710 relatively easy to get wrong. Should also expose TLS library linkage
713 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
714 64-bit ${eval} (JH/03).
716 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
717 GNU libc to support some of the 64-bit stuff, should not lead to
718 conflicts. Defined before os.h is pulled in, so if a given platform
719 needs to override this, it can.
721 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
722 protection layer was required, which is not implemented.
723 Bugzilla 1254, patch from Wolfgang Breyha.
725 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
726 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
727 tls_dhparam take prime identifiers. Also unbreak combination of
728 OpenSSL+DH_params+TLSSNI.
730 PP/39 Disable SSLv2 by default in OpenSSL support.
736 PP/01 Solaris build fix for Oracle's LDAP libraries.
737 Bugzilla 1109, patch from Stephen Usher.
739 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
741 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
744 TF/02 Fix a couple more cases where we did not log the error message
745 when unlink() failed. See also change 4.74-TF/03.
747 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
748 lock up or crash if it happened to be inside a call to libc when it
749 got a SIGUSR1 from exiwhat.
751 The SIGUSR1 handler appends the current process status to the process
752 log which is later printed by exiwhat. It used to use the general
753 purpose logging code to do this, but several functions it calls are
754 not safe for signals.
756 The new output code in the SIGUSR1 handler is specific to the process
757 log, and simple enough that it's easy to inspect for signal safety.
758 Removing some special cases also simplifies the general logging code.
759 Removing the spurious timestamps from the process log simplifies
762 TF/04 Improved ratelimit ACL condition.
764 The /noupdate option has been deprecated in favour of /readonly which
765 has clearer semantics. The /leaky, /strict, and /readonly update modes
766 are mutually exclusive. The update mode is no longer included in the
767 database key; it just determines when the database is updated. (This
768 means that when you upgrde Exim will forget old rate measurements.)
770 Exim now checks that the per_* options are used with an update mode that
771 makes sense for the current ACL. For example, when Exim is processing a
772 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
773 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
774 must specify per_mail/readonly. If you omit the update mode it defaults to
775 /leaky where that makes sense (as before) or /readonly where required.
777 The /noupdate option is now undocumented but still supported for
778 backwards compatibility. It is equivalent to /readonly except that in
779 ACLs where /readonly is required you may specify /leaky/noupdate or
780 /strict/noupdate which are treated the same as /readonly.
782 A useful new feature is the /count= option. This is a generalization
783 of the per_byte option, so that you can measure the throughput of other
784 aggregate values. For example, the per_byte option is now equivalent
785 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
787 The per_rcpt option has been generalized using the /count= mechanism
788 (though it's more complicated than the per_byte equivalence). When it is
789 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
790 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
791 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
792 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
793 Note that using per_rcpt with a non-readonly update mode in more than
794 one ACL will cause the recipients to be double-counted. (The per_mail
795 and per_byte options don't have this problem.)
797 The handling of very low rates has changed slightly. If the computed rate
798 is less than the event's count (usually one) then this event is the first
799 after a long gap. In this case the rate is set to the same as this event's
800 count, so that the first message of a spam run is counted properly.
802 The major new feature is a mechanism for counting the rate of unique
803 events. The new per_addr option counts the number of different
804 recipients that someone has sent messages to in the last time period. It
805 behaves like per_rcpt if all the recipient addresses are different, but
806 duplicate recipient addresses do not increase the measured rate. Like
807 the /count= option this is a general mechanism, so the per_addr option
808 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
809 example, measure the rate that a client uses different sender addresses
810 with the options per_mail/unique=$sender_address. There are further
811 details in the main documentation.
813 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
815 TF/06 Removed a few PCRE remnants.
817 TF/07 Automatically extract Exim's version number from tags in the git
818 repository when doing development or release builds.
820 PP/02 Raise smtp_cmd_buffer_size to 16kB.
821 Bugzilla 879. Patch from Paul Fisher.
823 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
824 Heavily based on revision 40f9a89a from Simon Arlott's tree.
827 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
829 PP/05 Variable $av_failed, true if the AV scanner deferred.
830 Bugzilla 1078. Patch from John Horne.
832 PP/06 Stop make process more reliably on build failure.
833 Bugzilla 1087. Patch from Heiko Schlittermann.
835 PP/07 Make maildir_use_size_file an _expandable_ boolean.
836 Bugzilla 1089. Patch from Heiko Schlittermann.
838 PP/08 Handle ${run} returning more data than OS pipe buffer size.
839 Bugzilla 1131. Patch from Holger Weiß.
841 PP/09 Handle IPv6 addresses with SPF.
842 Bugzilla 860. Patch from Wolfgang Breyha.
844 PP/10 GnuTLS: support TLS 1.2 & 1.1.
846 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
849 PP/11 match_* no longer expand right-hand-side by default.
850 New compile-time build option, EXPAND_LISTMATCH_RHS.
851 New expansion conditions, "inlist", "inlisti".
853 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
855 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
857 PP/14 fix log_write() format string regression from TF/03.
858 Bugzilla 1152. Patch from Dmitry Isaikin.
864 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
866 PP/02 Harmonised TLS library version reporting; only show if debugging.
867 Layout now matches that introduced for other libraries in 4.74 PP/03.
869 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
871 PP/04 New "dns_use_edns0" global option.
873 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
876 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
877 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
879 TK/01 Updated PolarSSL code to 0.14.2.
880 Bugzilla 1097. Patch from Andreas Metzler.
882 PP/07 Catch divide-by-zero in ${eval:...}.
885 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
888 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
889 format-string attack -- SECURITY: remote arbitrary code execution.
891 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
892 time unintentionally subject to list matching rules, letting the header
893 cause arbitrary Exim lookups (of items which can occur in lists, *not*
894 arbitrary string expansion). This allowed for information disclosure.
896 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
897 INT_MIN/-1 -- value coerced to INT_MAX.
903 NM/01 Workround for PCRE version dependency in version reporting
906 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
907 This fixes portability to compilers other than gcc, notably
908 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
910 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
911 makefiles for portability to HP-UX and POSIX correctness.
913 PP/01 Permit LOOKUP_foo enabling on the make command-line.
914 Also via indented variable definition in the Makefile.
915 (Debugging by Oliver Heesakkers).
917 PP/02 Restore caching of spamd results with expanded spamd_address.
918 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
920 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
921 Improves build reliability. Fix from: Frank Elsner
923 NM/02 Fix wide character breakage in the rfc2047 coding
924 Fixes bug 1064. Patch from Andrey N. Oktyabrski
926 NM/03 Allow underscore in dnslist lookups
927 Fixes bug 1026. Patch from Graeme Fowler
929 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
930 Code patches from Adam Ciarcinski of NetBSD.
932 NM/04 Fixed exiqgrep to cope with mailq missing size issue
935 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
936 is logged, to avoid truncation. Patch from John Horne.
938 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
939 Patch from Jakob Hirsch.
941 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
942 SQL string expansion failure details.
943 Patch from Andrey Oktyabrski.
945 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
946 Patch from Simon Arlott.
948 PP/09 New lookups functionality failed to compile on old gcc which rejects
949 extern declarations in function scope.
950 Patch from Oliver Fleischmann
952 PP/10 Use sig_atomic_t for flags set from signal handlers.
953 Check getgroups() return and improve debugging.
954 Fixed developed for diagnosis in bug 927 (which turned out to be
957 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
958 Patch from Mark Zealey.
960 PP/12 Bugzilla 1056: Improved spamd server selection.
961 Patch from Mark Zealey.
963 PP/13 Bugzilla 1086: Deal with maildir quota file races.
964 Based on patch from Heiko Schlittermann.
966 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
967 Patch from Uwe Doering, sign-off by Michael Haardt.
969 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
970 variable declaration deep within a block. Bug and patch from
973 PP/15 lookups-Makefile IRIX compatibilty coercion.
975 PP/16 Make DISABLE_DKIM build knob functional.
977 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
978 Patch by Simon Arlott
980 TF/03 Fix valgrind.h portability to C89 compilers that do not support
981 variable argument macros. Our copy now differs from upstream.
987 TF/01 Failure to get a lock on a hints database can have serious
988 consequences so log it to the panic log.
990 TF/02 Log LMTP confirmation messages in the same way as SMTP,
991 controlled using the smtp_confirmation log selector.
993 TF/03 Include the error message when we fail to unlink a spool file.
995 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
996 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
997 for maintaining out-of-tree patches for some time.
999 PP/01 Bugzilla 139: Documentation and portability issues.
1000 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1001 Handle per-OS dynamic-module compilation flags.
1003 PP/02 Let /dev/null have normal permissions.
1004 The 4.73 fixes were a little too stringent and complained about the
1005 permissions on /dev/null. Exempt it from some checks.
1006 Reported by Andreas M. Kirchwitz.
1008 PP/03 Report version information for many libraries, including
1009 Exim version information for dynamically loaded libraries. Created
1010 version.h, now support a version extension string for distributors
1011 who patch heavily. Dynamic module ABI change.
1013 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1014 privilege escalation vulnerability whereby the Exim run-time user
1015 can cause root to append content of the attacker's choosing to
1018 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1021 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1022 If dropping privileges for untrusted macros, we disabled normal logging
1023 on the basis that it would fail; for the Exim run-time user, this is not
1024 the case, and it resulted in successful deliveries going unlogged.
1025 Fixed. Reported by Andreas Metzler.
1031 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1032 only prepend for the Resent-* case. Fixes regression introduced in
1033 Exim 4.70 by NM/22 for Bugzilla 607.
1035 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1036 increasing numbers of administrators be bitten by this.
1038 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1040 PP/03 Bugzilla 994: added openssl_options main configuration option.
1042 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1044 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1046 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1048 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1049 without a peer certificate, leading to a segfault because of an
1050 assumption that peers always have certificates. Be a little more
1051 paranoid. Problem reported by Martin Tscholak.
1053 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1054 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1055 NB: ClamAV planning to remove STREAM in "middle of 2010".
1056 CL also introduces -bmalware, various -d+acl logging additions and
1057 more caution in buffer sizes.
1059 PP/09 Implemented reverse_ip expansion operator.
1061 PP/10 Bugzilla 937: provide a "debug" ACL control.
1063 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1065 PP/12 Bugzilla 973: Implement --version.
1067 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1069 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1071 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1073 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1074 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1075 ignore trailing whitespace.
1077 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1079 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1082 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1083 Notification from Dr Andrew Aitchison.
1085 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1086 ExtendedDetectionInfo response format.
1087 Notification from John Horne.
1089 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1092 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1093 XSL and documented dependency on system catalogs, with examples of how
1096 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1099 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1100 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1101 configuration file which is writeable by the Exim user or group.
1103 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1104 of configuration files to cover files specified with the -C option if
1105 they are going to be used with root privileges, not just the default
1108 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1109 option (effectively making it always true).
1111 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1112 files to be used while preserving root privileges.
1114 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1115 that rogue child processes cannot use them.
1117 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1118 run-time user, instead of root.
1120 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1121 Exim run-time user without dropping privileges.
1123 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1124 result string, instead of calling string_vformat() twice with the same
1127 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1128 for other users. Others should always drop root privileges if they use
1129 -C on the command line, even for a whitelisted configure file.
1131 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1133 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1139 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1140 $data_path, and $header_path variables; fixed documentation bugs and
1143 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1144 exipick to access non-standard spools, including the "frozen" queue
1147 NM/01 Bugzilla 965: Support mysql stored procedures.
1148 Patch from Alain Williams
1150 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1152 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1153 Patch from Andreas Metzler
1155 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1156 Patch from Kirill Miazine
1158 NM/05 Bugzilla 671: Added umask to procmail example.
1160 JJ/03 installed exipick 20100323.0, fixing doc bug
1162 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1163 directory. Notification and patch from Dan Rosenberg.
1165 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1167 TK/02 Improve log output when DKIM signing operation fails.
1169 MH/01 Treat the transport option dkim_domain as a colon separated
1170 list, not as a single string, and sign the message with each element,
1171 omitting multiple occurences of the same signer.
1173 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1174 Bugzilla 985, 986. Patch by Simon Arlott
1176 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1177 Patch by Simon Arlott
1179 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1180 MBX locking. Notification from Dan Rosenberg.
1186 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1188 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1190 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1192 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1194 NM/04 Bugzilla 915: Flag broken perl installation during build.
1200 TK/01 Added patch by Johannes Berg that expands the main option
1201 "spamd_address" if it starts with a dollar sign.
1203 TK/02 Write list of recipients to X-Envelope-Sender header when building
1204 the mbox-format spool file for content scanning (suggested by Jakob
1207 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1208 (http://www.dcc-servers.net/) support via dccifd. Activated by
1209 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1211 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1212 by Mark Daniel Reidel <mr@df.eu>.
1214 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1215 When building exim an external PCRE library is now needed -
1216 PCRE is a system library on the majority of modern systems.
1217 See entry on PCRE_LIBS in EDITME file.
1219 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1220 conversation. Added nologin parameter to request.
1221 Patch contributed by Kirill Miazine.
1223 TF/01 Do not log submission mode rewrites if they do not change the address.
1225 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1227 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1228 log files in place. Contributed by Roberto Lima.
1230 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1232 TF/03 Bugzilla 615: When checking the local_parts router precondition
1233 after a local_part_suffix or local_part_prefix option, Exim now
1234 does not use the address's named list lookup cache, since this
1235 contains cached lookups for the whole local part.
1237 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1238 Robert Millan. Documentation is in experimental-spec.txt.
1240 TF/04 Bugzilla 668: Fix parallel build (make -j).
1242 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1244 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1245 Patch provided by Jan Srzednicki.
1247 TF/05 Leading white space used to be stripped from $spam_report which
1248 wrecked the formatting. Now it is preserved.
1250 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1251 that they are available at delivery time.
1253 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1255 TF/08 TLS error reporting now respects the incoming_interface and
1256 incoming_port log selectors.
1258 TF/09 Produce a more useful error message if an SMTP transport's hosts
1259 setting expands to an empty string.
1261 NM/06 Bugzilla 744: EXPN did not work under TLS.
1262 Patch provided by Phil Pennock.
1264 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1265 Patch provided by Richard Godbee.
1267 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1268 acl_smtp_notquit, added index entry.
1270 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1271 Patch provided by Eugene Bujak.
1273 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1274 accept(). Patch provided by Maxim Dounin.
1276 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1277 Patch provided by Phil Pennock.
1279 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1281 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1282 Patch provided by Brad "anomie" Jorsch.
1284 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1285 Patch provided by Dean Brooks.
1287 TK/05 Add native DKIM support (does not depend on external libraries).
1289 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1290 Patch provided by Graeme Fowler.
1292 NM/16 Bugzilla 851: Documentation example syntax fix.
1294 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1296 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1299 NM/19 Bugzilla 745: TLS version reporting.
1300 Patch provided by Phil Pennock.
1302 NM/20 Bugzilla 167: bool: condition support.
1303 Patch provided by Phil Pennock.
1305 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1306 clients. Patch provided by Phil Pennock.
1308 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1309 Patch provided by Brad "anomie" Jorsch.
1311 NM/23 Bugzilla 687: Fix misparses in eximstats.
1312 Patch provided by Heiko Schlittermann.
1314 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1315 Patch provided by Heiko Schlittermann.
1317 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1318 plus update to original patch.
1320 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1322 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1323 Patch provided by David Brownlee.
1325 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1327 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1329 NM/30 Bugzilla 888: TLS documentation bugfixes.
1331 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1333 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1334 Unlike the original bugzilla I have changed all shell scripts in src tree.
1336 NM/33 Bugzilla 898: Transport filter timeout fix.
1337 Patch by Todd Rinaldo.
1339 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1340 Patch by Serge Demonchaux.
1342 NM/35 Bugzilla 39: Base64 decode bug fixes.
1343 Patch by Jakob Hirsch.
1345 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1347 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1349 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1351 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1357 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1358 ALT-N's libdkim that I have put here:
1359 http://duncanthrax.net/exim-experimental/
1361 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1362 were called 'true' and it seems that C99 defines that as a reserved
1363 keyword to be used with 'bool' variable types. That means you could
1364 not include C99-style headers which use bools without triggering
1365 build errors in sieve.c.
1367 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1368 as mailq or other aliases. Changed the --help handling significantly
1369 to do whats expected. exim_usage() emits usage/help information.
1371 SC/01 Added the -bylocaldomain option to eximstats.
1373 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1375 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1377 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1383 PH/01 Another patch from the Sieve maintainer.
1385 PH/02 When an IPv6 address is converted to a string for single-key lookup
1386 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1387 dots are used instead of colons so that keys in lsearch files need not
1388 contain colons. This was done some time before quoting was made available
1389 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1390 (notated using the quote facility) so as to distinguish them from IPv4
1391 keys. This meant that lookups for IP addresses in host lists did not work
1392 for iplsearch lookups.
1394 This has been fixed by arranging for IPv6 addresses to be expressed with
1395 colons if the lookup type is iplsearch. This is not incompatible, because
1396 previously such lookups could never work.
1398 The situation is now rather anomolous, since one *can* have colons in
1399 ordinary lsearch keys. However, making the change in all cases is
1400 incompatible and would probably break a number of configurations.
1402 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1405 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1406 conversion specification without a maximum field width, thereby enabling
1407 a rogue spamd server to cause a buffer overflow. While nobody in their
1408 right mind would setup Exim to query an untrusted spamd server, an
1409 attacker that gains access to a server running spamd could potentially
1410 exploit this vulnerability to run arbitrary code as the Exim user.
1412 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1413 $primary_hostname instead of what libspf2 thinks the hosts name is.
1415 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1416 a directory entry by the name of the lookup key. Previously, if a
1417 symlink pointed to a non-existing file or a file in a directory that
1418 Exim lacked permissions to read, a lookup for a key matching that
1419 symlink would fail. Now it is enough that a matching directory entry
1420 exists, symlink or not. (Bugzilla 503.)
1422 PH/03 The body_linecount and body_zerocount variables are now exported in the
1425 PH/04 Added the $dnslist_matched variable.
1427 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1428 This means they are set thereafter only if the connection becomes
1431 PH/06 Added the client_condition to authenticators so that some can be skipped
1432 by clients under certain conditions.
1434 PH/07 The error message for a badly-placed control=no_multiline_responses left
1435 "_responses" off the end of the name.
1437 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1439 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1440 (without spaces) instead of just copying the configuration text.
1442 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1444 PH/11 Added $max_received_linelength.
1446 PH/12 Added +ignore_defer and +include_defer to host lists.
1448 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1449 way in which PCRE > 7.0 is built.
1451 PH/14 Implemented queue_only_load_latch.
1453 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1454 MAIL command. The effect was to mangle the value on 64-bit systems.
1456 PH/16 Another patch from the Sieve maintainer.
1458 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1460 PH/18 If a system quota error occurred while trying to create the file for
1461 a maildir delivery, the message "Mailbox is full" was not appended to the
1462 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1463 only to a quota excession during the actual writing of the file.
1465 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1466 characters?) which causes problems in log lines. The DN values are now
1467 passed through string_printing() before being added to log lines.
1469 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1470 and InterBase are left for another time.)
1472 PH/21 Added message_body_newlines option.
1474 PH/22 Guard against possible overflow in moan_check_errorcopy().
1476 PH/23 POSIX allows open() to be a macro; guard against that.
1478 PH/24 If the recipient of an error message contained an @ in the local part
1479 (suitably quoted, of course), incorrect values were put in $domain and
1480 $local_part during the evaluation of errors_copy.
1486 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1487 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1490 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1491 issue a MAIL command.
1493 PH/02 In an ACL statement such as
1495 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1497 if a client was not listed at all, or was listed with a value other than
1498 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1499 the condition was not true (as it should be), so access was not denied.
1500 The bug was that the ! inversion was incorrectly passed on to the second
1501 item. This has been fixed.
1503 PH/03 Added additional dnslists conditions == and =& which are different from
1504 = and & when the dns lookup returns more than one IP address.
1506 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1507 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1509 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1510 FSYNC, which compiles an option called disable_fsync that allows for
1511 bypassing fsync(). The documentation is heavily laced with warnings.
1513 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1515 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1516 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1517 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1518 including adding "make clean"; (3) Added -fPIC when compiling the test
1519 dynamically loaded module, to get rid of a warning.
1521 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1522 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1523 = 0s. The bug is otherwise harmless.
1525 PH/07 There was a bug in the dovecot authenticator such that the value of
1526 $auth1 could be overwritten, and so not correctly preserved, after a
1527 successful authentication. This usually meant that the value preserved by
1528 the server_setid option was incorrect.
1530 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1532 PH/09 Installed PCRE release 7.0.
1534 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1535 run for batched SMTP input. It is now run at the start of every message
1536 in the batch. While fixing this I discovered that the process information
1537 (output by running exiwhat) was not always getting set for -bs and -bS
1538 input. This is fixed, and it now also says "batched" for BSMTP.
1540 PH/11 Added control=no_pipelining.
1542 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1543 patch, slightly modified), and move the expansion of helo_data till after
1544 the connection is made in the smtp transport (so it can use these
1547 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1549 PH/14 Added log_selector = +pid.
1551 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1553 PH/16 Add ${if forany and ${if forall.
1555 PH/17 Added dsn_from option to vary the From: line in DSNs.
1557 PH/18 Flush SMTP output before performing a callout, unless control =
1558 no_callout_flush is set.
1560 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1561 was true (the default) a successful delivery failed to delete the retry
1562 item, thus causing premature timeout of the address. The bug is now
1565 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1567 PH/21 Long custom messages for fakedefer and fakereject are now split up
1568 into multiline reponses in the same way that messages for "deny" and
1569 other ACL rejections are.
1571 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1572 with slight modification.
1574 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1575 draft, changing the syntax and factoring some duplicate code".
1577 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1578 for deliveries of the second and subsequent messages over the same SMTP
1581 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1582 ${reduce, with only minor "tidies".
1584 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1586 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1587 expansion side effects.
1589 PH/27 When a message times out after an over-quota error from an Exim-imposed
1590 quota, the bounce message says "mailbox is full". This message was not
1591 being given when it was a system quota that was exceeded. It now should
1594 MH/03 Made $recipients available in local_scan(). local_scan() already has
1595 better access to the recipient list through recipients_list[], but
1596 $recipients can be useful in postmaster-provided expansion strings.
1598 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1599 in the case of a MAIL command with additional options following the
1600 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1601 were accidentally chopped off.
1603 PH/29 SMTP synchronization checks are implemented when a command is read -
1604 there is a check that no more input is waiting when there shouldn't be
1605 any. However, for some commands, a delay in an ACL can mean that it is
1606 some time before the response is written. In this time, more input might
1607 arrive, invalidly. So now there are extra checks after an ACL has run for
1608 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1609 pipelining has not been advertised.
1611 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1613 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1614 returned in the SMTP response when smtp_return_error_details was false.
1615 This has been fixed.
1617 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1618 instead of the C I/O that was originally supplied, because problems were
1619 reported on Solaris.
1621 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1622 Exim which did not show up earlier: it was assuming that a call to
1623 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1624 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1625 was a macro that became an assignment, so it seemed to work. This has
1626 changed to a proper function call with a void return, hence the compile
1627 error. Exim's code has been fixed.
1629 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1632 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1634 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1636 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1637 because it turns out to be too big or there is a timeout) there is no
1638 "Completed" line in the log. When some messages of this type were
1639 selected by exigrep, they were listed as "not completed". Others were
1640 picked up by some special patterns. I have improved the selection
1641 criteria to be more general.
1643 PH/38 The host_find_failed option in the manualroute router can now be set
1644 to "ignore", to completely ignore a host whose IP address cannot be
1645 found. If all hosts are ignored, the behaviour is controlled by the new
1646 host_all_ignored option.
1648 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1649 homing or because of multiple MX records with /mx) generated more than
1650 one IP address, and the following item turned out to be the local host,
1651 all the secondary addresses of the first item were incorrectly removed
1652 from the list, along with the local host and any following hosts (which
1653 is what is supposed to happen).
1655 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1656 whoever called Exim into the -H file. In the case of the daemon it was
1657 behaving confusingly. When first started, it used values for whoever
1658 started the daemon, but after a SIGHUP it used the Exim user (because it
1659 calls itself on a restart). I have changed the code so that it now always
1662 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1663 message are rejected with the same error (e.g. no authentication or bad
1664 sender address), and a DATA command is nevertheless sent (as can happen
1665 with PIPELINING or a stupid MUA), the error message that was given to the
1666 RCPT commands is included in the rejection of the DATA command. This is
1667 intended to be helpful for MUAs that show only the final error to their
1670 PH/42 Another patch from the Sieve maintainer.
1672 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1673 Eximstats - Fixed some broken HTML links and added missing column headers
1675 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1676 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1678 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1680 PH/43 Yet another patch from the Sieve maintainer.
1682 PH/44 I found a way to check for a TCP/IP connection going away before sending
1683 the response to the final '.' that terminates a message, but only in the
1684 case where the client has not sent further data following the '.'
1685 (unfortunately, this is allowed). However, in many cases there won't be
1686 any further data because there won't be any more messages to send. A call
1687 to select() can be used: if it shows that the input is "ready", there is
1688 either input waiting, or the socket has been closed. An attempt to read
1689 the next input character can distinguish the two cases. Previously, Exim
1690 would have sent an OK response which the client would never have see.
1691 This could lead to message repetition. This fix should cure that, at
1692 least in a lot of common cases.
1694 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1695 advertised in response to EHLO.
1701 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1702 fixed by 4.65/MH/01 (is this a record?) are fixed:
1704 (i) An empty string was always treated as zero by the numeric comparison
1705 operators. This behaviour has been restored.
1707 (ii) It is documented that the numeric comparison operators always treat
1708 their arguments as decimal numbers. This was broken in that numbers
1709 starting with 0 were being interpreted as octal.
1711 While fixing these problems I realized that there was another issue that
1712 hadn't been noticed. Values of message_size_limit (both the global option
1713 and the transport option) were treated as octal if they started with 0.
1714 The documentation was vague. These values are now always treated as
1715 decimal, and I will make that clear in the documentation.
1721 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1722 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1725 MH/01 Don't check that the operands of numeric comparison operators are
1726 integers when their expansion is in "skipping" mode (fixes bug
1727 introduced by 4.64-PH/07).
1729 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1730 child addresses, Exim now panics and dies. Previously, because the count
1731 is held in a short int, deliveries were likely to be lost. As such a
1732 large number of recipients for a single message is ridiculous
1733 (performance will be very, very poor), I have chosen to impose a limit
1734 rather than extend the field.
1740 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1741 leftover -K file (the existence of which was triggered by #402).
1742 While we were at it, introduced process PID as part of the -K
1743 filename. This should rule out race conditions when creating
1746 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1747 processing considerably. Previous code took too long for large mails,
1748 triggering a timeout which in turn triggers #401.
1750 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1751 in the DK code in transports.c. sendfile() is not really portable,
1752 hence the _LINUX specificness.
1754 TF/01 In the add_headers option to the mail command in an Exim filter,
1755 there was a bug that Exim would claim a syntax error in any
1756 header after the first one which had an odd number of characters
1759 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1760 callout verification, Exim cached a "reject" for the entire domain. This
1761 is correct for most verifications, but it is not correct for a recipient
1762 verification with use_sender or use_postmaster set, because in that case
1763 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1764 case of MAIL FROM:<> rejection from other early rejections (e.g.
1765 rejection of HELO). When verifying a recipient using a non-null MAIL
1766 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1767 Whatever the result of the callout, the value of the domain cache is
1768 left unchanged (for any other kind of callout, getting as far as trying
1769 RCPT means that the domain itself is ok).
1771 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1774 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1775 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1776 session, the daemon ignores ECONNECT errors and logs others; it now
1777 ignores EPIPE as well.
1779 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1780 (quoted-printable decoding).
1782 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1783 later the small subsequent patch to fix an introduced bug.
1785 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1787 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1789 PH/08 An error is now given if message_size_limit is specified negative.
1791 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1792 to be given (somewhat) arbitrary names.
1794 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1797 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1798 miscellaneous code fixes
1800 PH/10 Added the log_reject_target ACL modifier to specify where to log
1803 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1804 hostname. This is wrong, because it relates to the incoming message (and
1805 probably the interface on which it is arriving) and not to the outgoing
1806 callout (which could be using a different interface). This has been
1807 changed to use the value of the helo_data option from the smtp transport
1808 instead - this is what is used when a message is actually being sent. If
1809 there is no remote transport (possible with a router that sets up host
1810 addresses), $smtp_active_hostname is used.
1812 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1813 tweaks were necessary in order to get it to work (see also 21 below):
1814 (a) The code assumed that strncpy() returns a negative number on buffer
1815 overflow, which isn't the case. Replaced with Exim's string_format()
1817 (b) There were several signed/unsigned issues. I just did the minimum
1818 hacking in of casts. There is scope for a larger refactoring.
1819 (c) The code used strcasecmp() which is not a standard C function.
1820 Replaced with Exim's strcmpic() function.
1821 (d) The code set only $1; it now sets $auth1 as well.
1822 (e) A simple test gave the error "authentication client didn't specify
1823 service in request". It would seem that Dovecot has changed its
1824 interface. Fortunately there's a specification; I followed it and
1825 changed what the client sends and it appears to be working now.
1827 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1830 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1831 address A is aliased to B and C, where B exists and C does not. Without
1832 -v the output is "A verified" because verification stops after a
1833 successful redirection if more than one address is generated. However,
1834 with -v the child addresses are also verified. Exim was outputting "A
1835 failed to verify" and then showing the successful verification for C,
1836 with its parentage. It now outputs "B failed to verify", showing B's
1837 parentage before showing the successful verification of C.
1839 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1840 look up a TXT record in a specific list after matching in a combined
1843 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1844 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1845 they consult the DNS. I had assumed they would set it the way they
1846 wanted; and indeed my experiments on Linux seem to show that in some
1847 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1848 To be on the safe side, however, I have now made the interface to
1849 host_find_byname() similar to host_find_bydns(), with an argument
1850 containing the DNS resolver options. The host_find_byname() function now
1851 sets these options at its start, just as host_find_bydns() does. The smtp
1852 transport options dns_qualify_single and dns_search_parents are passed to
1853 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1854 of host_find_byname() use the default settings of RES_DEFNAMES
1855 (qualify_single) but not RES_DNSRCH (search_parents).
1857 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1858 spool_read_header() do less string testing, by means of a preliminary
1859 switch on the second character of optional "-foo" lines. (This is
1860 overdue, caused by the large number of possibilities that now exist.
1861 Originally there were few.) While I was there, I also converted the
1862 str(n)cmp tests so they don't re-test the leading "-" and the first
1863 character, in the hope this might squeeze out yet more improvement.
1865 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1866 flag allowing group syntax was set by the header_syntax check but not
1867 turned off, possible causing trouble later; (2) The flag was not being
1868 set at all for the header_verify test, causing "group"-style headers to
1869 be rejected. I have now set it in this case, and also caused header_
1870 verify to ignore an empty address taken from a group. While doing this, I
1871 came across some other cases where the code for allowing group syntax
1872 while scanning a header line wasn't quite right (mostly, not resetting
1873 the flag correctly in the right place). These bugs could have caused
1874 trouble for malformed header lines. I hope it is now all correct.
1876 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1877 with the "reply" argument non-NULL. The code, however (which originally
1878 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1879 but it didn't always do it. This confused somebody who was copying the
1880 code for some other use. I have removed all the tests.
1882 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1883 feature that was used to support insecure browsers during the U.S. crypto
1884 embargo. It requires special client support, and Exim is probably the
1885 only MTA that supported it -- and would never use it because real RSA is
1886 always available. This code has been removed, because it had the bad
1887 effect of slowing Exim down by computing (never used) parameters for the
1888 RSA_EXPORT functionality.
1890 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1891 authenticator to fail if there's a tab character in the incoming data
1892 (there should never be unless someone is messing about, as it's supposed
1893 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1894 if the connection is using TLS or if the remote IP is the same as the
1895 local IP, and the "valid-client-cert option" if a client certificate has
1898 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1899 authenticators. This can be used for authorization after authentication
1900 succeeds. (In the case of plaintext, it servers for both authentication
1903 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1904 if any retry times were supplied.
1906 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1907 connection, where there is no client IP address to check. In this
1908 situation, the verify now always succeeds.
1910 PH/25 Applied John Jetmore's -Mset patch.
1912 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1914 PH/27 In a string expansion for a processed (not raw) header when multiple
1915 headers of the same name were present, leading whitespace was being
1916 removed from all of them, but trailing whitespace was being removed only
1917 from the last one. Now trailing whitespace is removed from each header
1918 before concatenation. Completely empty headers in a concatenation (as
1919 before) are ignored.
1921 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1922 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1924 PH/29 [Removed. This was a change that I later backed out, and forgot to
1925 correct the ChangeLog entry (that I had efficiently created) before
1926 committing the later change.]
1928 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1929 address errors (4xx response to RCPT) over the same connection as other
1930 messages routed to the same hosts. Such deliveries are always "forced",
1931 so retry times are not inspected. This resulted in far too many retries
1932 for the affected addresses. The effect occurred only when there were more
1933 hosts than the hosts_max_try setting in the smtp transport when it had
1934 the 4xx errors. Those hosts that it had tried were not added to the list
1935 of hosts for which the message was waiting, so if all were tried, there
1936 was no problem. Two fixes have been applied:
1938 (i) If there are any address or message errors in an SMTP delivery, none
1939 of the hosts (tried or untried) are now added to the list of hosts
1940 for which the message is waiting, so the message should not be a
1941 candidate for sending over the same connection that was used for a
1942 successful delivery of some other message. This seems entirely
1943 reasonable: after all the message is NOT "waiting for some host".
1944 This is so "obvious" that I'm not sure why it wasn't done
1945 previously. Hope I haven't missed anything, but it can't do any
1946 harm, as the worst effect is to miss an optimization.
1948 (ii) If, despite (i), such a delivery is accidentally attempted, the
1949 routing retry time is respected, so at least it doesn't keep
1950 hammering the server.
1952 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1953 in ${readsocket because some servers need this prod.
1955 PH/32 Added some extra debug output when updating a wait-xxx database.
1957 PH/33 The hint "could be header name not terminated by colon", which has been
1958 given for certain expansion errors for a long time, was not being given
1959 for the ${if def:h_colon_omitted{... case.
1961 PH/34 The spec says: "With one important exception, whenever a domain list is
1962 being scanned, $domain contains the subject domain." There was at least
1963 one case where this was not true.
1965 PH/35 The error "getsockname() failed: connection reset by peer" was being
1966 written to the panic log as well as the main log, but it isn't really
1967 panic-worthy as it just means the connection died rather early on. I have
1968 removed the panic log writing for the ECONNRESET error when getsockname()
1971 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1972 runs only) independently of the message's sender address. This meant
1973 that, if the 4xx error was in fact related to the sender, a different
1974 message to the same recipient with a different sender could confuse
1975 things. In particualar, this can happen when sending to a greylisting
1976 server, but other circumstances could also provoke similar problems.
1977 I have changed the default so that the retry time for these errors is now
1978 based a combination of the sender and recipient addresses. This change
1979 can be overridden by setting address_retry_include_sender=false in the
1982 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1983 remote server are returned as part of bounce messages. This was not
1984 happening for LMTP over a pipe (the lmtp transport), but now it is the
1985 same for both kinds of LMTP.
1987 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1988 in header lines that were in fact CNAMEs. This is no longer the case.
1990 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1991 and queue runs started by the daemon processed all messages. This has
1992 been fixed so that -R and -S can now usefully be given with -q<time>.
1994 PH/40 Import PCRE release 6.7 (fixes some bugs).
1996 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1998 PH/42 Give an error if -q is specified more than once.
2000 PH/43 Renamed the variables $interface_address and $interface_port as
2001 $received_ip_address and $received_port, to make it clear that these
2002 values apply to message reception, and not to the outgoing interface when
2003 a message is delivered. (The old names remain recognized, of course.)
2005 PH/44 There was no timeout on the connect() call when using a Unix domain
2006 socket in the ${readsocket expansion. There now is.
2008 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2009 be meaningful with "accept".
2011 SC/01 Eximstats V1.43
2012 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2014 SC/02 Eximstats V1.44
2015 Use a glob alias rather than an array ref in the generated
2016 parser. This improves both readability and performance.
2018 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2019 Collect SpamAssassin and rejection statistics.
2020 Don't display local sender or destination tables unless
2021 there is data to show.
2022 Added average volumes into the top table text output.
2024 SC/04 Eximstats V1.46
2025 Collect data on the number of addresses (recipients)
2026 as well as the number of messages.
2028 SC/05 Eximstats V1.47
2029 Added 'Message too big' to the list of mail rejection
2030 reasons (thanks to Marco Gaiarin).
2032 SC/06 Eximstats V1.48
2033 Mainlog lines which have GMT offsets and are too short to
2034 have a flag are now skipped.
2036 SC/07 Eximstats V1.49 (Alain Williams)
2037 Added the -emptyok flag.
2039 SC/08 Eximstats V1.50
2040 Fixes for obtaining the IP address from reject messages.
2042 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2043 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2044 whitesspace changes from 4.64-PH/27
2046 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2049 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2050 are found, allow negative numbers in numeric criteria)
2052 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2054 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2055 to match changes made in 4.64-PH/43
2057 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2059 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2060 "domain" is actually an IP address, and force a failure. This locks out
2061 those revolvers/nameservers that support "A-for-A" lookups, in
2062 contravention of the specifications.
2064 PH/48 When a host name was looked up from an IP address, and the subsequent
2065 forward lookup of the name timed out, the host name was left in
2066 $sender_host_name, contrary to the specification.
2068 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2069 restricted to single-key lookups, Exim was not diagnosing an error if
2070 * or *@ was used with a query-style lookup.
2072 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2074 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2075 long ago, but noone interested enough thought of it. Let's just say that
2076 the "1.1" means that there are some new functions that weren't there at
2077 some point in the past.
2079 PH/51 Error processing for expansion failure of helo_data from an smtp
2080 transport during callout processing was broken.
2082 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2083 tested/used via the -bh/-bhc/-bs options.
2085 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2086 bug, fixed in subsequent PCRE releases).
2088 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2089 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2091 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2097 SC/01 Use a glob alias rather than an array ref in eximstats generated
2098 parser. This improves both readability and performance.
2100 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2101 Don't display local sender or destination tables in eximstats unless
2102 there is data to show.
2103 Added average volumes into the eximstats top table text output.
2105 SC/03 Collect data on the number of addresses (recipients) as well
2106 as the number of messages in eximstats.
2108 TF/01 Correct an error in the documentation for the redirect router. Exim
2109 does (usually) call initgroups() when daemonizing.
2111 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2112 with consistent privilege compared to when running as a daemon.
2114 TF/03 Note in the spec that $authenticated_id is not set for local
2115 submissions from trusted users.
2117 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2118 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2120 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2121 by adding some example configuration directives to the default
2122 configuration file. A little bit of work is required to uncomment the
2123 directives and define how usernames and passwords are checked, but
2124 there is now a framework to start from.
2126 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2127 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2128 without this. I don't know how relevant this is to other LDAP libraries.
2130 PH/02 Add the verb name to the "unknown ACL verb" error.
2132 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2134 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2136 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2137 directory not expanded when it should be if an expanded home directory
2138 was set for the address (which is overridden by the transport).
2140 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2143 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2144 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2145 because it is too late at that time, and has no effect.
2147 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2148 security issue with \' (bugzilla #107). I could not use the
2149 PQescapeStringConn() function, because it needs a PGconn value as one of
2152 PH/08 When testing addresses using -bt, indicate those final addresses that
2153 are duplicates that would not cause an additional delivery. At least one
2154 person was confused, thinking that -bt output corresponded to deliveries.
2155 (Suppressing duplicates isn't a good idea as you lose the information
2156 about possibly different redirections that led to the duplicates.)
2158 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2159 systems where poll() doesn't work, in particular OS X.
2161 PH/10 Added more information to debugging output for retry time not reached.
2163 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2164 operations in malware.c.
2166 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2169 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2170 syslog_duplication set false, log lines that would normally be written
2171 both the the main log and to the reject log were not written to syslog at
2174 PH/14 In the default configuration, change the use of "message" in ACL warn
2175 statements to "add_header".
2177 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2178 not followed by a command (e.g. "seen endif").
2180 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2181 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2184 PH/17 Added extra conditions to the default value of delay_warning_condition
2188 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2189 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2190 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2193 The Auto-Submitted: and various List- headers are standardised, whereas I
2194 don't think Precedence: ever was.
2196 PH/18 Refactored debugging code in route_finduser() to show more information,
2197 in particular, the error code if getpwnam() issues one.
2199 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2200 This is apparently needed in addition to the PH/07 change above to avoid
2201 any possible encoding problems.
2203 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2204 but not after initializing Perl.
2206 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2207 output them only if debugging. By default they are written stderr,
2208 apparently, which is not desirable.
2210 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2213 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2216 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2218 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2219 authenticated or an ident call has been made. Suppress the default
2220 values for $authenticated_id and $authenticated_sender (but permit -oMai
2221 and -oMas) when testing with -bh.
2223 PH/24 Re-jigged the order of the tests in the default configuration so that the
2224 tests for valid domains and recipients precede the DNS black list and CSA
2225 tests, on the grounds that those ones are more expensive.
2227 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2228 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2229 This bug exists in every version of Exim that I still have, right back to
2232 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2233 However, an attempt to turn on case-sensitivity in a regex key by
2234 including (?-i) didn't work because the subject string was already
2235 lowercased, and the effects were non-intuitive. It turns out that a
2236 one-line patch can be used to allow (?-i) to work as expected.
2242 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2243 other effects) broke the use of negated acl sub-conditions.
2245 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2248 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2249 "Deny" causes Exim to reject the incoming connection with a 554 error.
2250 Unfortunately, if there is a major crisis, such as a disk failure,
2251 tcp-wrappers gives "deny", whereas what one would like would be some
2252 kind of temporary error. A kludge has been added to help with this.
2253 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2254 554 error is used if errno is still zero or contains ENOENT (which occurs
2255 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2258 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2260 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2261 errors. Otherwise a message that provokes a temporary error (when other
2262 messages do not) can cause a whole host to time out.
2264 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2265 addresses were routed directly to files or pipes from a redirect router.
2266 File deliveries just didn't batch; pipe deliveries might have suffered
2269 PH/06 A failure to get a lock for a hints database would erroneously always say
2270 "Failed to get write lock", even when it was really a read lock.
2272 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2273 of 0600. This has been changed to use the value of the lockfile_mode
2274 option (which defaults to 0600).
2276 PH/08 Applied small patch from the Sieve maintainer.
2278 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2279 folder from quota calculations, a direct delivery into this folder messed
2280 up the contents of the maildirsize file. This was because the regex was
2281 used only to exclude .Trash (or whatever) when the size of the mailbox
2282 was calculated. There was no check that a delivery was happening into an
2283 excluded directory. This bug has been fixed by ignoring all quota
2284 processing for deliveries into excluded directories.
2286 PH/10 Added the maildirfolder_create_regex option to appendfile.
2292 PH/01 The code for finding all the local interface addresses on a FreeBSD
2293 system running IPv6 was broken. This may well have applied to all BSD
2294 systems, as well as to others that have similar system calls. The broken
2295 code found IPv4 interfaces correctly, but gave incorrect values for the
2296 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2297 that it would not match correctly against @[] and not recognize the IPv6
2300 PH/02 The ipliteral router was not recognizing addresses of the form user@
2301 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2303 PH/03 Added disable_ipv6.
2305 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2306 decoded form, because it is most often used to construct To: headers
2307 lines in autoreplies, and the decoded form may well be syntactically
2308 invalid. However, $reply_address has leading white space removed, and all
2309 newlines turned into spaces so that the autoreply transport does not
2312 PH/05 If group was specified without a user on a router, and no group or user
2313 was specified on a transport, the group from the router was ignored.
2315 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2316 for visible compile-time settings that can be used to change these
2317 numbers, for those that want even more. Backwards compatibility with old
2318 spool files has been maintained. However, going back to a previous Exim
2319 release will lost any variables that are in spool files.
2321 PH/07 Two small changes when running in the test harness: increase delay when
2322 passing a TCP/IP connection to a new process, in case the original
2323 process has to generate a bounce, and remove special handling of
2324 127.0.0.2 (sic), which is no longer necessary.
2326 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2327 be the same on different OS.
2329 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2332 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2333 whether --show-vars was specified or not
2335 JJ/02 exipick: Added support for new ACL variable spool format introduced
2338 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2339 syntactically invalid From: or Reply-to: line, and a filter used this to
2340 generate an autoreply, and therefore failed to obtain an address for the
2341 autoreply, Exim could try to deliver to a non-existent relative file
2342 name, causing unrelated and misleading errors. What now happens is that
2343 it logs this as a hard delivery error, but does not attempt to create a
2346 PH/11 The exinext utility has a -C option for testing purposes, but although
2347 the given file was scanned by exinext itself; it wasn't being passed on
2348 when Exim was called.
2350 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2351 an end-of-file indication when reading a command response.
2353 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2354 compiled. In many other places in Exim, IPv6 addresses are always
2355 recognized, so I have changed this. It also means that IPv4 domain
2356 literals of the form [IPV4:n.n.n.n] are now always recognized.
2358 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2359 used if the router is not running as root, for example, when verifying at
2360 ACL time, or when using -bh. The debugging output from this situation was
2361 non-existent - all you got was a failure to exec. I have made two
2364 (a) Failures to set uid/gid, the current directory, or a process leader
2365 in a subprocess such as that created by queryprogram now generate
2366 suitable debugging ouput when -d is set.
2368 (b) The queryprogram router detects when it is not running as root,
2369 outputs suitable debugging information if -d is set, and then runs
2370 the subprocess without attempting to change uid/gid.
2372 PH/15 Minor change to Makefile for building test_host (undocumented testing
2375 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2376 additional section of a DNS packet that returns MX or SRV records.
2377 Instead, it always explicitly searches for A/AAAA records. This avoids
2378 major problems that occur when a DNS server includes only records of one
2379 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2380 fixed another bug: if SRV records were looked up and the corresponding
2381 address records were *not* found in the additional section, the port
2382 values from the SRV records were lost.
2384 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2385 using the correct key (the original address) when searching the retry
2386 rules in order to find which one to use for generating the retry hint.
2388 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2389 adding the default one. Similarly, if it contains a Reply-To: header, the
2390 errors_reply_to option, if set, is not used.
2392 PH/19 When calculating a retry time, Exim used to measure the "time since
2393 failure" by looking at the "first failed" field in the retry record. Now
2394 it does not use this if it is later than than the arrival time of the
2395 message. Instead it uses the arrival time. This makes for better
2396 behaviour in cases where some deliveries succeed, thus re-setting the
2397 "first failed" field. An example is a quota failure for a huge message
2398 when small messages continue to be delivered. Without this change, the
2399 "time since failure" will always be short, possible causing more frequent
2400 delivery attempts for the huge message than are intended.
2401 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2403 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2404 $1, $2, $3) because the numerical variables can be reset during some
2405 expansion items (e.g. "match"), thereby losing the authentication data.
2407 PH/21 Make -bV show the size of off_t variables so that the test suite can
2408 decide whether to run tests for quotas > 2G.
2410 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2411 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2412 filecount value is greater than 2G or if a quota value is greater than 2G
2413 on a system where the size of off_t is not greater than 4, a panic error
2416 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2417 never match. The debug and -bh output now contains an explicit error
2418 message indicating a malformed IPv4 address or mask.
2420 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2421 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2422 PH/23 above applies.
2424 PH/25 Do not write to syslog when running in the test harness. The only
2425 occasion when this arises is a failure to open the main or panic logs
2426 (for which there is an explicit test).
2428 PH/26 Added the /no_tell option to "control=freeze".
2430 PH/27 If a host name lookup failed very early in a connection, for example, if
2431 the IP address matched host_lookup and the reverse lookup yielded a name
2432 that did not have a forward lookup, an error message of the form "no IP
2433 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2434 could be logged. Now it outputs the IP address instead of "NULL".
2436 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2437 allows the sender and the authenticated sender to be set when
2438 submitting a message from within Exim. Since child_open_exim() is
2439 documented for local_scan(), the new function should be too.
2441 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2442 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2443 results in an empty string is now treated as unset.
2445 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2447 PH/31 Added sender_verify_fail logging option.
2449 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2450 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2451 tidied the source and removed it altogether.
2453 PH/33 When a queue run was abandoned because the load average was too high, a
2454 log line was always written; now it is written only if the queue_run log
2455 selector is set. In addition, the log line for abandonment now contains
2456 information about the queue run such as the pid. This is always present
2457 in "start" and "stop" lines but was omitted from the "abandon" line.
2459 PH/34 Omit spaces between a header name and the colon in the error message that
2460 is given when verify = headers_syntax fails (if there are lots of them,
2461 the message gets confusing).
2463 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2464 names, as there are now some PTR records that contain slashes. This check
2465 is only to protect against broken name servers that fall over on strange
2466 characters, so the fact that it applies to all lookups doesn't matter.
2468 PH/36 Now that the new test suite is complete, we can remove some of the
2469 special code in Exim that was needed for the old test suite. For example,
2470 sorting DNS records because real resolvers return them in an arbitrary
2471 order. The new test suite's fake resolver always returns records in the
2474 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2475 bounces) except when queue_only is set, to avoid logging races between
2476 the different processes.
2478 PH/38 Panic-die if .include specifies a non-absolute path.
2480 PH/39 A tweak to the "H" retry rule from its user.
2482 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2483 a label. They prevented compilation on older perls.
2485 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2486 a warning to be raised on newish perls.
2488 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2489 on queue. Changes to match documented behaviour of showing count of
2490 messages matching specified criteria.
2492 PH/40 Changed the default ident timeout from 30s to 5s.
2494 PH/41 Added support for the use of login_cap features, on those BSD systems
2495 that have them, for controlling the resources used by pipe deliveries.
2497 PH/42 The content-scanning code uses fopen() to create files in which to put
2498 message data. Previously it was not paying any attention to the mode of
2499 the files. Exim runs with umask(0) because the rest of the code creates
2500 files with open(), and sets the required mode explicitly. Thus, these
2501 files were ending up world-writeable. This was not a big issue, because,
2502 being within the spool directory, they were not world-accessible. I have
2503 created a function called modefopen, which takes an additional mode
2504 argument. It sets umask(777), creates the file, chmods it to the required
2505 mode, then resets the umask. All the relevant calls to fopen() in the
2506 content scanning code have been changed to use this function.
2508 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2509 to 24 hours. This avoids potential overflow problems when processing G
2510 and H retry rules. I suspect nobody ever tinkers with this value.
2512 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2514 PH/45 When the plaintext authenticator is running as a client, the server's
2515 challenges are checked to ensure they are valid base64 strings. By
2516 default, the authentication attempt is cancelled if an invalid string is
2517 received. Setting client_ignore_invalid_base64 true ignores these errors.
2518 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2519 they are received. Thus, the responses can be made to depend on the
2520 challenges. If an invalid string is ignored, an empty string is placed in
2523 PH/46 Messages that are created by the autoreply transport now contains a
2524 References: header, in accordance with RFCs 2822 and 3834.
2526 PH/47 Added authenticated_sender_force to the smtp transport.
2528 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2530 PH/49 Installed latest patch from the Sieve maintainer.
2532 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2533 was also set, the appendfile transport was unnecessarily scanning a
2534 directory of message files (e.g. for maildir delivery) to find the count
2535 of files (along with the size), even though it did not need this
2536 information. It now does the scan only if it needs to find either the
2537 size of the count of files.
2539 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2541 PH/52 Two bugs concerned with error handling when the smtp transport is
2544 (i) Exim was not creating retry information for temporary errors given
2545 for individual recipients after the DATA command when the smtp transport
2546 was used in LMTP mode. This meant that they could be retried too
2547 frequently, and not timed out correctly.
2549 (ii) Exim was setting the flag that allows error details to be returned
2550 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2551 recipients that were returned after the DATA command.
2553 PH/53 This is related to PH/52, but is more general: for any failing address,
2554 when detailed error information was permitted to be returned to the
2555 sender, but the error was temporary, then after the final timeout, only
2556 "retry timeout exceeded" was returned. Now it returns the full error as
2557 well as "retry timeout exceeded".
2559 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2560 do this, and (what is worse) MTAs that accept it.
2562 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2563 will now be deprecated.
2565 PH/56 New os.c-cygwin from the Cygwin maintainer.
2567 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2568 formats (previously only available in exim formats via -bpr, -bpru,
2569 and -bpra. Now also available in native and exiqgrep formats)
2571 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2572 with very large, slow to parse queues
2574 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2576 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2578 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2579 responses to authentication challenges, though it was showing the
2580 challenges; (ii) I've removed the CR characters from the debug output for
2583 PH/58 Allow for the insertion of a newline as well as a space when a string
2584 is turned into more than one encoded-word during RFC 2047 encoding. The
2585 Sieve code now uses this.
2587 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2588 data_4xx, lost_connection, tls_required.
2590 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2591 message was being sent as an SMTP response.
2593 PH/61 Add -l and -k options to exicyclog.
2595 PH/62 When verifying, if an address was redirected to one new address, so that
2596 verification continued, and the new address failed or deferred after
2597 having set something in $address_data, the value of $address_data was not
2598 passed back to the ACL. This was different to the case when no
2599 redirection occurred. The value is now passed back in both cases.
2601 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2602 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2603 use login_cap.h, so on its own it isn't the distinguishing feature. The
2604 new name refers directly to the setclassresources() function.
2606 PH/65 Added configuration files for NetBSD3.
2608 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2610 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2611 is preferred over IPv4.
2613 PH/68 The bounce_return_message and bounce_return_body options were not being
2614 honoured for bounces generated during the reception of non-SMTP messages.
2615 In particular, this applied to messages rejected by the ACL. This bug has
2616 been fixed. However, if bounce_return_message is true and bounce_return_
2617 body is false, the headers that are returned for a non-SMTP message
2618 include only those that have been read before the error was detected.
2619 (In the case of an ACL rejection, they have all been read.)
2621 PH/69 The HTML version of the specification is now built in a directory called
2622 spec_html instead of spec.html, because the latter looks like a path with
2623 a MIME-type, and this confuses some software.
2625 PH/70 Catch two compiler warnings in sieve.c.
2627 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2628 function verify_get_ident() calls ip_connect() to connect a socket, but
2629 if the "connect()" function timed out, ip_connect() used to close the
2630 socket. However, verify_get_ident() also closes the socket later, and in
2631 between Exim writes to the log, which may get opened at this point. When
2632 the socket was closed in ip_connect(), the log could get the same file
2633 descriptor number as the socket. This naturally causes chaos. The fix is
2634 not to close the socket in ip_connect(); the socket should be closed by
2635 the function that creates it. There was only one place in the code where
2636 this was missing, in the iplookup router, which I don't think anybody now
2637 uses, but I've fixed it anyway.
2639 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2640 well as to direct DNS lookups. Otherwise the handling of names in host
2641 lists is inconsistent and therefore confusing.
2647 PH/01 Two changes to the default runtime configuration:
2649 (1) Move the checks for relay_from_hosts and authenticated clients from
2650 after to before the (commented out) DNS black list checks.
2652 (2) Add control=submission to the relay_from_hosts and authenticated
2653 clients checks, on the grounds that messages accepted by these
2654 statements are most likely to be submissions.
2656 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2658 (1) Generate an error if the third argument for the ${prvs expansion is
2661 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2664 (3) Reset the variables that are obtained from the first argument of
2665 ${prvscheck and used in the second argument before leaving the code,
2666 because their memory is reclaimed, so using them afterwards may do
2669 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2670 one (it's much easier than Tom thought :-).
2672 (5) Because of (4), we can now allow for the use of $prvscheck_result
2673 inside the third argument.
2675 PH/03 For some reason, the default setting of PATH when running a command from
2676 a pipe transport was just "/usr/bin". I have changed it to
2679 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2680 anything to be listed in the output from -bV.
2682 PH/05 When a filter generated an autoreply, the entire To: header line was
2683 quoted in the delivery log line, like this:
2685 => >A.N.Other <ano@some.domain> <original@ddress> ...
2687 This has been changed so that it extracts the operative address. There
2688 may be more than one such address. If so, they are comma-separated, like
2691 => >ano@some.domain,ona@other.domain <original@ddress> ...
2693 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2694 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2695 not being looked up in the rDNS to get a host name, Exim was showing the
2696 IP address twice in Received: lines, even though the IP addresses were
2697 identical. For example:
2699 Received: from [1.2.3.4] (helo=[1.2.3.4])
2701 However, if the real host name was known, it was omitting the HELO data
2702 if it matched the actual IP address. This has been tidied up so that it
2703 doesn't show the same IP address twice.
2705 PH/07 When both +timestamp and +memory debugging was on, the value given by
2706 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2707 called by the debug printing, thereby overwriting the timestamp buffer.
2708 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2710 PH/08 When the original message was included in an autoreply transport, it
2711 always said "this is a copy of the message, including all the headers",
2712 even if body_only or headers_only was set. It now gives an appropriate
2715 PH/09 Applied a patch from the Sieve maintainer which:
2717 o fixes some comments
2718 o adds the (disabled) notify extension core
2719 o adds some debug output for the result of if/elsif tests
2720 o points to the current vacation draft in the documentation
2721 and documents the missing references header update
2725 o fixes a bug in processing the envelope test (when testing
2726 multiple envelope elements, the last element determinted the
2729 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2730 Electronic Mail") by including:
2732 Auto-submitted: auto-generated
2734 in the messages that it generates (bounce messages and others, such as
2735 warnings). In the case of bounce messages for non-SMTP mesages, there was
2736 also a typo: it was using "Auto_submitted" (underscore instead of
2737 hyphen). Since every message generated by Exim is necessarily in response
2738 to another message, thes have all been changed to:
2740 Auto-Submitted: auto-replied
2742 in accordance with these statements in the RFC:
2744 The auto-replied keyword:
2746 - SHOULD be used on messages sent in direct response to another
2747 message by an automatic process,
2749 - MUST NOT be used on manually-generated messages,
2751 - MAY be used on Delivery Status Notifications (DSNs) and Message
2752 Disposition Notifications (MDNs),
2754 - MUST NOT be used on messages generated by automatic or periodic
2755 processes, except for messages which are automatic responses to
2758 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2759 to the default Received: header definition.
2761 PH/12 Added log selector acl_warn_skipped (default on).
2763 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2764 variables because (a) they are in the wrong storage pool and (b) even if
2765 they were copied, it wouldn't work properly because of the caching.
2767 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2768 checking when decoding. Apparently there are clients that generate
2769 overlong encoded strings. Why am I not surprised?
2771 PH/15 If the first argument of "${if match_address" was not empty, but did not
2772 contain an "@" character, Exim crashed. Now it writes a panic log message
2773 and treats the condition as false.
2775 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2777 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2778 extension "envelope-auth". The code is finished and in agreement with
2779 other implementations, but there is no documentation so far and in fact,
2780 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2781 not changing the active code.
2783 Print executed "if" and "elsif" statements when debugging is used. This
2784 helps a great deal to understand what a filter does.
2786 Document more things not specified clearly in RFC3028. I had all this
2787 sorted out, when out of a sudden new issues came to my mind. Oops."
2789 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2792 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2793 canonical form (as documented). However, after a host name lookup from
2794 the IP address, check_host() was doing a simple string comparison with
2795 addresses acquired from the DNS when checking that the found name did
2796 have the original IP as one of its addresses. Since any found IPv6
2797 addresses are likely to be in abbreviated form, the comparison could
2798 fail. Luckily, there already exists a function for doing the comparison
2799 by converting both addresses to binary, so now that is used instead of
2800 the text comparison.
2802 PH/20 There was another similar case to PH/19, when a complete host name was
2803 given in a host list; looking up its IP address could give an abbreviated
2804 form, whereas the current host's name might or might not be abbreviated.
2805 The same fix has been applied.
2811 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2812 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2815 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2816 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2818 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2820 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2821 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2822 Also, for "Auto-Submitted", treat anything other than "no" as
2823 non-personal, in accordance with RFC 3834. (Previously it treated
2824 anything starting "auto-" as non-personal.)
2826 TF/01 The control=submission/name=... option had a problem with syntax
2827 errors if the name included a slash character. The /name= option
2828 now slurps the rest of the string, so it can include any characters
2829 but it must come last in the list of options (after /sender_retain
2832 PH/05 Some modifications to the interface to the fake nameserver for the new
2840 TK/01 Added the "success_on_redirect" address verification option. See
2841 NewStuff for rationale and an example.
2843 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2845 PH/02 Patch to exigrep to allow it to work on syslog lines.
2847 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2848 fread() to skip over the body file's header line, because in Cygwin the
2849 header line is locked and is inaccessible.
2851 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2852 co-exist for some time) to make it clear that it is the Exim ID that is
2853 referenced, not the Message-ID: header line.
2855 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2856 string_format() function, because snprintf() does not exist on all
2859 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2860 ${stat: expansion item.
2862 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2863 protocol synchronization error", to keep the pedants happy.
2865 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2866 well as for IRIX systems, when gcc is being used. See the host.c source
2869 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2871 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2874 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2875 transport and to the smtp transport in LMTP mode.
2877 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2879 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2880 run a filter in a subprocess. This could lead to confusion in subsequent
2881 lookups in the parent process. There should also be a search_tidyup() at
2882 the end of the subprocess.
2884 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2885 only if the host matched helo_try_verify_hosts, which caused the
2886 verification to occur when the EHLO/HELO command was issued. The ACL just
2887 tested the remembered result. Now, if a previous verification attempt has
2888 not happened, "verify = helo" does it there and then.
2890 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2892 TK/03 Fix log output including CR from clamd.
2894 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2895 exist provoked a memory error which could cause a segfault.
2897 PH/15 Installed PCRE 6.2
2899 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2901 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2902 of the problem. Specifically, suggested +O2 rather than +O1 for the
2905 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2907 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2908 record, the port was not correctly logged when the outgoing_port log
2909 selector was set (it logged the transort's default port).
2911 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2912 fallback_hosts, and "hosts" in the smtp transport.
2914 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2915 host errors such as "Connection refused".
2917 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2918 authentication with radiusclient 0.4.9:
2920 - Error returned from rc_read_config was caught wrongly
2921 - Username/password not passed on to radius server due to wrong length.
2923 The presumption is that some radiusclient API changes for 4.51/PH/17
2924 were not taken care of correctly. The code is still untested by me (my
2925 Linux distribution still has 0.3.2 of radiusclient), but it was
2926 contributed by a Radius user.
2928 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2929 expanding the "port" option of the smtp transport.
2931 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2932 while reading a MIME header. Thanks to Tom Hughes for a patch.
2934 PH/24 Include config.h inside local_scan.h so that configuration settings are
2937 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2938 that in an ACL for RCPT (for example), you can examine exactly what was
2941 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2942 commands, but it was not correctly comparing the address with the actual
2943 client host address. Thus, it would show the EHLO address in Received:
2944 header lines when this was not necessary.
2946 PH/27 Added the % operator to ${eval:}.
2948 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2949 it should be ignoring failures (because with -C, for example, it has lost
2950 privilege). It wasn't ignoring creation failures other than "already
2953 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2956 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2957 because an input line was too long, either on its own, or by virtue of
2958 too many continuations, the temporary file was not being removed, and the
2959 return code was incorrect.
2961 PH/31 Missing "BOOL" in function definition in filtertest.c.
2963 PH/32 Applied Sieve patches from the maintainer.
2965 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2967 PH/33 Added "verify = not_blind".
2969 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2970 Local/Makefile (with some defaults set). These are used in built scripts
2971 such as exicyclog, but they have never been used in the exim_install
2972 script (though there are many overriding facilities there). I have
2973 arranged that the exim_install script now takes note of these two
2976 PH/35 Installed configuration files for Dragonfly.
2978 PH/36 When a locally submitted message by a trusted user did not contain a
2979 From: header, and the sender address was obtained from -f or from an SMTP
2980 MAIL command, and the trusted user did not use -F to supply a sender
2981 name, $originator_name was incorrectly used when constructing a From:
2982 header. Furthermore, $originator_name was used for submission mode
2983 messages from external hosts without From: headers in a similar way,
2984 which is clearly wrong.
2986 PH/37 Added control=suppress_local_fixups.
2988 PH/38 When log_selector = +received_sender was set, and the addition of the
2989 sender made the log line's construction buffer exactly full, or one byte
2990 less than full, an overflow happened when the terminating "\n" was
2993 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2994 when the result of a list match is failure because a DNS lookup failed.
2996 PH/40 RM_COMMAND is now used in the building process.
2998 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2999 the "build-* directories that it finds.
3001 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3002 address in a domain literal was a prefix of an interface address.
3004 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3005 when verifying a sender address, unless rewrite_headers is false.
3007 PH/44 Wrote a long comment about why errors_to addresses are verified as
3008 recipients, not senders.
3010 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3011 the ratelimit ACL was added.
3013 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3015 PH/46 Added extra information about PostgreSQL errors to the error string.
3017 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3018 avoiding the need to install special zones in a real server. This is
3019 backwards compatible; if it can't find the fake resolver, it drops back.
3020 Thus, both old and new test suites can be run.
3022 TF/02 Added util/ratelimit.pl
3024 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3025 clock is set back in time.
3027 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3028 Candler <B.Candler@pobox.com>.
3030 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3031 OK for addresses that are the result of redirections.
3033 PH/48 A number of further additions for the benefit of the new test suite,
3034 including a fake gethostbyname() that interfaces to the fake DNS resolver
3037 TF/06 The fix for widen_domains has also been applied to qualify_single and
3038 search_parents which are the other dnslookup options that can cause
3041 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3044 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3046 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3047 service" if it encountered an MX record with an empty target hostname.
3048 The message is now "an MX or SRV record indicated no SMTP service".
3050 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3051 if the DNS of the sending site is misconfigured. This is quite a
3052 common situation. This change restores the behaviour of treating a
3053 helo verification defer as a failure.
3055 PH/51 If self=fail was set on a router, the bounce message did not include the
3056 actual error message.
3062 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3064 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3065 command itself ended in error, the underlying message about the transport
3066 filter timeout was being overwritten with the pipe command error. Now the
3067 underlying error message should be appended to the second error message.
3069 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3071 PH/02 Reduce the amount of output that "make" produces by default. Full output
3072 can still be requested.
3074 PH/03 The warning log line about a condition test deferring for a "warn" verb
3075 was being output only once per connection, rather than after each
3076 occurrence (because it was using the same function as for successful
3077 "warn" verbs). This seems wrong, so I have changed it.
3079 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3080 it should not have, which might have caused a crash in the right
3081 circumstances, but probably never did.
3083 PH/04 Installed a modified version of Tony Finch's patch to make submission
3084 mode fix the return path as well as the Sender: header line, and to
3085 add a /name= option so that you can make the user's friendly name appear
3088 TF/03 Added the control = fakedefer ACL modifier.
3090 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3091 Mark Lowes for thorough testing.
3093 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3095 TK/03 Merged latest SRS patch from Miles Wilton.
3097 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3098 with the definition in sysexits.h (which is #included earlier).
3099 Fortunately, Exim does not actually use EX_OK. The code used to try to
3100 preserve the sysexits.h value, by assumimg that macro definitions were
3101 scanned for macro replacements. I have been disabused of this notion,
3102 so now the code just undefines EX_OK before #including unistd.h.
3104 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3105 in the smtp transport. When a block could not be written in a single
3106 write() function, the timeout was being re-applied to each part-write.
3107 This seems wrong - if the receiver was accepting one byte at a time it
3108 would take for ever. The timeout is now adjusted when this happens. It
3109 doesn't have to be particularly precise.
3111 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3112 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3114 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3115 without a domain if the check to <postmaster@domain> fails.
3117 SC/01 Eximstats: added -xls and the ability to specify output files
3118 (patch written by Frank Heydlauf).
3120 SC/02 Eximstats: use FileHandles for outputing results.
3122 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3124 SC/04 Eximstats: fixed display of large numbers with -nvr option
3126 SC/05 Eximstats: fixed merging of reports with empty tables.
3128 SC/06 Eximstats: added the -include_original_destination flag
3130 SC/07 Eximstats: removed tabs and trailing whitespace.
3132 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3134 TK/06 MBOX spool code: Add real "From " MBOX separator line
3135 so the .eml file is really in mbox format (even though
3136 most programs do not really care). Patch from Alex Miller.
3138 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3139 The latter is generated from $received_to and is only set if the
3140 message has one envelope recipient. SA can use these headers,
3141 obviously out-of-the-box. Patch from Alex Miller.
3143 PH/08 The ${def test on a variable was returning false if the variable's
3144 value was "0", contrary to what the specification has always said!
3145 The result should be true unless the variable is empty.
3147 PH/09 The syntax error of a character other than { following "${if
3148 def:variable_name" (after optional whitespace) was not being diagnosed.
3149 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3150 accidental colon was present, for example, could give incorrect results.
3152 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3153 result is used (not including appendfile, where other changes are about
3156 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3157 This involved changing a lot of size variables from int to off_t. It
3158 should work with maildirs and everything.
3160 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3161 spamd dying while we are connected to it.
3163 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3166 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3167 function for BDB 4.3.
3169 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3171 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3172 the sources for PCRE have been re-arranged and more files are now
3175 PH/15 The code I had for printing potentially long long variables in PH/11
3176 above was not the best (it lost precision). The length of off_t variables
3177 is now inspected at build time, and an appropriate printing format (%ld
3178 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3179 to be "long long int" or "long int". This is needed for the internal
3180 formatting function string_vformat().
3182 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3183 the configuration file to be ":syslog", then the script "guesses" where
3184 the logs files are, rather than using the compiled in default. In our
3185 case the guess is not the same as the compiled default, so the script
3186 suddenly stopped working when I started to use syslog. The patch checks
3187 to see if log_file_path is "". If so, it attempts to read it from exim
3188 with no configuration file to get the compiled in version, before it
3189 falls back to the previous guessing code."
3191 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3192 implementing BATV in an Exim configuration. See NewStuff for the gory
3195 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3196 Makefile that are specific to HP-UX.
3198 PH/18 If the "use_postmaster" option was set for a recipient callout together
3199 with the "random" option, the postmaster address was used as the MAIL
3200 FROM address for the random test, but not for the subsequent recipient
3201 test. It is now used for both.
3203 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3204 patch removes a few documentation additions to RFC 3028, because the
3205 latest draft now contains them. It adds the new en;ascii-case comparator
3206 and a new error check for 8bit text in MIME parts. Comparator and
3207 require names are now matched exactly. I enabled the subaddress
3208 extension, but it is not well tested yet (read: it works for me)."
3210 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3211 rework some of the code of TK/09 above to avoid the hardwired use of
3212 "%lld" and "long long". Replaced the call to snprintf() with a call to
3215 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3216 records point to non-existent hosts", "retry timeout exceeded", and
3217 "retry time not reached for any host after a long failure period".
3219 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3220 experimental DomainKeys support:
3222 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3223 (2) On an error such as an illegally used "control", the wrong name for
3224 the control was given.
3226 These problems did NOT occur unless DomainKeys support was compiled.
3228 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3230 PH/24 Added ${if match_ip condition.
3232 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3233 databases so that it will be absolutely obvious if a crash occurs in the
3234 DB library. This is a regular occurrence (often caused by mis-matched
3237 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3238 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3239 picked up on a user's system that detects such things. There doesn't seem
3240 to be a gcc warning option for this - only an attribute that has to be
3241 put on the function's prototype. It seems that in Fedora Core 4 they have
3242 set this on a number of new functions. No doubt there will be more in due
3245 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3246 specify a transport. However, if an address that was verified by such a
3247 router was the subject of a callout, Exim crashed because it tried to
3248 read the rcpt_include_affixes from the non-existent transport. Now it
3249 just assumes that the setting of that option is false. This bug was
3250 introduced by 4.51/PH/31.
3252 PH/28 Changed -d+all to exclude +memory, because that information is very
3253 rarely of interest, but it makes the output a lot bigger. People tend to
3254 do -d+all out of habit.
3256 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3257 code in os-type was giving problems when libc.so lives in lib64, like on
3260 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3261 aren't the modern standard, and it seems that some systems' include files
3262 don't always have them. Exim was already checking for some of the newer
3263 ones like T_AAAA, and defining it itself. I've added checks for all the
3264 record types that Exim uses.
3266 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3267 not automatically generating a new one, as it is supposed to. This
3268 prevented TLS from working. If the file did exist, but contained invalid
3269 data, a new version was generated, as expected. It was only the case of a
3270 non-existent file that was broken.
3272 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3273 with a change in libdomainkeys > 0.64.
3275 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3276 from DNS. If the selector record carries the flag, it now has
3277 precedence over the domain-wide flag.
3279 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3281 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3282 the use of a port name that isn't defined in /etc/services) occurred, the
3283 message was deferred as in a normal delivery, and thus remained on the
3284 spool, instead of being failed because of the mua_wrapper setting. This
3285 is now fixed, and I tidied up some of the mua_wrapper messages at the
3288 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3289 the messages in a hash of arrays rather than using individual hashes.
3290 This is a bit cleaner and results in dramatic memory savings, albeit
3291 at a slight CPU cost.
3293 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3294 as requested by Marc Sherman.
3296 SC/10 Eximstats: added histograms for user specified patterns as requested
3299 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3301 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3302 fopen() in the content-scanning modules that did not already have it.
3308 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3309 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3311 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3313 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3315 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3316 to test Sieve filters that use "vacation".
3318 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3319 that changes the way the GnuTLS parameters are stored in the cache file.
3320 The new format can be generated externally. For backward compatibility,
3321 if the data in the cache doesn't make sense, Exim assumes it has read an
3322 old-format file, and it generates new data and writes a new file. This
3323 means that you can't go back to an older release without removing the
3326 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3327 work if there are any delivery delays because "one_time" forces the
3328 parent to be marked "delivered", so its unseen clone is never tried
3329 again. For this reason, Exim now forbids the simultaneous setting of
3332 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3333 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3334 entry if you want to know the details. The fix, however, neglected to
3335 consider the case when local delivery batching is involved. The test for
3336 "previously delivered" was not happening when checking to see if an
3337 address could be batched with a previous (undelivered) one; under
3338 certain circumstances this could lead to multiple deliveries to the same
3341 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3342 in its include files, and this causes problems building Exim.
3344 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3345 header_syntax) but Exim was just ignoring anything given after a slash.
3346 In particular, this caused confusion with an attempt to use "verify =
3347 reverse_host_lookup/defer_ok". An error is now given when options are
3348 supplied for verify items that do not have them. (Maybe reverse_host_
3349 lookup should have a defer_ok option, but that's a different point.)
3351 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3352 defined by RFC 821) to 2048, because there were problems with some AUTH
3353 commands, and RFC 1869 says the size should be increased for extended
3354 SMTP commands that take arguments.
3356 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3359 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3360 "unknown" error; now it says that the functionality isn't in the binary.
3362 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3363 an address' error message when a string expansion fails (syntax or
3364 whatever). Otherwise the password may appear in the log. Following change
3365 PH/42 below, there is no longer a chance of it appearing in a bounce
3368 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3370 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3371 of its addresses was ever tried. (Bugzilla bug #2.)
3373 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3374 the result incorrectly in the debug output. (It correctly added a newline
3375 to what was transported.)
3377 TF/01 Added $received_time.
3379 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3380 commented out examples of how to interface to a virus scanner and to
3381 SpamAssassin. Also added commented examples of av_scanner and
3382 spamd_address settings.
3384 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3385 and controls are allowed in which ACLs. There were a couple of minor
3386 errors. Some of the entries in the conditions table (which is a table of
3387 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3388 negation of where the condition IS allowed.
3390 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3392 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3393 header file does not have a version number, so I've had to invent a new
3394 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3395 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3396 radiusclient), but it was contributed by a Radius user.
3398 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3399 files or directories, for OpenSSL.
3401 PH/19 When an Exim process that is running as root has to create an Exim log
3402 file, it does so in a subprocess that runs as exim:exim so as to get the
3403 ownership right at creation (otherwise, other Exim processes might see
3404 the file with the wrong ownership). There was no test for failure of this
3405 fork() call, which would lead to the process getting stuck as it waited
3406 for a non-existent subprocess. Forks do occasionally fail when resources
3407 run out. I reviewed all the other calls to fork(); they all seem to check
3410 PH/20 When checking for unexpected SMTP input at connect time (before writing
3411 the banner), Exim was not dealing correctly with a non-positive return
3412 from the read() function. If the client had disconnected by this time,
3413 the result was a log entry for a synchronization error with an empty
3414 string after "input=" when read() returned zero. If read() returned -1
3415 (an event I could not check), uninitialized data bytes were printed.
3416 There were reports of junk text (parts of files, etc) appearing after
3419 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3421 PH/22 Added support for macro redefinition, and (re)definition in between
3422 driver and ACL definitions.
3424 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3425 forgetting to use the resulting value; it was using the unexpanded value.
3427 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3428 hadn't been configured. The fix is from Juergen Kreileder, who
3429 understands it better than I do:
3431 "Here's what I see happening with three configured cyrus_sasl
3432 authenticators configured (plain, login, cram-md5):
3434 On startup auth_cyrus_sasl_init() gets called for each of these.
3435 This means three calls to sasl_listmech() without a specified mech_list.
3436 => SASL tests which mechs of all available mechs actually work
3437 => three warnings about OTP not working
3438 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3440 With the patch, sasl_listmech() also gets called three times. But now
3441 SASL's mech_list option is set to the server_mech specified in the the
3442 authenticator. Or in other words, the answer from sasl_listmech()
3443 gets limited to just the mech you're testing for (which is different
3445 => the return list contains just 'plain' or 'login', 'cram-md5' or
3446 nothing depending on the value of ob->server_mech.
3448 I've just tested the patch: Authentication still works fine,
3449 unavailable mechs specified in the exim configuration are still
3450 caught, and the auth.log warnings about OTP are gone."
3452 PH/25 When debugging is enabled, the contents of the command line are added
3453 to the debugging output, even when log_selector=+arguments is not
3456 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3457 answer is "GNU", and only if the return is "GNU/something" is the answer
3460 PH/27 $acl_verify_message is now set immediately after the failure of a
3461 verification in an ACL, and so is available in subsequent modifiers. In
3462 particular, the message can be preserved by coding like this:
3464 warn !verify = sender
3465 set acl_m0 = $acl_verify_message
3467 Previously, $acl_verify_message was set only while expanding "message"
3468 and "log_message" when a very denied access.
3470 PH/28 Modified OS/os.c-Linux with
3472 -#ifndef OS_LOAD_AVERAGE
3473 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3475 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3476 nomenclature these days.)
3478 PH/29 Installed patch from the Sieve maintainer that adds the options
3479 sieve_useraddress and sieve_subaddress to the redirect router.
3481 PH/30 In these circumstances:
3482 . Two addresses routed to the same list of hosts;
3483 . First host does not offer TLS;
3484 . First host accepts first address;
3485 . First host gives temporary error to second address;
3486 . Second host offers TLS and a TLS session is established;
3487 . Second host accepts second address.
3488 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3489 and peerdn, if requested) that were in fact used only for the second
3492 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3493 attention to any local part prefix or suffix that was matched by the
3494 router that accepted the address. It now behaves in the same way as it
3495 does for delivery: the affixes are removed from the local part unless
3496 rcpt_include_affixes is set on the transport.
3498 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3499 timeout during the DATA phase of an incoming message.
3501 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3502 applied a patch sanctioned by the Sieve maintainer.
3504 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3505 the uid or gid is negative. A case of a negative gid caused this to be
3506 noticed. The fix allows for either to be negative.
3508 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3509 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3512 PH/36 The change PH/12 above was broken. Fixed it.
3514 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3515 the grounds that routing the same address twice would always produce the
3516 same answer. This might have been true once, but it is certainly no
3517 longer true now. Routing a child address may depend on the previous
3518 routing that produced that child. Some complicated redirection strategies
3519 went wrong when messages had multiple recipients, and made Exim's
3520 behaviour dependent on the order in which the addresses were given.
3522 I have moved the duplicate checking until after the routing is complete.
3523 Exim scans the addresses that are assigned to local and remote
3524 transports, and removes any duplicates. This means that more work will be
3525 done, as duplicates will always all be routed, but duplicates are
3526 presumably rare, so I don't expect this is of any significance.
3528 For deliveries to pipes, files, and autoreplies, the duplicate checking
3529 still happens during the routing process, since they are not going to be
3532 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3533 It corrects a timeout issue with spamd. This is Ian's comment: "The
3534 background is that sometimes spamd either never reads data from a
3535 connection it has accepted, or it never writes response data. The exiscan
3536 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3537 blindly assumes that writes won't block so it may never time out."
3539 PH/39 Allow G after quota size as well as K and M.
3541 PH/40 The value set for $authenticated_id in an authenticator may not contain
3542 binary zeroes or newlines because the value is written to log lines and
3543 to spool files. There was no check on this. Now the value is run through
3544 the string_printing() function so that such characters are converted to
3545 printable escape sequences.
3547 PH/41 $message_linecount is a new variable that contains the total number of
3548 lines in the message. Compare $body_linecount, which is the count for the
3551 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3552 bounce and delay warning messages, except in certain special cases, which
3555 (a) An SMTP error message from a remote host;
3556 (b) A message specified in a :fail: redirection;
3557 (c) A message specified in a "fail" command in a system filter;
3558 (d) A message specified in a FAIL return from the queryprogram router;
3559 (e) A message specified by the cannot_route_message router option.
3561 In these cases only, Exim does include the error details in bounce and
3562 warning messages. There are also a few cases where bland messages such
3563 as "unrouteable address" or "local delivery error" are given.
3565 PH/43 $value is now also set for the "else" part of a ${run expansion.
3567 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3568 being worked on, but at least Exim now implements the latest version to
3571 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3572 process to complete was treated as a delivery failure, a timeout while
3573 writing the message to the pipe was logged, but erroneously treated as a
3574 successful delivery. Such timeouts include transport filter timeouts. For
3575 consistency with the overall process timeout, these timeouts are now
3576 treated as errors, giving rise to delivery failures by default. However,
3577 there is now a new Boolean option for the pipe transport called
3578 timeout_defer, which, if set TRUE, converts the failures into defers for
3579 both kinds of timeout. A transport filter timeout is now identified in
3582 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3583 systems where "make" and "gmake" are different, calling "gmake" at top
3584 level broke things. I've arranged for the value of $(MAKE) to be passed
3585 from the Makefile to this script so that it can call the same version of
3589 A note about Exim versions 4.44 and 4.50
3590 ----------------------------------------
3592 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3593 changes of various kinds. As a consequence, a big documentation update was
3594 needed. This delayed the release for rather longer than seemed good, especially
3595 in the light of a couple of (minor) security issues. Therefore, the changes
3596 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3597 release. So 4.44 and 4.50 are in effect two different branches that both start
3600 I have left the 4.50 change log unchanged; it contains all the changes since
3601 4.43. The change log for 4.44 is below; many of its items are identical to
3602 those for 4.50. This seems to be the most sensible way to preserve the
3603 historical information.
3609 1. Minor wording change to the doc/README.SIEVE file.
3611 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3612 computation of the current number of files was incorrect.
3614 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3615 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3618 4. Give more explanation in the error message when the command for a transport
3619 filter fails to execute.
3621 5. There are several places where Exim runs a non-Exim command in a
3622 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3623 was being done only for the command run by the queryprogram router. It is
3624 now done for all such subprocesses. The other cases are: ${run, transport
3625 filters, and the commands run by the lmtp and pipe transports.
3627 6. Added CONFIGURE_GROUP build-time option.
3629 7. Some older OS have a limit of 256 on the maximum number of file
3630 descriptors. Exim was using setrlimit() to set 1000 as a large value
3631 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3632 systems. I've change it so that if it can't get 1000, it tries for 256.
3634 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3635 was an oversight, and furthermore, ever since the addition of extra
3636 controls (e.g. 4.43/32), the checks on when to allow different forms of
3637 "control" were broken. There should now be diagnostics for all cases when a
3638 control that does not make sense is encountered.
3640 9. Added the /retain_sender option to "control=submission".
3642 10. $recipients is now available in the predata ACL (oversight).
3644 11. Tidy the search cache before the fork to do a delivery from a message
3645 received from the command line. Otherwise the child will trigger a lookup
3646 failure and thereby defer the delivery if it tries to use (for example) a
3647 cached ldap connection that the parent has called unbind on.
3649 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3650 of $address_data from the recipient verification was clobbered by the
3651 sender verification.
3653 13. The value of address_data from a sender verification is now available in
3654 $sender_address_data in subsequent conditions in the ACL statement.
3656 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3658 15. Added a new option "connect=<time>" to callout options, to set a different
3661 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3662 was its contents. (It was OK if the option was not defined at all.)
3664 17. A "Completed" log line is now written for messages that are removed from
3665 the spool by the -Mrm option.
3667 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3668 information about exactly what failed.
3670 19. Added -dd to debug only the daemon process.
3672 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3673 handles timeouts, both on the server side and network timeouts. Renamed the
3674 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3676 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3677 It is now set to "smtps".
3679 22. $host_address is now set to the target address during the checking of
3680 ignore_target_hosts.
3682 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3683 being passed; this would have caused $sender_host_name to have been used if
3684 matching the list had actually called for a host name (not very likely,
3685 since this list is usually IP addresses). A host name is now passed as
3688 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3689 code that specifies a non-restarting handler (typically sigaction() in
3690 modern systems) in an attempt to fix a rare and obscure crash bug.
3692 25. Narrowed the window for a race in the daemon that could cause it to ignore
3693 SIGCHLD signals. This is not a major problem, because they are used only to
3694 wake it up if nothing else does.
3696 26. A malformed maildirsize file could cause Exim to calculate negative values
3697 for the mailbox size or file count. Odd effects could occur as a result.
3698 The maildirsize information is now recalculated if the size or filecount
3701 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3702 support for a long time. Removed HAVE_SYS_VFS_H.
3704 28. Installed the latest version of exipick from John Jetmore.
3706 29. In an address list, if the pattern was not a regular expression, an empty
3707 subject address (from a bounce message) matched only if the pattern was an
3708 empty string. Non-empty patterns were not even tested. This was the wrong
3709 because it is perfectly reasonable to use an empty address as part of a
3710 database query. An empty address is now tested by patterns that are
3711 lookups. However, all the other forms of pattern expect the subject to
3712 contain a local part and a domain, and therefore, for them, an empty
3713 address still always fails if the pattern is not itself empty.
3715 30. Exim went into a mad DNS loop when attempting to do a callout where the
3716 host was specified on an smtp transport, and looking it up yielded more
3717 than one IP address.
3719 31. Re-factored the code for checking spool and log partition space into a
3720 function that finds that data and another that does the check. The former
3721 is then used to implement four new variables: $spool_space, $log_space,
3722 $spool_inodes, and $log_inodes.
3724 32. The RFC2047 encoding function was originally intended for short strings
3725 such as real names; it was not keeping to the 75-character limit for
3726 encoded words that the RFC imposes. It now respects the limit, and
3727 generates multiple encoded words if necessary. To be on the safe side, I
3728 have increased the buffer size for the ${rfc2047: expansion operator from
3731 33. It is now permitted to omit both strings after an "if" condition; if the
3732 condition is true, the result is "true". As before, when the second string
3733 is omitted, a false condition yields an empty string. This makes it less
3734 cumbersome to write custom ACL and router conditions.
3736 34. Failure to deliver a bounce message always caused it to be frozen, even if
3737 there was an errors_to setting on the router. The errors_to setting is now
3740 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3741 canonical form (fully expanded) before being placed in
3742 $sender_host_address.
3744 36. The table in the code that translates DNS record types into text (T_A to
3745 "A" for instance) was missing entries for NS and CNAME. It is just possible
3746 that this could have caused confusion if both these types were looked up
3747 for the same domain, because the text type is used as part of Exim's
3748 per-process caching. But the chance of anyone hitting this buglet seems
3751 37. The dnsdb lookup has been extended in a number of ways.
3753 (1) There is a new type, "zns", which walks up the domain tree until it
3754 finds some nameserver records. It should be used with care.
3756 (2) There is a new type, "mxh", which is like "mx" except that it returns
3757 just the host names, not the priorities.
3759 (3) It is now possible to give a list of domains (or IP addresses) to be
3760 looked up. The behaviour when one of the lookups defers can be
3761 controlled by a keyword.
3763 (4) It is now possible to specify the separator character for use when
3764 multiple records are returned.
3766 38. The dnslists ACL condition has been extended: it is now possible to supply
3767 a list of IP addresses and/or domains to be looked up in a particular DNS
3770 39. Added log_selector=+queue_time_overall.
3772 40. When running the queue in the test harness, wait just a tad after forking a
3773 delivery process, to get repeatability of debugging output.
3775 41. Include certificate and key file names in error message when GnuTLS fails
3776 to set them up, because the GnuTLS error message doesn't include the name
3777 of the failing file when there is a problem reading it.
3779 42. Allow both -bf and -bF in the same test run.
3781 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3783 44. The "Exiscan patch" is now merged into the mainline Exim source.
3785 45. Sometimes the final signoff response after QUIT could fail to get
3786 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3787 < 0 before doing a fflush(). This bug looks as though it goes back to the
3788 introduction of TLS in release 3.20, but "sometimes" must have been rare
3789 because the tests only now provoked it.
3791 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3792 (this can affect the format of dates).
3794 47. exim_tidydb, when checking for the continued existence of a message for
3795 which it has found a message-specific retry record, was not finding
3796 messages that were in split spool directories. Consequently, it was
3797 deleting retry records that should have stayed in existence.
3799 48. Steve fixed some bugs in eximstats.
3801 49. The SPA authentication driver was not abandoning authentication and moving
3802 on to the next authenticator when an expansion was forced to fail,
3803 contradicting the general specification for all authenticators. Instead it
3804 was generating a temporary error. It now behaves as specified.
3806 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3807 (the order specifies the preference for clients). The order is now AES256,
3808 AES128, 3DES, ARCFOUR128.
3810 51. Small patch to Sieve code - explicitly set From: when generating an
3813 52. Exim crashed if a remote delivery caused a very long error message to be
3814 recorded - for instance if somebody sent an entire SpamAssassin report back
3815 as a large number of 550 error lines. This bug was coincidentally fixed by
3816 increasing the size of one of Exim's internal buffers (big_buffer) that
3817 happened as part of the Exiscan merge. However, to be on the safe side, I
3818 have made the code more robust (and fixed the comments that describe what
3821 53. Now that there can be additional text after "Completed" in log lines (if
3822 the queue_time_overall log selector is set), a one-byte patch to exigrep
3823 was needed to allow it to recognize "Completed" as not the last thing in
3826 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3827 patch that reportedly fixes this has been added. I am not expert enough to
3828 create a test for it. This is what the patch creator wrote:
3830 "I found a little strange behaviour of ldap code when working with
3831 Windows 2003 AD Domain, where users was placed in more than one
3832 Organization Units. When I tried to give exim partial DN, the exit code
3833 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3834 But simultaneously result of request was absolutely normal ldap result,
3835 so I produce this patch..."
3837 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3838 REFERENCE, so I have modified the code to exclude the patch when that macro
3841 55. Some experimental protocols are using DNS PTR records for new purposes. The
3842 keys for these records are domain names, not reversed IP addresses. The
3843 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3844 leaves it alone. Component reversal etc. now happens only for IP addresses.
3847 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3849 57. Double the size of the debug message buffer (to 2048) so that more of very
3850 long debug lines gets shown.
3852 58. The exicyclog utility now does better if the number of log files to keep
3853 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3855 59. Two changes related to the smtp_active_hostname option:
3857 (1) $smtp_active_hostname is now available as a variable.
3858 (2) The default for smtp_banner uses $smtp_active_hostname instead
3859 of $primary_hostname.
3861 60. The host_aton() function is supposed to be passed a string that is known
3862 to be a valid IP address. However, in the case of IPv6 addresses, it was
3863 not checking this. This is a hostage to fortune. Exim now panics and dies
3864 if the condition is not met. A case was found where this could be provoked
3865 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3866 components; fortuitously, this particular loophole had already been fixed
3867 by change 4.50/55 above.
3869 If there are any other similar loopholes, the new check in host_aton()
3870 itself should stop them being exploited. The report I received stated that
3871 data on the command line could provoke the exploit when Exim was running as
3872 exim, but did not say which command line option was involved. All I could
3873 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3874 running as the user.
3877 61. There was a buffer overflow vulnerability in the SPA authentication code
3878 (which came originally from the Samba project). I have added a test to the
3879 spa_base64_to_bits() function which I hope fixes it.
3882 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3883 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3885 63. The daemon start-up calls getloadavg() while still root for those OS that
3886 need the first call to be done as root, but it missed one case: when
3887 deliver_queue_load_max is set with deliver_drop_privilege. This is
3888 necessary for the benefit of the queue runner, because there is no re-exec
3889 when deliver_drop_privilege is set.
3891 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3892 This has been fixed.
3894 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3895 was in use, was not putting the data itself into the right store pool;
3896 consequently, it could be overwritten for a subsequent message in the same
3897 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3900 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3902 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3903 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3904 the function were treating the return as a boolean value, which happened to
3905 work because 0=false and not-0=true, but is not correct code.
3907 68. The host_aton() function was not handling scoped IPv6 addresses (those
3908 with, for example, "%eth0" on the end) correctly.
3910 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3911 negated items (that is, ~something) in unsigned ints. Some compilers
3912 apparently mutter when there is no cast.
3914 70. If an address verification called from an ACL failed, and did not produce a
3915 user-specific message (i.e. there was only a "system" message), nothing was
3916 put in $acl_verify_message. In this situation, it now puts the system
3919 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3920 session; change 4.31/43 added the unwanted input to the log line - except
3921 that it did not do this in the start of session case. It now does.
3923 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3924 This is wrong and can cause the other end to generate a synchronization
3925 error if it is another Exim or anything else that does the synchronization
3926 check. A QUIT command is no longer sent after a timeout.
3928 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3929 during host lookups.
3931 74. The defer_ok option of callout verification was not working if it was used
3932 when verifying addresses in header lines, that is, for this case:
3934 verify = header_sender/callout=defer_ok
3936 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3937 those file descriptors could be used for SMTP connections. If anything
3938 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3939 could be sent to the SMTP client, causing chaos. The daemon now opens
3940 stdin, stdout, and stderr to /dev/null when it puts itself into the
3943 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3944 log by default. The user can override this with suitable Perl magic.
3946 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3947 the log message when discard triggers, was not working for the DATA ACL or
3948 for the non-SMTP ACL.
3950 78. Error message wording change in sieve.c.
3952 79. If smtp_accept_max_per_host was set, the number of connections could be
3953 restricted to fewer than expected, because the daemon was trying to set up
3954 a new connection before checking whether the processes handling previous
3955 connections had finished. The check for completed processes is now done
3956 earlier. On busy systems, this bug wouldn't be noticed because something
3957 else would have woken the daemon, and it would have reaped the completed
3960 80. If a message was submitted locally by a user whose login name contained one
3961 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3962 It caused a spool format error. I have fixed the spool reading code. A
3963 related problem was that the "from" clause in the Received: line became
3964 illegal because of the space(s). It is now covered by ${quote_local_part.
3966 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3969 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3971 83. Patch from Sieve maintainer to fix unterminated string problem in
3972 "vacation" handling.
3974 84. Some minor changes to the Linux configuration files to help with other
3975 OS variants using glibc.
3977 85. One more patch for Sieve to update vacation handling to latest spec.
3980 ----------------------------------------------------
3981 See the note above about the 4.44 and 4.50 releases.
3982 ----------------------------------------------------
3988 1. Change 4.43/35 introduced a bug that caused file counts to be
3989 incorrectly computed when quota_filecount was set in an appendfile
3992 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3993 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3996 3. Give more explanation in the error message when the command for a transport
3997 filter fails to execute.
3999 4. There are several places where Exim runs a non-Exim command in a
4000 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4001 was being done only for the command run by the queryprogram router. It is
4002 now done for all such subprocesses. The other cases are: ${run, transport
4003 filters, and the commands run by the lmtp and pipe transports.
4005 5. Some older OS have a limit of 256 on the maximum number of file
4006 descriptors. Exim was using setrlimit() to set 1000 as a large value
4007 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4008 systems. I've change it so that if it can't get 1000, it tries for 256.
4010 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4011 was an oversight, and furthermore, ever since the addition of extra
4012 controls (e.g. 4.43/32), the checks on when to allow different forms of
4013 "control" were broken. There should now be diagnostics for all cases when a
4014 control that does not make sense is encountered.
4016 7. $recipients is now available in the predata ACL (oversight).
4018 8. Tidy the search cache before the fork to do a delivery from a message
4019 received from the command line. Otherwise the child will trigger a lookup
4020 failure and thereby defer the delivery if it tries to use (for example) a
4021 cached ldap connection that the parent has called unbind on.
4023 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4024 of $address_data from the recipient verification was clobbered by the
4025 sender verification.
4027 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4028 was its contents. (It was OK if the option was not defined at all.)
4030 11. A "Completed" log line is now written for messages that are removed from
4031 the spool by the -Mrm option.
4033 12. $host_address is now set to the target address during the checking of
4034 ignore_target_hosts.
4036 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4037 being passed; this would have caused $sender_host_name to have been used if
4038 matching the list had actually called for a host name (not very likely,
4039 since this list is usually IP addresses). A host name is now passed as
4042 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4043 code that specifies a non-restarting handler (typically sigaction() in
4044 modern systems) in an attempt to fix a rare and obscure crash bug.
4046 15. Narrowed the window for a race in the daemon that could cause it to ignore
4047 SIGCHLD signals. This is not a major problem, because they are used only to
4048 wake it up if nothing else does.
4050 16. A malformed maildirsize file could cause Exim to calculate negative values
4051 for the mailbox size or file count. Odd effects could occur as a result.
4052 The maildirsize information is now recalculated if the size or filecount
4055 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4056 support for a long time. Removed HAVE_SYS_VFS_H.
4058 18. Updated exipick to current release from John Jetmore.
4060 19. Allow an empty sender to be matched against a lookup in an address list.
4061 Previously the only cases considered were a regular expression, or an
4064 20. Exim went into a mad DNS lookup loop when doing a callout where the
4065 host was specified on the transport, if the DNS lookup yielded more than
4068 21. The RFC2047 encoding function was originally intended for short strings
4069 such as real names; it was not keeping to the 75-character limit for
4070 encoded words that the RFC imposes. It now respects the limit, and
4071 generates multiple encoded words if necessary. To be on the safe side, I
4072 have increased the buffer size for the ${rfc2047: expansion operator from
4075 22. Failure to deliver a bounce message always caused it to be frozen, even if
4076 there was an errors_to setting on the router. The errors_to setting is now
4079 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4080 canonical form (fully expanded) before being placed in
4081 $sender_host_address.
4083 24. Updated eximstats to version 1.33
4085 25. Include certificate and key file names in error message when GnuTLS fails
4086 to set them up, because the GnuTLS error message doesn't include the name
4087 of the failing file when there is a problem reading it.
4089 26. Expand error message when OpenSSL has problems setting up cert/key files.
4092 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4093 (this can affect the format of dates).
4095 28. exim_tidydb, when checking for the continued existence of a message for
4096 which it has found a message-specific retry record, was not finding
4097 messages that were in split spool directories. Consequently, it was
4098 deleting retry records that should have stayed in existence.
4100 29. eximstats updated to version 1.35
4101 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4102 1.35 - bugfix such that pie charts by volume are generated correctly
4104 30. The SPA authentication driver was not abandoning authentication and moving
4105 on to the next authenticator when an expansion was forced to fail,
4106 contradicting the general specification for all authenticators. Instead it
4107 was generating a temporary error. It now behaves as specified.
4109 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4110 (the order specifies the preference for clients). The order is now AES256,
4111 AES128, 3DES, ARCFOUR128.
4113 31. Small patch to Sieve code - explicitly set From: when generating an
4116 32. Exim crashed if a remote delivery caused a very long error message to be
4117 recorded - for instance if somebody sent an entire SpamAssassin report back
4118 as a large number of 550 error lines. This bug was coincidentally fixed by
4119 increasing the size of one of Exim's internal buffers (big_buffer) that
4120 happened as part of the Exiscan merge. However, to be on the safe side, I
4121 have made the code more robust (and fixed the comments that describe what
4124 33. Some experimental protocols are using DNS PTR records for new purposes. The
4125 keys for these records are domain names, not reversed IP addresses. The
4126 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4127 leaves it alone. Component reversal etc. now happens only for IP addresses.
4130 34. The host_aton() function is supposed to be passed a string that is known
4131 to be a valid IP address. However, in the case of IPv6 addresses, it was
4132 not checking this. This is a hostage to fortune. Exim now panics and dies
4133 if the condition is not met. A case was found where this could be provoked
4134 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4135 components; fortuitously, this particular loophole had already been fixed
4136 by change 4.50/55 or 4.44/33 above.
4138 If there are any other similar loopholes, the new check in host_aton()
4139 itself should stop them being exploited. The report I received stated that
4140 data on the command line could provoke the exploit when Exim was running as
4141 exim, but did not say which command line option was involved. All I could
4142 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4143 running as the user.
4146 35. There was a buffer overflow vulnerability in the SPA authentication code
4147 (which came originally from the Samba project). I have added a test to the
4148 spa_base64_to_bits() function which I hope fixes it.
4151 36. The daemon start-up calls getloadavg() while still root for those OS that
4152 need the first call to be done as root, but it missed one case: when
4153 deliver_queue_load_max is set with deliver_drop_privilege. This is
4154 necessary for the benefit of the queue runner, because there is no re-exec
4155 when deliver_drop_privilege is set.
4157 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4158 was in use, was not putting the data itself into the right store pool;
4159 consequently, it could be overwritten for a subsequent message in the same
4160 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4163 38. Sometimes the final signoff response after QUIT could fail to get
4164 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4165 < 0 before doing a fflush(). This bug looks as though it goes back to the
4166 introduction of TLS in release 3.20, but "sometimes" must have been rare
4167 because the tests only now provoked it.
4173 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4174 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4175 Now it returns an integer. A number of calls were still expecting a T/F
4176 return. Fortuitously, in all cases, the tests worked in OK situations,
4177 which is the norm. However, things would have gone wrong on any write
4178 failures on the smtp file descriptor. This function is used when sending
4179 messages over SMTP and also when doing verify callouts.
4181 2. When Exim is called to do synchronous delivery of a locally submitted
4182 message (the -odf or -odi options), it no longer closes stderr before doing
4185 3. Implemented the mua_wrapper option.
4187 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4189 5. Implemented the functions header_remove(), header_testname(),
4190 header_add_at_position(), and receive_remove_recipient(), and exported them
4193 6. If an ACL "warn" statement specified the addition of headers, Exim already
4194 inserted X-ACL-Warn: at the start if there was no header name. However, it
4195 was not making this test for the second and subsequent header lines if
4196 there were newlines in the string. This meant that an invalid header could
4197 be inserted if Exim was badly configured.
4199 7. Allow an ACL "warn" statement to add header lines at the start or after all
4200 the Received: headers, as well as at the end.
4202 8. Added the rcpt_4xx retry error code.
4204 9. Added postmaster_mailfrom=xxx to callout verification option.
4206 10. Added mailfrom=xxxx to the callout verification option, for verify=
4209 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4210 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4212 12. Inserted some casts to stop certain compilers warning when using pointer
4213 differences as field lengths or precisions in printf-type calls (mostly
4214 affecting debugging statements).
4216 13. Added optional readline() support for -be (dynamically loaded).
4218 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4219 same clock tick as a message's arrival, so that its received time was the
4220 same as the "first fail" time on the retry record, and that message
4221 remained on the queue past the ultimate address timeout, every queue runner
4222 would try a delivery (because it was past the ultimate address timeout) but
4223 after another failure, the ultimate address timeout, which should have then
4224 bounced the address, did not kick in. This was a "< instead of <=" error;
4225 in most cases the first failure would have been in the next clock tick
4226 after the received time, and all would be well.
4228 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4229 being recognized when the domain list was tested by the match_domain
4230 condition in an expansion string.
4232 16. Added the ${str2b64: operator.
4234 17. Exim was always calling setrlimit() to set a large limit for the number of
4235 processes, without checking whether the existing limit was already
4236 adequate. (It did check for the limit on file descriptors.) Furthermore,
4237 errors from getrlimit() and setrlimit() were being ignored. Now they are
4238 logged to the main and panic logs, but Exim does carry on, to try to do its
4239 job under whatever limits there are.
4241 18. Imported PCRE 5.0.
4243 19. Trivial typo in log message " temporarily refused connection" (the leading
4246 20. If the log selector return_path_on_delivery was set and an address was
4247 redirected to /dev/null, the delivery process crashed because it assumed
4248 that a return path would always be set for a "successful" delivery. In this
4249 case, the whole delivery is bypassed as an optimization, and therefore no
4252 21. Internal re-arrangement: the function for sending a challenge and reading
4253 a response while authentication was assuming a zero-terminated challenge
4254 string. It's now changed to take a pointer and a length, to allow for
4255 binary data in such strings.
4257 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4259 23. Exim was not respecting finduser_retries when seeking the login of the
4260 uid under which it was called; it was always trying 10 times. (The default
4261 setting of finduser_retries is zero.) Also, it was sleeping after the final
4262 failure, which is pointless.
4264 24. Implemented tls_on_connect_ports.
4266 25. Implemented acl_smtp_predata.
4268 26. If the domain in control=submission is set empty, Exim assumes that the
4269 authenticated id is a complete email address when it generates From: or
4270 Sender: header lines.
4272 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4273 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4274 chown and chgrp in /bin and hostname in /usr/bin.
4276 28. Exim was keeping the "process log" file open after each use, just as it
4277 does for the main log. This opens the possibility of it remaining open for
4278 long periods when the USR1 signal hits a daemon. Occasional processlog
4279 errors were reported, that could have been caused by this. Anyway, it seems
4280 much more sensible not to leave this file open at all, so that is what now
4283 29. The long-running daemon process does not normally write to the log once it
4284 has entered its main loop, and it closes the log before doing so. This is
4285 so that log files can straightforwardly be renamed and moved. However,
4286 there are a couple of unusual error situations where the daemon does write
4287 log entries, and I had neglected to close the log afterwards.
4289 30. The text of an SMTP error response that was received during a remote
4290 delivery was being truncated at 512 bytes. This is too short for some of
4291 the long messages that one sometimes sees. I've increased the limit to
4294 31. It is now possible to make retry rules that apply only when a message has a
4295 specific sender, in particular, an empty sender.
4297 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4298 it possible to be selective about when SMTP synchronization is enforced.
4300 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4302 32. Implemented hosts_connection_nolog.
4304 33. Added an ACL for QUIT.
4306 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4309 35. Added mailbox_size and mailbox_filecount to appendfile.
4311 36. Added control = no_multiline_responses to ACLs.
4313 37. There was a bug in the logic of the code that waits for the clock to tick
4314 in the case where the clock went backwards by a substantial amount such
4315 that the microsecond fraction of "now" was more than the microsecond
4316 fraction of "then" (but the whole seconds number was less).
4318 38. Added support for the libradius Radius client library this is found on
4319 FreeBSD (previously only the radiusclient library was supported).
4325 1. When certain lookups returned multiple values in the form name=value, the
4326 quoting of the values was not always being done properly. Specifically:
4327 (a) If the value started with a double quote, but contained no whitespace,
4329 (b) If the value contained whitespace other than a space character (i.e.
4330 tabs or newlines or carriage returns) it was not quoted.
4331 This fix has been applied to the mysql and pgsql lookups by writing a
4332 separate quoting function and calling it from the lookup code. The fix
4333 should probably also be applied to nisplus, ibase and oracle lookups, but
4334 since I cannot test any of those, I have not disturbed their existing code.
4336 2. A hit in the callout cache for a specific address caused a log line with no
4337 reason for rejecting RCPT. Now it says "Previous (cached) callout
4338 verification failure".
4340 3. There was an off-by-one bug in the queryprogram router. An over-long
4341 return line was truncated at 256 instead of 255 characters, thereby
4342 overflowing its buffer with the terminating zero. As well as fixing this, I
4343 have increased the buffer size to 1024 (and made a note to document this).
4345 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4346 when Exim is waiting for an SMTP response from a remote server, Exim
4347 restarts its select() call on the socket, thereby resetting its timeout.
4348 This is not a problem when such interrupts are rare. Somebody set up a cron
4349 job to run exiwhat every 2 minutes, which is less than the normal select()
4350 timeout (5 or 10 minutes). This meant that the select() timeout never
4351 kicked in because it was always reset. I have fixed this by comparing the
4352 time when an interrupt arrives with the time at the start of the first call
4353 to select(). If more time than the timeout has elapsed, the interrupt is
4354 treated as a timeout.
4356 5. Some internal re-factoring in preparation for the addition of Sieve
4357 extensions (by MH). In particular, the "personal" test is moved to a
4358 separate function, and given an option for scanning Cc: and Bcc: (which is
4359 not set for Exim filters).
4361 6. When Exim created an email address using the login of the caller as the
4362 local part (e.g. when creating a From: or Sender: header line), it was not
4363 quoting the local part when it contained special characters such as @.
4365 7. Installed new OpenBSD configuration files.
4367 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4368 try to make them clearer.
4370 9. Callout options, other than the timeout value, were being ignored when
4371 verifying sender addresses in header lines. For example, when using
4373 verify = header_sender/callout=no_cache
4375 the cache was (incorrectly) being used.
4377 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4378 only the Cygwin environment.
4380 11. When return_path_on_delivery was set as a log selector, if different remote
4381 addresses in the same message used different return paths and parallel
4382 remote delivery occurred, the wrong values would sometimes be logged.
4383 (Whenever a remote delivery process finished, the return path value from
4384 the most recently started remote delivery process was logged.)
4386 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4387 lines when AUTH and/or TLS are in use. This is the "received protocol"
4388 field. Exim used to use "asmtp" for authenticated SMTP, without any
4389 indication (in the protocol name) for TLS use. Now it follows the RFC and
4390 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4391 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4392 names appear in log lines as well as in Received: header lines.
4394 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4395 extensions, and comparison tests, and to fix some bugs.
4397 14. Changes to the "personal" filter test:
4399 (1) The test was buggy in that it was just doing the equivalent of
4400 "contains" tests on header lines. For example, if a user's address was
4401 anne@some.where, the "personal" test would incorrectly be true for
4403 To: susanne@some.where
4405 This test is now done by extracting each address from the header in turn,
4406 and checking the entire address. Other tests that are part of "personal"
4407 are now done using regular expressions (for example, to check local parts
4408 of addresses in From: header lines).
4410 (2) The list of non-personal local parts in From: addresses has been
4411 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4412 taken from the Sieve specification recommendations.
4414 (3) If the message contains any header line starting with "List-" it is
4415 treated as non-personal.
4417 (4) The test for "circular" in the Subject: header line has been removed
4418 because it now seems ill-conceived.
4420 15. Minor typos in src/EDITME comments corrected.
4422 16. Installed latest exipick from John Jetmore.
4424 17. If headers_add on a router specified a text string that was too long for
4425 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4426 of string_sprintf() is now avoided.
4428 18. $message_body_size was not set (it was always zero) when running the DATA
4429 ACL and the local_scan() function.
4431 19. For the "mail" command in an Exim filter, no default was being set for
4432 the once_repeat time, causing a random time value to be used if "once" was
4433 specified. (If the value happened to be <= 0, no repeat happened.) The
4434 default is now 0s, meaning "never repeat". The "vacation" command was OK
4435 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4436 (I found it when inspecting the code).
4438 20. There is now an overall timeout for performing a callout verification. It
4439 defaults to 4 times the callout timeout, which applies to individual SMTP
4440 commands during the callout. The overall timeout applies when there is more
4441 than one host that can be tried. The timeout is checked before trying the
4442 next host. This prevents very long delays if there are a large number of
4443 hosts and all are timing out (e.g. when the network connections are timing
4444 out). The value of the overall timeout can be changed by specifying an
4445 additional sub-option for "callout", called "maxwait". For example:
4447 verify = sender/callout=5s,maxwait=20s
4449 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4450 to the end before writing, but this should make it even safer).
4452 22. Exim was forgetting that it had advertised PIPELINING for the second and
4453 subsequent messages on an SMTP connection. It was also not resetting its
4454 memory on STARTTLS and an internal HELO.
4456 23. When Exim logs an SMTP synchronization error within a session, it now
4457 records whether PIPELINING has been advertised or not.
4459 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4460 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4461 rather than long int.
4463 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4465 26. Added the never_mail option to autoreply.
4471 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4472 crash if the getsockname() call failed; this can happen if a connection is
4473 closed very soon after it is established. The problem was simply in the
4474 order in which certain operations were done, causing Exim to try to write
4475 to the SMTP stream before it had set up the file descriptor. The bug has
4476 been fixed by making things happen in the correct order.
4482 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4483 before the connection was closed, thus losing the rejection response.
4485 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4486 some early Solaris releases, but causes trouble in current releases where
4487 socklen_t is defined.
4489 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4492 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4495 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4496 that releasing the top of it at the end releases what was used for sub-
4497 expansions (unless the block got too big). However, discard this block if
4498 the first thing is a variable or header, so that we can use its block when
4499 it is dynamic (useful for very large $message_headers, for example).
4501 6. Lookups now cache *every* query, not just the most recent. A new, separate
4502 store pool is used for this. It can be recovered when all lookup caches are
4503 flushed. Lookups now release memory at the end of their result strings.
4504 This has involved some general refactoring of the lookup sources.
4506 7. Some code has been added to the store_xxx() functions to reduce the amount
4507 of flapping under certain conditions.
4509 8. log_incoming_interface used to affect only the <= reception log lines. Now
4510 it causes the local interface and port to be added to several more SMTP log
4511 lines, for example "SMTP connection from", and rejection lines.
4513 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4515 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4517 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4518 could be overwritten at the end of the current message (or the start of a
4519 new message if it was set in a HELO ACL). The value is now preserved for
4520 the duration of the SMTP connection.
4522 12. If a transport had a headers_rewrite setting, and a matching header line
4523 contained an unqualified address, that address was qualified, even if it
4524 did not match any rewriting rules. The underlying bug was that the values
4525 of the flags that permit the existence of unqualified sender and recipient
4526 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4527 non-local messages, and by -bnq for local messages) were not being
4528 preserved with the message after it was received.
4530 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4531 "next input=" as part of the text comprising the host identity instead of
4532 the correct text. The code was using the same buffer for two different
4533 strings. However, depending on which order the printing function evaluated
4534 its arguments, the bug did not always show up. Under Linux, for example, my
4535 test suite worked just fine.
4537 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4538 doesn't work with some older versions of Perl. It has been changed to "my",
4539 which in any case is probably the better facility to use.
4541 15. A really picky compiler found some instances of statements for creating
4542 error messages that either had too many or two few arguments for the format
4545 16. The size of the buffer for calls to the DNS resolver has been increased
4546 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4547 for addresses that have a lot of PTR records. This alleviates a problem; it
4548 does not fully solve it.
4550 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4551 buffer now truncates the list and logs the incident, which is the same
4552 action as happens when Exim is looking up a host name and its aliases.
4553 Previously in this situation something unpredictable would happen;
4554 sometimes it was "internal error: store_reset failed".
4556 18. If a server dropped the connection unexpectedly when an Exim client was
4557 using GnuTLS and trying to read a response, the client delivery process
4558 crashed while trying to generate an error log message.
4560 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4561 string, for example:
4563 warn message = H1: something\nH2: something
4565 the text was added as a single header line from Exim's point of view
4566 though it ended up OK in the delivered message. However, searching for the
4567 second and subsequent header lines using $h_h2: did not work. This has been
4568 fixed. Similarly, if a system filter added multiple headers in this way,
4569 the routers could not see them.
4571 20. Expanded the error message when iplsearch is called with an invalid key to
4572 suggest using net-iplsearch in a host list.
4574 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4575 ACLs are no longer actually imposed (and a message to that effect is
4578 22. If a "gecos" field in a passwd entry contained escaped characters, in
4579 particular, if it contained a \" sequence, Exim got it wrong when building
4580 a From: or a Sender: header from that name. A second bug also caused
4581 incorrect handling when an unquoted " was present following a character
4582 that needed quoting.
4584 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4585 was not being matched caselessly.
4587 24. Arranged for all hyphens in the exim.8 source to be escaped with
4590 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4591 properly. Recipient callout cache records were still being keyed to include
4592 the sender, even when use_sender was set false. This led to far more
4593 callouts that were necessary. The sender is no longer included in the key
4594 when use_sender is false.
4596 26. Added "control = submission" modifier to ACLs.
4598 27. Added the ${base62d: operator to decode base 62 numbers.
4600 28. dnsdb lookups can now access SRV records.
4602 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4603 the configuration file.
4605 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4606 (-v) mode. This makes the output for a verbose queue run more intelligible.
4608 31. Added a use_postmaster feature to recipient callouts.
4610 32. Added the $body_zerocount variable, containing the number of binary zero
4611 bytes in the message body.
4613 33. The time of last modification of the "new" subdirectory is now used as the
4614 "mailbox time last read" when there is a quota error for a maildir
4617 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4619 35. Added +ignore_unknown as a special item in host lists.
4621 36. Code for decoding IPv6 addresses in host lists is now included, even if
4622 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4623 address was recognized as an IP address, but was then not correctly decoded
4624 into binary, causing unexpected and incorrect effects when compared with
4631 1. Very minor rewording of debugging text in manualroute to say "list of
4632 hosts" instead of "hostlist".
4634 2. If verify=header_syntax was set, and a header line with an unqualified
4635 address (no domain) and a large number of spaces between the end of the
4636 name and the colon was received, the reception process suffered a buffer
4637 overflow, and (when I tested it) crashed. This was caused by some obsolete
4638 code that should have been removed. The fix is to remove it!
4640 3. When running in the test harness, delay a bit after writing a bounce
4641 message to get a bit more predictability in the log output.
4643 4. Added a call to search_tidyup() just before forking a reception process. In
4644 theory, someone could use a lookup in the expansion of smtp_accept_max_
4645 per_host which, without the tidyup, could leave open a database connection.
4647 5. Added the variables $recipient_data and $sender_data which get set from a
4648 lookup success in an ACL "recipients" or "senders" condition, or a router
4649 "senders" option, similar to $domain_data and $local_part_data.
4651 6. Moved the writing of debug_print from before to after the "senders" test
4654 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4655 problems for message scanning, either using a data ACL, or using
4656 local_scan() because the Received: header was not generated till after they
4657 were called (in order to set the time as the time of reception completion).
4658 I have revised the way this works. The header is now generated after the
4659 body is received, but before the ACL or local_scan() are called. After they
4660 are run, the timestamp in the header is updated.
4666 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4667 before starting a queue runner without re-exec. This happened only when
4668 deliver_drop_privilege was set or when the Exim user was set to root. The
4669 effect of the bug was that timeouts during subsequent deliveries caused
4670 crashes instead of being properly handled. The handler is now left at its
4671 default (and expected) setting.
4673 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4674 message, again when deliver_drop_privilege is set or Exim is run as root.
4675 The bug described in (1) was not present in this case, but the tidying up
4676 of the other signals was missing. I have made the two cases consistent.
4678 3. The ignore_target_hosts setting on a manualroute router was being ignored
4679 for hosts that were looked up using the /MX notation.
4681 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4684 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4685 operated on the sender address. After changing the $sender_address to <>
4686 for the sender address verify, Exim was re-instated it as the original
4687 (before rewriting) address, but remembering that it had rewritten it, so it
4688 wasn't rewriting it again. This bug also had the effect of breaking the
4689 sender address verification caching when the sender address was rewritten.
4691 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4692 This has been changed so that if the ip literal address matches
4693 ignore_target_hosts, the router declines.
4695 7. Added expansion conditions match_domain, match_address, and match_local_
4696 part (NOT match_host).
4698 8. The placeholder for the Received: header didn't have a length field set.
4700 9. Added code to Exim itself and to exim_lock to test for a specific race
4701 condition that could lead to file corruption when using MBX delivery. The
4702 issue is with the lockfile that is created in /tmp. If this file is removed
4703 after a process has opened it but before that process has acquired a lock,
4704 there is the potential for a second process to recreate the file and also
4705 acquire a lock. This could lead to two Exim processes writing to the file
4706 at the same time. The added code performs the same test as UW imapd; it
4707 checks after acquiring the lock that its file descriptor still refers to
4708 the same named file.
4710 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4711 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4714 11. Added the smtp_active_hostname option. If used, this will typically be made
4715 to depend on the incoming interface address. Because $interface_address is
4716 not set up until the daemon has forked a reception process, error responses
4717 that can happen earlier (such as "too many connections") no longer contain
4720 12. If an expansion in a condition on a "warn" statement fails because a lookup
4721 defers, the "warn" statement is abandoned, and the next ACL statement is
4722 processed. Previously this caused the whole ACL to be aborted.
4724 13. Added the iplsearch lookup type.
4726 14. Added ident_timeout as a log selector.
4728 15. Added tls_certificate_verified as a log selector.
4730 16. Added a global option tls_require_ciphers (compare the smtp transport
4731 option of the same name). This controls incoming TLS connections.
4733 17. I finally figured out how to make tls_require_ciphers do a similar thing
4734 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4735 before starting the TLS session.
4737 18. Tabs are now shown as \t in -bP output.
4739 19. If the log selector return_path_on_delivery was set, Exim crashed when
4740 bouncing a message because it had too many Received: header lines.
4742 20. If two routers both had headers_remove settings, and the first one included
4743 a superfluous trailing colon, the final name in the first list and the
4744 first name in the second list were incorrectly joined into one item (with a
4745 colon in the middle).
4751 1. Added -C and -D options to the exinext utility, mainly to make it easier
4752 to include in the automated testing, but these could be helpful when
4753 multiple configurations are in use.
4755 2. The exinext utility was not formatting the output nicely when there was
4756 an alternate port involved in the retry record key, nor when there was a
4757 message id as well (for retries that were specific to a specific message
4758 and a specific host). It was also confused by IPv6 addresses, because of
4759 the additional colons they contain. I have fixed the IPv4 problem, and
4760 patched it up to do a reasonable job for IPv6.
4762 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4763 delivery, the log line now contains "pipelined" if PIPELINING was used.
4765 4. An SMTP transport process used to panic and die if the bind() call to set
4766 an explicit outgoing interface failed. This has been changed; it is now
4767 treated in the same way as a connect() failure.
4769 5. A reference to $sender_host_name in the part of a conditional expansion
4770 that was being skipped was still causing a DNS lookup. This no longer
4773 6. The def: expansion condition was not recognizing references to header lines
4774 that used bh_ and bheader_.
4776 7. Added the _cache feature to named lists.
4778 8. The code for checking quota_filecount in the appendfile transport was
4779 allowing one more file than it should have been.
4781 9. For compatibility with Sendmail, the command line option
4789 and sets the incoming protocol and host name (for trusted callers). The
4790 host name and its colon can be omitted when only the protocol is to be set.
4791 Note the Exim already has two private options, -pd and -ps, that refer to
4792 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4793 "s", but I don't think that's a major issue.
4795 10. A number of refactoring changes to the code, none of which should affect
4798 (a) The number of logging options was getting close to filling up the
4799 32-bit word that was used as a bit map. I have split them into two classes:
4800 those that are passed in the argument to log_write(), and those that are
4801 only ever tested independently outside of that function. These are now in
4802 separate 32-bit words, so there is plenty of room for expansion again.
4803 There is no change in the user interface or the logging behaviour.
4805 (b) When building, for example, log lines, the code previously used a
4806 macro that called string_cat() twice, in order to add two strings. This is
4807 not really sufficiently general. Furthermore, there was one instance where
4808 it was actually wrong because one of the argument was used twice, and in
4809 one call a function was used. (As it happened, calling the function twice
4810 did not affect the overall behaviour.) The macro has been replaced by a
4811 function that can join an arbitrary number of extra strings onto a growing
4814 (c) The code for expansion conditions now uses a table and a binary chop
4815 instead of a serial search (which was left over from when there were very
4816 few conditions). Also, it now recognizes conditions like "pam" even when
4817 the relevant support is not compiled in: a suitably worded error message is
4818 given if an attempt is made to use such a condition.
4820 11. Added ${time_interval:xxxxx}.
4822 12. A bug was causing one of the ddress fields not to be passed back correctly
4823 from remote delivery subprocesses. The field in question was not being
4824 subsequently used, so this caused to problems in practice.
4826 13. Added new log selectors queue_time and deliver_time.
4828 14. Might have fixed a bug in maildirsizefile handling that threw up
4829 "unexpected character" debug warnings, and recalculated the data
4830 unnecessarily. In any case, I expanded the warning message to give more
4833 15. Added the message "Restricted characters in address" to the statements in
4834 the default ACL that block characters like @ and % in local parts.
4836 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4837 Three changes have been made:
4839 (a) There was a serious bug; a negative response to MAIL caused the whole
4840 recipient domain to be cached as invalid, thereby blocking all messages
4841 to all local parts at the same domain, from all senders. This bug has
4842 been fixed. The domain is no longer cached after a negative response to
4843 MAIL if the sender used is not empty.
4845 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4848 (c) A new callout option, "use_sender" has been added for people who want
4849 the modified behaviour.
4855 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4858 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4859 indeed breaks things for older releases.
4861 3. Added additional logging to the case where there is a problem reading data
4862 from a filter that is running in a subprocess using a pipe, in order to
4863 try to track down a specific problem.
4865 4. Testing facility fudge: when running in the test harness and attempting
4866 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4867 getting "No route to host". Convert this to a timeout.
4869 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4872 6. Some OS don't have socklen_t but use size_t instead. This affects the
4873 fifth argument of getsockopt() amongst other things. This is now
4874 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4875 can be set for individual OS. I have set it for SunOS5, OSF1, and
4876 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4877 some earlier ones do not.
4879 7. Change 4.30/15 was not doing the test caselessly.
4881 8. The standard form for an IPv6 address literal was being rejected by address
4882 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4883 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4884 this, as well as the form without the "IPv6" on the front (but only when
4885 address literals are enabled, of course).
4887 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4889 10. Exim crashed if a message with an empty sender address specified by -f
4890 encountered a router with an errors_to setting. This could be provoked only
4891 by a command such as
4895 where an empty string was supplied; "<>" did not hit this bug.
4897 11. Installed PCRE release 4.5.
4899 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4900 remained set. It is now erased.
4902 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4903 times from message ids (which are base 36 rather than the normal 62).
4905 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4906 were being counted as actual protocol errors, and logged if the log
4907 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4908 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4909 and DATA following a set of rejected RCPTs do not count as protocol errors.
4910 In other words, Exim assumes they were pipelined, though this may not
4911 actually be the case. Of course, in all cases the client gets an
4912 appropriate error code.
4914 15. If a lookup fails in an ACL condition, a message about the failure may
4915 be available; it is used if testing the ACL cannot continue, because most
4916 such messages specify what the cause of the deferral is. However, some
4917 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4918 that caused an old message to be retained and used if a later statement
4919 caused a defer, replacing the real cause of the deferral.
4921 16. If an IP address had so many PTR records that the DNS lookup buffer
4922 was not large enough to hold them, Exim could crash while trying to process
4923 the truncated data. It now detects and logs this case.
4925 17. Further to 4.21/58, another change has been made: if (and only if) the
4926 first line of a message (the first header line) ends with CRLF, a bare LF
4927 in a subsequent header line has a space inserted after it, so as not to
4928 terminate the header.
4930 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4931 unnecessarily, used atoi() instead of strtol(), and didn't check the
4932 termination when getting file sizes from file names by regex.
4934 19. Completely re-implemented the support for maildirsize files, in the light
4935 of a number of problems with the previous contributed implementation
4936 (4.30/29). In particular:
4938 . If the quota is zero, the maildirsize file is maintained, but no quota is
4941 . If the maildir directory does not exist, it is created before any attempt
4942 to write a maildirsize file.
4944 . The quota value in the file is just a cache; if the quota is changed in
4945 the transport, the new value overrides.
4947 . A regular expression is available for excluding directories from the
4950 20. The autoreply transport checks the characters in options that define the
4951 message's headers; it allows continued headers, but it was checking with
4952 isspace() after an embedded newline instead of explicitly looking for a
4955 21. If all the "regular" hosts to which an address was routed had passed their
4956 expiry times, and had not reached their retry times, the address was
4957 bounced, even if fallback hosts were defined. Now Exim should go on to try
4960 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4961 equivalent code in the SMTP transport. Some hosts send humungous responses
4962 to HELO/EHLO, more than 1024 it seems.
4964 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4965 gives compiler warnings in some environments. I've now done it "properly",
4968 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4969 (because of problems with the built-in one) was declared to return uschar *
4970 instead of char *, causing compiler failure.
4972 25. Fixed a file descriptor leak when processing alias/forward files.
4974 26. Fixed a minor format string issue in dbfn.c.
4976 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4978 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4979 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4982 29. When a sender address is verified, it is cached, to save repeating the test
4983 when there is more than one recipient in a message. However, when the
4984 verification involves a callout, it is possible for different callout
4985 options to be set for different recipients. It is too complicated to keep
4986 track of this in the cache, so now Exim always runs a verification when a
4987 callout is required, relying on the callout cache for the optimization.
4988 The overhead is duplication of the address routing, but this should not be
4991 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4992 to be verified with callout=postmaster, and the main callout worked but the
4993 postmaster check failed, the verification correctly failed. However, if a
4994 subsequent RCPT command asked for sender verification *without* the
4995 postmaster check, incorrect caching caused this verification also to fail,
4998 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4999 it was not caching the DNS options (qualify_single, search_parents) that
5000 were used when the lookup failed. A subsequent lookup with different
5001 options therefore always gave the same answer, though there were cases
5002 where it should not have. (Example: a "domains = !$mx_any" option on a
5003 dnslookup router: the "domains" option is always processed without any
5004 widening, but the router might have qualify_single set.) Now Exim uses the
5005 cached value only when the same options are set.
5007 32. Added John Jetmore's "exipick" utility to the distribution.
5009 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5010 than a timeout (e.g. a certificate is required, and is not provided), an
5011 Exim server now closes the connection immediately. Previously it waited for
5012 the client to close - but if the client is SSL, it seems that they each
5013 wait for each other, leading to a delay before one of them times out.
5015 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5016 maintained 0.8.x compatibility because I don't think many are using it, and
5017 it is clearly obsolete.
5019 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5022 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5023 client certificate was expired. A simple patch fixes this, though I don't
5024 understand the full logic of why the verify callback is called multiple
5027 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5028 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5029 which causes problems with some clients (such as the Certicom SSL Plus
5030 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5031 disables the coutermeasure allowing Eudora to connect."
5033 38. Exim was not checking that a write() to a log file succeeded. This could
5034 lead to Bad Things if a log got too big, in particular if it hit a file
5035 size limit. Exim now panics and dies if it cannot write to a log file, just
5036 as it does if it cannot open a log file.
5038 39. Modified OS/Makefile-Linux so that it now contains
5040 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5042 The two -D definitions ensure that Exim is compiled with large file
5043 support, which makes it possible to handle log files that are bigger than
5046 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5047 instance) a domain was checked against a named list that involved a lookup,
5048 causing $domain_data to be set, then another domain was checked against the
5049 same list, then the first domain was re-checked, the value of $domain_data
5050 after the final check could be wrong. In particular, if the second check
5051 failed, it could be set empty. This bug probably also applied to
5054 41. The strip_trailing_dot option was not being applied to the address given
5055 with the -f command-line option.
5057 42. The code for reading a message's header from the spool was incrementing
5058 $received_count, but never initializing it. This meant that the value was
5059 incorrect (doubled) while delivering a message in the same process in which
5060 it was received. In the most common configuration of Exim, this never
5061 happens - a fresh exec is done - but it can happen when
5062 deliver_drop_privilege is set.
5064 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5065 it now includes up to 150 characters of the unexpected data in the log
5068 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5069 and building data strings. The size of both of these buffers was 10 000
5070 bytes - far larger than anybody would *ever* want, thought I. Needless to
5071 say, somebody hit the limit. I have increased the maximum line length to
5072 20 000 and the maximum data length of concatenated lines to 100 000. I have
5073 also fixed two bugs, because there was no checking on these buffers. Tsk,
5074 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5075 buffer is too small.
5077 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5078 lsearch lookups. Now it does.
5080 46. When parsing a route_list item in a manualroute router, a fixed-length
5081 buffer was used for the list of hosts. I made this 1024 bytes long,
5082 thinking that nobody would ever have a list of hosts that long. Wrong.
5083 Somebody had a whole pile of complicated expansion conditions, and the
5084 string was silently truncated, leading to an expansion error. It turns out
5085 that it is easier to change to an unlimited length (owing to other changes
5086 that have happened since this code was originally written) than to build
5087 structure for giving a limitation error. The length of the item that
5088 expands into the list of hosts is now unlimited.
5090 47. The lsearch lookup could not handle data where the length of text line was
5091 more than 4095 characters. Such lines were truncated, leading to shortened
5092 data being returned. It should now handle lines of any length.
5094 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5095 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5098 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5099 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5100 no longer happens in comment lines. A list of replacements is now placed
5101 at the head of all of the source files, except those whose only change is
5102 to replace PERL_COMMAND in the very first #! line.
5104 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5105 messages on the queue, with a bottom-up merge sort, using code contributed
5106 by Michael Haardt. This should make operations like -bp somewhat faster on
5107 large queues. It won't affect queue runners, except when queue_run_in_order
5110 51. Installed eximstats 1.31 in the distribution.
5112 52. Added support for SRV lookups to the dnslookup router.
5114 53. If an ACL referred to $message_body or $message_body_end, the value was not
5115 reset for any messages that followed in the same SMTP session.
5117 54. The store-handling optimization for building very long strings was not
5118 differentiating between the different store pools. I don't think this
5119 actually made any difference in practice, but I've tidied it.
5121 55. While running the routers to verify a sender address, $sender_address
5122 was still set to the sender address. This is wrong, because when routing to
5123 send a bounce to the sender, it would be empty. Therefore, I have changed
5124 it so that, while verifying a sender address, $sender_address is set to <>.
5125 (There is no change to what happens when verifying a recipient address.)
5127 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5128 target A or AAAA records (if not already returned) without resetting the
5129 qualify_single or search_parents options of the DNS resolver. These are
5130 inappropriate in this case because the targets of MX and SRV records must
5131 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5132 target that, when qualified, matched something in the local domain. These
5133 two options are now turned off when doing these lookups.
5135 57. It seems that at least some releases of Reiserfs (which does not have the
5136 concept of a fixed number of inodes) returns zero and not -1 for the
5137 number of available inodes. This interacted badly with check_spool_inodes,
5138 which assumed that -1 was the "no such thing" setting. What I have done is
5139 to check that the total number of inodes is greater than zero before doing
5140 the test of how many are available.
5142 58. When a "warn" ACL statement has a log_message modifier, the message is
5143 remembered, and not repeated. This is to avoid a lot of repetition when a
5144 message has many recipients that cause the same warning to be written.
5145 Howewer, Exim was preserving the list of already written lines for an
5146 entire SMTP session, which doesn't seem right. The memory is now reset if a
5147 new message is started.
5149 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5150 debugging output unless log_rewrite was also set.
5152 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5153 of (int)(handle) when we know that handle contains (void *)(-1).
5155 61. The Exim daemon panic-logs an error return when it closes the incoming
5156 connection. However "connection reset by peer" seems to be common, and
5157 isn't really an error worthy of noting specially, so that particular error
5160 62. When Exim is trying to find all the local interfaces, it used to panic and
5161 die if the ioctl to get the interface flags failed. However, it seems that
5162 on at least one OS (Solaris 9) it is possible to have an interface that is
5163 included in the list of interfaces, but for which you get a failure error
5164 for this call. This happens when the interface is not "plumbed" into a
5165 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5166 failure of the "get flags" call assumes that the interface is down.
5168 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5169 makes life easier for people who are doing arithmetic on fields extracted
5170 from dates, where you often get leading zeros that should not be
5171 interpreted as octal.
5173 64. Added qualify_domain to the redirect router, to override the global
5176 65. If a pathologically long header line contained very many addresses (the
5177 report of this problem mentioned 10 000) and each of them was rewritten,
5178 Exim could use up a very large amount of memory. (It kept on making new
5179 copies of the header line as it rewrote, and never released the old ones.)
5180 At the expense of a bit more processing, the header rewriting function has
5181 been changed so that it no longer eats memory in this way.
5183 66. The generation of the Received: header has been moved from the time that a
5184 message starts to be received, to the time that it finishes. The timestamp
5185 in the Received: header should now be very close to that of the <= log
5186 line. There are two side-effects of this change:
5188 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5189 logged header lines no longer include the local Received: line, because
5190 it has not yet been created. The same applies to a copy of the message
5191 that is returned to a non-SMTP sender when a message is rejected.
5193 (b) When a filter file is tested using -bf, no additional Received: header
5194 is added to the test message. After some thought, I decided that this
5197 This change does not affect the value of $received_for. It is still set
5198 after address rewriting, but before local_scan() is called.
5200 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5202 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5203 gave an unhelpful panic error message, and a defer error. I have managed to
5204 change this behaviour so that it now rejects any supplied certificate,
5205 which seems right, as the list of acceptable certificates is empty.
5207 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5208 gave an unhelpful defer error. I have not managed to make this reject any
5209 supplied certificates, but the error message it gives is "no certificate
5210 supplied", which is not helpful.
5212 70. exigrep's output now also includes lines that are not associated with any
5213 message, but which match the given pattern. Implemented by a patch from
5214 Martin Sluka, which also tidied up the Perl a bit.
5216 71. Recipient callout verification, like sender verification, was using <> in
5217 the MAIL FROM command. This isn't really the right thing, since the actual
5218 sender may affect whether the remote host accepts the recipient or not. I
5219 have changed it to use the actual sender in the callout; this means that
5220 the cache record is now keyed on a recipient/sender pair, not just the
5221 recipient address. There doesn't seem to be a real danger of callout loops,
5222 since a callout by the remote host to check the sender would use <>.
5223 [SEE ABOVE: changed after hitting problems.]
5225 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5226 temporary errors. However, in the case of such a code being given after
5227 the end of a data transmission (i.e. after ".") Exim was failing to write
5228 a retry record for the message. (Yes, there was some broken host that was
5229 actually sending 8xx at this point.)
5231 73. An unknown lookup type in a host list could cause Exim to panic-die when
5232 the list was checked. (An example that provoked this was putting <; in the
5233 middle of a list instead of at the start.) If this happened during a DATA
5234 ACL check, a -D file could be left lying around. This kind of configuration
5235 error no longer causes Exim to die; instead it causes a defer errror. The
5236 incident is still logged to the main and panic logs.
5238 74. Buglet left over from Exim 3 conversion. The message "too many messages
5239 in one connection" was written to the rejectlog but not the mainlog, except
5240 when address rewriting (yes!) was being logged.
5242 75. Added write_rejectlog option.
5244 76. When a system filter was run not as root (that is, when system_filter_user
5245 was set), the values of the $n variables were not being returned to the
5246 main process; thus, they were not subsequently available in the $sn
5249 77. Added +return_path_on_delivery log selector.
5251 78. A connection timeout was being treated differently from recipients deferred
5252 when testing hosts_max_try with a message that was older than the host's
5253 retry timeout. (The host should not be counted, thus allowing all hosts to
5254 be tried at least once before bouncing.) This may have been the cause of an
5255 occasionally reported bug whereby a message would remain on the queue
5256 longer than the retry timeout, but would be bounced if a delivery was
5257 forced. I say "may" because I never totally pinned down the problem;
5258 setting up timeout/retry tests is difficult. See also the next item.
5260 79. The ultimate address timeout was not being applied to errors that involved
5261 a combination of host plus message (for example, a timeout on a MAIL
5262 command). When an address resolved to a number of possible hosts, and they
5263 were not all tried for each delivery (e.g. because of hosts_max_try), a
5264 message could remain on the queue longer than the retry timeout.
5266 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5269 81. Fixed an obscure SMTP outgoing bug which required at least the following
5270 conditions: (a) there was another message waiting for the same server;
5271 (b) the server returned 5xx to all RCPT commands in the first message so
5272 that the message was not completed; (c) the server dropped the connection
5273 or gave a negative response to the RSET that Exim sends to abort the
5274 transaction. The observed case was a dropped connection after DATA that had
5275 been sent in pipelining mode. That is, the server had advertised PIPELINING
5276 but was not implementing it correctly. The effect of the bug was incorrect
5277 behaviour, such as trying another host, and this could lead to a crash.
5283 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5284 and daemon.c were passed as pointers to ints; they should have been
5285 pointers to socklen_t variables (which are typically unsigned ints).
5287 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5290 3. Fixed a really odd bug that affected only the testing scheme; patching a
5291 certain fixed string in the binary changed the value of another string that
5292 happened to be identical to the end of the original first string.
5294 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5295 name", it returns that address as the IP address. On some operating
5296 systems (e.g. Solaris), it also passes back the IP address string as the
5297 "host name". However, on others (e.g. Linux), it passes back an empty
5298 string. Exim wasn't checking for this, and was changing the host name to an
5299 empty string, assuming it had been canonicized.
5301 5. Although rare, it is permitted to have more than one PTR record for a given
5302 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5303 all the names associated with an address, because they do in Solaris.
5304 However, it seems that they do not in Linux for data that comes from the
5305 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5306 I found this out when I moved to a new Linux workstation and tried to run
5307 the Exim test suite.
5309 To get round this problem I have changed the code so that it now does its
5310 own call to the DNS to look up PTR records when searching for a host name.
5311 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5312 addresses that are only in /etc/hosts are still found.
5314 This behaviour is, however, controlled by an option called host_lookup_
5315 order, which defaults to "bydns:byaddr". If people want to use the other
5316 order, or indeed, just use one or the other means of lookup, they can
5317 specify it in this variable.
5319 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5320 some operating systems, this comes back from gethostbyaddr() as an empty
5321 string, and this is what Exim used to test for. However, it seems that in
5322 other systems, "." is yielded. Exim now tests for this case too.
5324 7. The values of check_spool_space and check_log_space are now held internally
5325 as a number of kilobytes instead of an absolute number of bytes. If a
5326 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5327 kilobyte. This means that much larger values can be stored.
5329 8. Exim monitor: an attempt to get the action menu when not actually pointing
5330 at a message produces an empty menu entitled "No message selected". This
5331 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5332 no entries in it ("Shell widget menu has zero width and/or height"). So I
5333 have added a single, blank menu entry in this case.
5335 9. Added ${quote_local_part.
5337 10. MIME decoding is now applied to the contents of Subject: header lines when
5340 11. Now that a reference to $sender_host_address automatically causes a reverse
5341 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5342 host lookup before query-style lookups in lists that might use this
5343 variable. This has therefore been abolished, and the "net-" prefix is no
5344 longer necessary for query-style lookups.
5346 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5347 have been a typo for LFLAGS, so it has been changed.
5349 13. The install script calls Exim with "-C /dev/null" in order to find the
5350 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5351 to be output. Howeve, since Exim outputs its version number before the
5352 error, it didn't break the script. It just looked ugly. I fixed this by
5353 always allowing "-C /dev/null" if the caller is root.
5355 14. Ignore overlarge ACL variable number when reading spool file - insurance
5356 against a later release with more variables having written the file.
5358 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5359 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5360 this, as well as the form without the "IPv6" on the front.
5362 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5363 OS/Makefile-Darwin file.
5365 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5366 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5368 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5369 variable. However, some OS use ints for inodes. I've added cast to long int
5370 to get rid of the compiler warning.
5372 19. I had forgotten to lock out "/../" in configuration file names when
5373 ALT_CONFIG_PREFIX was set.
5375 20. Routers used for verification do not need to specify transports. However,
5376 if such a router generated a host list, and callout was configured, Exim
5377 crashed, because it could not find a port number from the (non-existent)
5378 transport. It now assumes port 25 in this circumstance.
5380 21. Added the -t option to exigrep.
5382 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5383 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5384 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5385 them is set, LOOKUP_LSEARCH is forced.
5387 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5389 24. Added sender and host information to the "rejected by local_scan()" log
5390 line; previously there was no indication of these.
5392 25. Added .include_if_exists.
5394 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5395 Linux. It turns out that not all file systems support this. Apparently some
5396 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5397 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5398 not supported on the file descriptor, is now ignored when Exim is trying to
5399 sync a directory. This applies only to Linux.
5401 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5403 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5404 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5407 29. Added support for maildirsize files from supplied patch (modified a bit).
5409 30. The use of :fail: followed by an empty string could lead Exim to respond to
5410 sender verification failures with (e.g.):
5412 550 Verification failed for <xxx>
5413 550 Sender verify failed
5415 where the first response line was missing the '-' that indicates it is not
5416 the final line of the response.
5418 31. The loop for finding the name of the user that called Exim had a hardwired
5419 limit of 10; it now uses the value of finduser_retries, which is used for
5420 all other user lookups.
5422 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5425 33. Exim was neglecting to zero errno before one call of strtol() when
5426 expanding a string and expecting an integer value. On some systems this
5427 resulted in spurious "integer overflow" errors. Also, it was casting the
5428 result into an int without checking.
5430 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5431 did not work. The code looks as if it has *never* worked, though it appears
5432 to have been documented since at least releast 1.62. I have made it work.
5434 35. The "timeout_DNS" error in retry rules, also documented since at least
5435 1.62, also never worked. As it isn't clear exactly what this means, and
5436 clearly it isn't a major issue, I have abolished the feature by treating it
5437 as "timeout", and writing a warning to the main and panic logs.
5439 36. The display of retry rules for -brt wasn't always showing the error code
5442 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5443 timeout_connect_A, timeout_connect_MX.
5445 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5446 to the empty sender.
5448 39. The daemon was not analysing the content of -oX till after it had closed
5449 stderr and disconnected from the controlling terminal. This meant that any
5450 syntax errors were only noted on the panic log, and the return code from
5451 the command was 0. By re-arranging the code a little, I've made the
5452 decoding happen first, so such errors now appear on stderr, and the return
5453 code is 1. However, the actual setting up of the sockets still happens in
5454 the disconnected process, so errors there are still only recorded on the
5457 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5458 connections (as happens on some IP stacks) was logged at start up time as
5459 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5460 it from "IPv6 and IPv4", which means that two separate sockets are being
5463 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5464 says whether AF_INET or AF_INET6 was passed as an argument.
5466 42. Exiwhat output was messed up when time zones were included in log
5469 43. Exiwhat now gives more information about the daemon's listening ports,
5470 and whether -tls-on-connect was used.
5472 44. The "port" option of the smtp transport is now expanded.
5474 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5475 silently ignored. Now an error message is written to the main and panic
5478 46. There's a new ACL modifier called "logwrite" which writes to a log file
5479 as soon as it is encountered.
5481 47. Added $local_user_uid and $local_user_gid at routing time.
5483 48. Exim crashed when trying to verify a sender address that was being
5486 49. Exim was recognizing only a space character after ".include". It now also
5487 recognizes a tab character.
5489 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5490 extracting the relevant information from the specification. The man page no
5491 longer contains scrambled data for the -d option, and I've added a section
5492 at the front about calling Exim under different names.
5494 51. Added "extra_headers" argument to the "mail" command in filter files.
5496 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5499 53. Installed eximstats 1.29.
5501 54. Added transport_filter_timeout as a generic transport option.
5503 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5504 Cc: header lines. This was required by RFC 822, but it not required by RFC
5507 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5508 incoming messages that did not have them. Now it does so only if the
5509 message originates locally, that is, if there is no associated remote host
5510 address. When Resent- header lines are present, this applies to the Resent-
5511 lines rather than the non-Resent- lines.
5513 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5514 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5516 58. Messages for configuration errors now include the name of the main
5517 configuration file - useful now that there may be more than one file in a
5518 list (.included file names were always shown).
5520 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5521 for those rare installations that do not start the daemon as root or run it
5522 setuid root. I've cut out the call to initgroups() if the daemon is not
5525 60. The Exim user and group can now be bound into the binary as text strings
5526 that are looked up at the start of Exim's processing.
5528 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5530 62. Added $mailstore_basename variable.
5532 63. Installed patch to sieve.c from Michael Haardt.
5534 64. When Exim failed to open the panic log after failing to open the main log,
5535 the original message it was trying to log was written to stderr and debug
5536 output, but if they were not available (the usual case in production), it
5537 was lost. Now it is written to syslog before the two lines that record the
5538 failures to open the logs.
5540 65. Users' Exim filters run in subprocesses under the user's uid. It is
5541 possible for a "deliver" command or an alias in a "personal" command to
5542 provoke an address rewrite. If logging of address rewriting is configured,
5543 this fails because the process is not running as root or exim. There may be
5544 a better way of dealing with this, but for the moment (because 4.30 needs
5545 to be released), I have disabled address rewrite logging when running a
5546 filter in a non-root, non-exim process.
5552 1. The buildconfig auxiliary program wasn't quoting the value set for
5553 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5554 not defined. This bug was masked in 4.22 by the effect that was fixed in
5557 2. Some messages that were rejected after a message id was allocated were
5558 shown as "incomplete" by exigrep. It no longer does this for messages that
5559 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5561 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5562 have allow_domain_literals set, the ID did not get logged in the <= line.
5563 Domain literals are now always recognized in Message-ID: header lines.
5565 4. The first argument for a ${extract expansion item is the key name or field
5566 number. Leading and trailing spaces in this item were not being ignored,
5567 causing some misleading effects.
5569 5. When deliver_drop_privilege was set, single queue runner processes started
5570 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5571 same command in the process it spins off) were not dropping privilege.
5573 6. When the daemon running as "exim" started a queue runner, it always
5574 re-executed Exim in the spun-off process. This is a waste of effort when
5575 deliver_drop_privilege is set. The new process now just calls the
5576 queue-runner function directly.
5582 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5585 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5586 ignored. Though the use of -oP was forcing the writing of a pid file, it
5587 was always written to the default place.
5589 3. If the message "no IP address found for host xxxx" is generated during
5590 incoming verification, it is now followed by identification of the incoming
5591 connection (so you can more easily find what provoked it).
5593 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5595 5. Added some features to "harden" Exim a bit more against certain attacks:
5597 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5598 be put in Local/Makefile. This is like the never_users runtime option,
5599 but it cannot be overridden. The default setting is "root".
5601 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5602 prefix string with which any file named in a -C command line option
5605 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5606 is retained for -C and -D only if the caller of Exim is root. Without
5607 it, the exim user may also use -C and -D and retain privilege.
5609 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5610 command line option is disabled.
5612 6. Macro names set by the -D option must start with an upper case letter, just
5613 like macro names defined in the configuration file.
5615 7. Added "dereference=" facility to LDAP.
5617 8. Two instances of the typo "uknown" in the source files are fixed.
5619 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5620 the Configure-Makefile script screwed up while processing it.
5622 10. Incorporated PCRE 4.4.
5624 11. The SMTP synchronization check was not operating right at the start of an
5625 SMTP session. For example, it could not catch a HELO sent before the client
5626 waited for the greeting. There is now a check for outstanding input at the
5627 point when the greeting is written. Because of the duplex, asynchronous
5628 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5629 way, but not yet received, when the check is performed.
5631 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5632 on TCP/IP sockets, because this apparently causes some broken clients to
5635 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5636 unchanged) from the Cygwin maintainer.
5638 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5639 was supported instead of testing for mbx. Effectively a typo.
5641 15. The spa authenticator server code was not checking that the input it
5642 received was valid base64.
5644 16. The debug output line for the "set" modifier in ACLs was not showing the
5645 name of the variable that was being set.
5647 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5649 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5650 lookup on its own. Something else was needed to trigger the lookup. For
5651 example, a match in host_lookup or the need for a host name in a host list.
5652 Now, if $sender_host_name is referenced and the host name has not yet been
5653 looked up, a lookup is performed. If the lookup fails, the variable remains
5654 empty, and $host_lookup_failed is set to "1".
5656 19. Added "eqi" as a case-independent comparison operator.
5658 20. The saslauthd authentication condition could segfault if neither service
5659 nor realm was specified.
5661 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5662 error message that was logged was misleading, and incoming SMTP
5663 connections were dropped. The message is now more accurate, and temporary
5664 errors are given to SMTP connections.
5666 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5667 (except RSET) with a 421 error, until QUIT is received. However, it was
5668 failing to send a response to QUIT.
5670 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5671 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5672 ACL is now run after the helo_try_verify_hosts code.
5674 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5675 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5676 case-independent, so other case variants are also recognized). Apparently
5677 some systems use these upper case variants.
5679 25. If more than two messages were waiting for the same host, and a transport
5680 filter was specified for the transport, Exim sent two messages over the
5681 same TCP/IP connection, and then failed with "socket operation on non-
5682 socket" when it tried to send the third.
5684 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5686 27. The extern definition of crypt16() in expand.c was not being excluded when
5687 the OS had its own crypt16() function.
5689 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5690 as a preferred synonym for return_size_limit, both as an option and as an
5693 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5695 30. Changed the default configuration ACL to relax the local part checking rule
5696 for addresses that are not in any local domains. For these addresses,
5697 slashes and pipe symbols are allowed within local parts, but the sequence
5698 /../ is explicitly forbidden.
5700 31. SPA server authentication was not clearing the challenge buffer before
5703 32. log_message in a "warn" ACL statement was writing to the reject log as
5704 well as to the main log, which contradicts the documentation and doesn't
5705 seem right (because no rejection is happening). So I have stopped it.
5707 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5708 However, I am unable to do any testing of this.
5710 34. Fixed an infelicity in the appendfile transport. When checking directories
5711 for a mailbox, to see if any needed to be created, it was accidentally
5712 using path names with one or more superfluous leading slashes; tracing
5713 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5715 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5716 contents are added to the log line that is written for every discarded
5717 recipient. (Previously a log_message setting was ignored.)
5719 36. The ${quote: operator now quotes the string if it is empty.
5721 37. The install script runs exim in order to find its version number. If for
5722 some reason other than non-existence or emptiness, which it checks, it
5723 could not run './exim', it was installing it with an empty version number,
5724 i.e. as "exim-". This error state is now caught, and the installation is
5727 38. An argument was missing from the function that creates an error message
5728 when Exim fails to connect to the socket for saslauthd authentication.
5729 This could cause Exim to crash, or give a corrupted message.
5731 39. Added isip, isip4, and isip6 to ${if conditions.
5733 40. The ACL variables $acl_xx are now saved with the message, and can be
5734 accessed later in routers, transports, and filters.
5736 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5737 strings in the file are not string-expanded.
5739 42. If a MAIL command specified a SIZE value that was too large to fit into an
5740 int variable, the check against message_size_limit failed. Such values are
5741 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5742 day this will have to be increased, but I don't think I want to be around
5743 when emails are that large.
5750 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5751 iconv() is not standard in FreeBSD.
5753 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5754 IPv6 enabled. The observed symptom was a segmentation fault on return from
5755 the function os_common_find_running_interfaces() in src/os.c.
5757 3. In the check_special_case() function in daemon.c I had used "errno" as an
5758 argument name, which causes warnings on some systems. This was basically a
5759 typo, since it was named "eno" in the comments!
5761 4. The code that waits for the clock to tick (at a resolution of some fraction
5762 of a second) so as to ensure message-id uniqueness was always waiting for
5763 at least one whole tick, when it could have waited for less. [This is
5764 almost certainly not relevant at current processor speeds, where it is
5765 unlikely to ever wait at all. But we try to future-proof.]
5767 5. The function that sleeps for a time interval that includes fractions of a
5768 second contained a race. It did not block SIGALRM between setting the
5769 timer, and suspending (a couple of lines later). If the interval was short
5770 and the sigsuspend() was delayed until after it had expired, the suspension
5771 never ended. On busy systems this could lead to processes getting stuck for
5774 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5775 process, before it forks any delivery processes. The open lookup caching
5776 mechanism meant that the open file or database connection was passed into
5777 the delivery process. The problem was that delivery processes always tidy
5778 up cached lookup data. This could cause a problem for the next delivery
5779 process started by the queue runner, because the external queue runner
5780 process does not know about the closure. So the next delivery process
5781 still has data in the lookup cache. In the case of a file lookup, there was
5782 no problem because closing a file descriptor in a subprocess doesn't affect
5783 the parent. However, if the lookup was caching a connection to a database,
5784 the connection was closed, and the second delivery process was likely to
5785 see errors such as "PGSQL: query failed: server closed the connection
5786 unexpectedly". The problem has been fixed by closing all cached lookups
5787 in a queue runner before running a delivery process.
5789 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5790 seem to have the "const" qualifier which it has on other OS. I've
5793 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5794 *of the same type* (client or server) with the same public name that an
5795 error should be diagnosed.
5797 9. When Exim looked up a host name for an IP address, but failed to find the
5798 original IP address when looking up the host name (a safety check), it
5799 output the message "<ip address> does not match any IP for NULL", which was
5800 confusing, to say the least. The bug was that the host name should have
5801 appeared instead of "NULL".
5803 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5804 user that is built into the binary, and the -C or -D options is used, root
5805 privilege is dropped before the configuration file is read. In addition,
5806 logging is switched to stderr instead of the normal log files. If the
5807 configuration then re-defines the Exim user, the unprivileged environment
5808 is probably not what is expected, so Exim logs a panic warning message (but
5811 However, if deliver_drop_privilege is set, the unprivileged state may well
5812 be exactly what is intended, so the warning has been cut out in that case,
5813 and Exim is allowed to try to write to its normal log files.
5819 1. smtp_return_error_details was not giving details for temporary sender
5820 or receiver verification errors.
5822 2. Diagnose a configuration error if two authenticators have the same public
5825 3. Exim used not to create the message log file for a message until the first
5826 delivery attempt. This could be confusing when incoming messages were held
5827 for policy or load reasons. The message log file is now created at the time
5828 the message is received, and an initial "Received" line is written to it.
5830 4. The automatically generated man page for command line options had a minor
5831 bug that caused no ill effects; however, a more serious problem was that
5832 the procedure for building the man page automatically didn't always
5833 operate. Consequently, release 4.20 contains an out-of-date version. This
5834 shouldn't happen again.
5836 5. When building Exim with embedded Perl support, the script that builds the
5837 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5838 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5840 6. The freeze_tell option was not being used for messages that were frozen on
5841 arrival, either by an ACL or by local_scan().
5843 7. Added the smtp_incomplete_transaction log selector.
5845 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5846 was accepting AUTH without a new EHLO.
5848 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5849 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5852 10. Logging of TCP/IP connections (when configured) now happens in the main
5853 daemon process instead of the child process, so that the TCP/IP connection
5854 count is more accurate (but it can never be perfect).
5856 11. The use of "drop" in a nested ACL was not being handled correctly in the
5857 outer ACL. Now, if condition failure induced by the nested "drop" causes
5858 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5859 or "require"), the connection is dropped.
5861 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5862 that yield "discard" can now be used with an "accept" or a "discard" verb,
5863 but an error is generated for any others (because I can't see a useful way
5864 to define what should happen).
5866 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5867 are now processed in the same way as lines in the Exim configuration file.
5868 In particular, continuation lines are supported.
5870 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5872 15. Added -ti meaning -t -i.
5874 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5875 domains, and warn by logging if others are found.
5877 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5878 returned from GIFCONF when Exim is trying to find the list of interfaces on
5879 a host. The code in os.c has been modified to copy each ifreq to an aligned
5880 structure in all cases.
5882 Also, in some cases, the returned ifreq's were being copied to a 'struct
5883 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5884 means the last couple of bytes of an IPv6 address could be chopped if the
5885 ifreq contained only a normal sockaddr (14 bytes storage).
5887 18. Named domain lists were not supported in the hosts_treat_as_local option.
5888 An entry such as +xxxx was not recognized, and was treated as a literal
5891 19. Ensure that header lines added by a DATA ACL are included in the reject log
5892 if the ACL subsequently rejects the message.
5894 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5895 MD5 (which is deprecated).
5897 21. When testing a filter file using -bf, Exim was writing a message when it
5898 took the sender from a "From " line in the message, but it was not doing so
5899 when it took $return_path from a Return-Path: header line. It now does.
5901 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5902 with a valid header line field name (a series of printing characters
5903 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5905 23. Changed "disc" in the source to "disk" to conform to the documentation and
5906 the book and for uniformity.
5908 24. Ignore Sendmail's -Ooption=value command line item.
5910 25. When execve() failed while trying to run a command in a pipe transport,
5911 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5912 could be confused with a return value of 69 from the command itself. This
5913 has been changed to 127, the value the shell returns if it is asked to run
5914 a non-existent command. The wording for the related log line suggests a
5915 non-existent command as the problem.
5917 26. If received_header_text expands to an empty string, do not add a Received:
5918 header line to the message. (Well, it adds a token one on the spool, but
5919 marks it "old" so that it doesn't get used or transmitted.)
5921 27. Installed eximstats 1.28 (addition of -nt option).
5923 28. There was no check for failure on the call to getsockname() in the daemon
5924 code. This can fail if there is a shortage of resources on the system, with
5925 ENOMEM, for example. A temporary error is now given on failure.
5927 29. Contrary to the C standard, it seems that in some environments, the
5928 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5929 program. Exim now does this explicitly; it affects the formatting of
5930 timestamps using strftime().
5932 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5933 complaints. I've now initialized all the variables, to avoid this.
5935 32. Header lines added by a system filter were not being "seen" during
5936 transport-time rewrites.
5938 33. The info_callback() function passed to OpenSSL is set up with type void
5939 (*)(SSL *, int, int), as described somewhere. However, when calling the
5940 function (actually a macro) that sets it up, the type void(*)() is
5941 expected. I've put in a cast to prevent warnings from picky compilers.
5943 34. If a DNS black list lookup found a CNAME record, but there were no A
5944 records associated with the domain it pointed at, Exim crashed.
5946 35. If a DNS black list lookup returned more than one A record, Exim ignored
5947 all but the first. It now scans all returned addresses if a particular IP
5948 value is being sought. In this situation, the contents of the
5949 $dnslist_value variable are a list of all the addresses, separated by a
5952 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5953 to accept a host name and all its aliases if the forward lookup for any of
5954 them yielded the IP address of the incoming connection. Now it accepts only
5955 those names whose forward lookup yields the correct IP address. Any other
5956 names are discarded. This closes a loophole whereby a rogue DNS
5957 administrator could create reverse DNS records to break through a
5958 wildcarded host restriction in an ACL.
5960 37. If a user filter or a system filter that ran in a subprocess used any of
5961 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5962 the wrong values were passed to the pipe command ($thisaddress had the
5963 value of $0, $0 had the value of $1, etc). This bug was introduced by
5964 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5966 38. Improved the line breaking for long SMTP error messages from ACLs.
5967 Previously, if there was no break point between 40 and 75 characters, Exim
5968 left the rest of the message alone. Two changes have been made: (a) I've
5969 reduced the minimum length to 35 characters; (b) if it can't find a break
5970 point between 35 and 75 characters, it looks ahead and uses the first one
5971 that it finds. This may give the occasional overlong line, but at least the
5972 remaining text gets split now.
5974 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5975 file descriptors might be low, and that setting 1000 would always raise it.
5976 It turns out that in some environments, the limit is already over 1000 and
5977 that lowering it causes trouble. So now Exim takes care not to decrease it.
5979 40. When delivering a message, the value of $return_path is set to $sender_
5980 address at the start of routing (routers may change the value). By an
5981 oversight, this default was not being set up when an address was tested by
5982 -bt or -bv, which affected the outcome if any router or filter referred to
5985 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5986 log only when "message" or "log_message" are set. However, if one of the
5987 conditions was an address verification, or a call to a nested ACL, the
5988 messages generated by the underlying test were being passed through. This
5989 no longer happens. The underlying message is available in $acl_verify_
5990 message for both "message" and "log_message" expansions, so it can be
5991 passed through if needed.
5993 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5994 new expansion $bh_ to give the encoded byte string without charset
5995 translation. Translation happens only if iconv() is available; HAVE_ICONV
5996 indicates this at build time. HEADERS_CHARSET gives the charset to
5997 translate to; headers_charset can change it in the configuration, and
5998 "headers charset" can change it in an individual filter file.
6000 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6001 that creates RFC 2047 encoded "words" labels them as that charset instead
6002 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6003 expansion operator; (ii) when Exim creates a From: line for a local
6004 message; (iii) when a header line is rewritten to include a "phrase" part.
6006 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6007 buggy, causing it to skip the first lines of messages whose message ID
6008 ended in 'D'. This would not have bitten before Exim release 4.14, because
6009 message IDs were unlikely to end in 'D' before then. The effect was to have
6010 incorrect size information for certain domains.
6012 45. #include "config.h" was missing at the start of the crypt16.c module. This
6013 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6016 46. If there was a timeout during a "random" callout check, Exim treated it as
6017 a failure of the random address, and carried on sending RSET and the real
6018 address. If the delay was just some slowness somewhere, the response to the
6019 original RCPT would be taken as a response to RSET and so on, causing
6020 mayhem of various kinds.
6022 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6023 when I implemented it. It didn't allow for the fact that some option values
6024 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6025 the right test for positive values.
6027 48. Domain names in DNS records are case-independent. Exim always looks them up
6028 in lower case. Some resolvers return domain names in exactly the case they
6029 appear in the zone file, that is, they may contain uppercase letters. Not
6030 all resolvers do this - some return always lower case. Exim was treating a
6031 change of case by a resolver as a change of domain, similar to a widening
6032 of a domain abbreviation. This triggered its re-routing code and so it was
6033 trying to route what was effectively the same domain again. This normally
6034 caused routing to fail (because the router wouldn't handle the domain
6035 twice). Now Exim checks for this case specially, and just changes the
6036 casing of the domain that it ultimately uses when it transmits the message
6039 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6042 50. If a filter generated a file delivery with a non-absolute name (possible if
6043 no home directory exists for the router), the forbid_file option was not
6046 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6047 the existing equality matching.
6049 52. Exim was using ints instead of ino_t variables in some places where it was
6050 dealing with inode numbers.
6052 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6053 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6054 called TMPDIR, and if it finds it is different, it changes its value.
6056 54. The smtp_printf() function is now made available to local_scan() so
6057 additional output lines can be written before returning. There is also an
6058 smtp_fflush() function to enable the detection of a dropped connection.
6059 The variables smtp_input and smtp_batched_input are exported to
6062 55. Changed the default runtime configuration: the message "Unknown user"
6063 has been removed from the ACL, and instead placed on the localuser router,
6064 using the cannot_route_message feature. This means that any verification
6065 failures that generate their own messages won't get overridden. Similarly,
6066 the "Unrouteable address" message that was in the ACL for unverifiable
6067 relay addresses has also been removed.
6069 56. Added hosts_avoid_esmtp to the smtp transport.
6071 57. The exicyclog script was not checking for the esoteric option
6072 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6073 will work only if exicyclog is run under the appropriate euid.
6075 58. Following a discussion on the list, the rules by which Exim recognises line
6076 endings on incoming messages have been changed. The -dropcr and drop_cr
6077 options are now no-ops, retained only for backwards compatibility. The
6078 following line terminators are recognized: LF CRLF CR. However, special
6079 processing applies to CR:
6081 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6082 nor a local message in the state where . is a terminator.
6084 (ii) If a bare CR is encountered in a header line, an extra space is added
6085 after the line terminator so as not to end the header. The reasoning
6086 behind this is that bare CRs in header lines are most likely either
6087 to be mistakes, or people trying to play silly games.
6089 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6090 was being incorrectly given as 18 bytes larger than it should have been.
6091 This is a VOB (very old bug).
6093 60. This may never have affected anything current, but just in case it has:
6094 When the local host is found other than at the start of a list of hosts,
6095 the local host, those with the same MX, and any that follow, are discarded.
6096 When the list in question was part of a longer list of hosts, the following
6097 hosts (not currently being processed) were also being discarded. This no
6098 longer happens. I'm not sure if this situation could ever has previously
6101 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6104 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6105 header. This is something that is recommended in a new Internet Draft, and
6106 is something that is documented as being done by Sendmail. There are two
6107 possible values. For messages generated by the autoreply transport, Exim
6110 Auto-Submitted: auto-replied
6112 whereas for all other generated messages (e.g. bounces) it adds
6114 Auto-Submitted: auto-generated
6116 63. The "personal" condition in filters now includes a test for the
6117 Auto-Submitted: header. If it contains the string "auto-" the message it
6118 not considered personal.
6120 64. Added rcpt_include_affixes as a generic transport option.
6122 65. Added queue_only_override (default true).
6124 66. Added the syslog_duplication option.
6126 67. If what should have been the first header line of a message consisted of
6127 a space followed by a colon, Exim was mis-interpreting it as a header line.
6128 It isn't of course - it is syntactically invalid and should therefore be
6129 treated as the start of the message body. The misbehaviour could have
6130 caused a number of strange effects, including loss of data in subsequent
6131 header lines, and spool format errors.
6133 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6134 client host had authenticated. This control can now be exercised by an ACL
6135 for more flexibility.
6137 69. By default, callouts do not happen when testing with -bh. There is now a
6138 variant, -bhc, which does actually run the callout code, including
6139 consulting and updating the callout cache.
6141 70. Added support for saslauthd authentication, courtesy of Alexander
6144 71. If statvfs() failed on the spool or log directories while checking their
6145 size for availability, Exim confusingly gave the error "space shortage".
6146 Furthermore, in debugging mode it crashed with a floating point exception.
6147 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6148 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6149 serious problem, Exim now writes to the main and panic logs when this
6150 happens, with details of the failure. It then refuses to accept the
6151 incoming message, giving the message "spool directory problem" or "log
6152 directory problem" with a 421 code for SMTP messages.
6154 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6155 0, 1, and 2 exist, because some OS complain for execs without them (see
6156 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6157 descriptors. However, the code omitted to check that the open succeeded,
6158 causing mysterious errors if for some reason the permissions on /dev/null
6159 got screwed. Now Exim writes a message to the main and panic logs, and
6160 bombs out if it can't open /dev/null.
6162 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6163 interact so that it is all more flexible. It is supposed to remain
6164 backwards compatible. Also added extra_local_interfaces.
6166 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6167 to bomb out with an assertion failure - to the client this appears as a
6168 connection drop. This problem occurs in the part of the code that was taken
6169 from the Samba project. Fortunately, the assertion is in a very simple
6170 function, so I have fixed this by reproducing the function inline in the
6171 one place where it is called, and arranging for authentication to fail
6172 instead of killing the process with assert().
6174 75. The SPA client code was not working when the server requested OEM rather
6175 than Unicode encoding.
6177 76. Added code to make require_files with a specific uid setting more usable in
6178 the case where statting the file as root fails - usually a non-root-mounted
6179 NFS file system. When this happens and the failure is EACCES, Exim now
6180 forks a subprocess and does the per-uid checking as the relevant uid.
6182 77. Added process_log_path.
6184 78. If log_file_path was not explicitly set, a setting of check_log_space or
6185 check_log_inodes was ignored.
6187 79. If a space check for the spool or log partitions fails, the incident is now
6188 logged. Of course, in the latter case the data may get lost...
6190 80. Added the %p formatting code to string_format() so that it can be used to
6191 print addresses in debug_print(). Adjusted all the address printing in the
6192 debugging in store.c to use %p rather than %d.
6194 81. There was a concern that a line of code in smtp_in.c could overflow a
6195 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6196 initially expressed, the concern was not well-founded, because trailing
6197 spaces are removed early. However, if the trailing spaces were followed by
6198 a NULL, they did not get removed, so the overflow was possible. Two fixes
6201 (a) I re-wrote the offending code in a cleaner fashion.
6202 (b) If an incoming SMTP command contains a NULL character, it is rejected
6205 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6206 runs initgroups(), so that if the Exim user is in any additional groups,
6207 they will be used during message reception.
6213 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob