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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
42 TL/01 When running the test suite, the README says that variables such as
43 no_msglog_check are global and can be placed anywhere in a specific
44 test's script, however it was observed that placement needed to be near
45 the beginning for it to behave that way. Changed the runtest perl
46 script to read through the entire script once to detect and set these
47 variables, reset to the beginning of the script, and then run through
48 the script parsing/test process like normal.
50 TL/02 The BSD's have an arc4random API. One of the functions to induce
51 adding randomness was arc4random_stir(), but it has been removed in
52 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
53 function when detected.
55 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
56 cause callback expansion.
58 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
59 syntax errors in an expansion can be treated as a string instead of
60 logging or causing an error, due to the internal use of bool_lax
61 instead of bool when processing it.
63 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
64 server certificates when making smtp deliveries.
66 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
68 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
70 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
72 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
73 Merged patch from Sebastian Wiedenroth.
75 JH/05 Fix results-pipe from transport process. Several recipients, combined
76 with certificate use, exposed issues where response data items split
77 over buffer boundaries were not parsed properly. This eventually
78 resulted in duplicates being sent. This issue only became common enough
79 to notice due to the introduction of conection certificate information,
80 the item size being so much larger. Found and fixed by Wolfgang Breyha.
82 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
83 size buffer was used, resulting in syntax errors when an expansion
86 JH/07 Add support for directories of certificates when compiled with a GnuTLS
87 version 3.3.6 or later.
89 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
90 is EXPERIMENTAL_EVENT, the main-configuration and transport options
91 both become "event_action", the variables become $event_name, $event_data
92 and $event_defer_errno. There is a new variable $verify_mode, usable in
93 routers, transports and related events. The tls:cert event is now also
94 raised for inbound connections, if the main configuration event_action
97 TL/06 In test suite, disable OCSP for old versions of openssl which contained
98 early OCSP support, but no stapling (appears to be less than 1.0.0).
100 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
101 server certificate names available under the smtp transport option
102 "tls_verify_cert_hostname" now do not permit multi-component wildcard
105 JH/10 Time-related extraction expansions from certificates now use the main
106 option "timezone" setting for output formatting, and are consistent
107 between OpenSSL and GnuTLS compilations. Bug 1541.
109 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
110 encoded parameter in the incoming message. Bug 1558.
112 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
113 include certificate info, eximon was claiming there were spoolfile
116 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
118 JH/14 Log delivery-related information more consistently, using the sequence
119 "H=<name> [<ip>]" wherever possible.
121 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
122 are problematic for Debian distribution, omit them from the release
125 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
127 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
130 JH/17 Fix a null-indirection in certextract expansions when a nondefault
131 output list separator was used.
136 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
137 checkers that were complaining about end of non-void function with no
140 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
141 This was a regression intruduced in 4.83 by another bugfix.
143 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
145 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
146 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
152 TF/01 Correctly close the server side of TLS when forking for delivery.
154 When a message was received over SMTP with TLS, Exim failed to clear up
155 the incoming connection properly after forking off the child process to
156 deliver the message. In some situations the subsequent outgoing
157 delivery connection happened to have the same fd number as the incoming
158 connection previously had. Exim would try to use TLS and fail, logging
159 a "Bad file descriptor" error.
161 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
162 utilities have not been installed.
164 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
165 temporary space as the ACL may create new global variables.
167 TL/01 LDAP support uses per connection or global context settings, depending
168 upon the detected version of the libraries at build time.
170 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
171 to extract and use the src ip:port in logging and expansions as if it
172 were a direct connection from the outside internet. PPv2 support was
173 updated based on HAProxy spec change in May 2014.
175 JH/02 Add ${listextract {number}{list}{success}{fail}}.
177 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
178 Properly escape header and check for NULL return.
180 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
183 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
185 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
186 characters in header names, implemented as a verify condition.
187 Contributed by Michael Fischer v. Mollard.
189 TL/05 Rename SPF condition results err_perm and err_temp to standardized
190 results permerror and temperror. Previous values are deprecated but
191 still accepted. In a future release, err_perm and err_temp will be
192 completely removed, which will be a backward incompatibility if the
193 ACL tests for either of these two old results. Patch contributed by
194 user bes-internal on the mailing list.
196 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
198 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
199 selectors, in both main and reject logs.
201 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
204 JH/07 Add malware type "sock" for talking to simple daemon.
206 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
208 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
209 routers/transports under cutthrough routing.
211 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
212 numbers. Touch up "bool" conditional to keep the same definition.
214 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
216 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
218 JH/12 Expand items in router/transport headers_add or headers_remove lists
219 individually rather than the list as a whole. Bug 1452.
221 Required for reasonable handling of multiple headers_ options when
222 they may be empty; requires that headers_remove items with embedded
223 colons must have them doubled (or the list-separator changed).
225 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
226 view the policy declared in the DMARC record. Currently, $dmarc_status
227 is a combined value of both the record presence and the result of the
230 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
232 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
233 dnslookup router and the smtp transport (applying to the forward
236 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
237 of ldap servers used for a specific lookup. Patch provided by Heiko
240 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
241 New variable $lookup_dnssec_authenticated for observability.
243 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
244 Patch submitted by Lars Timman.
246 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
248 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
249 Requires trusted mode and valid format message id, aborts otherwise.
250 Patch contributed by Heiko Schlichting.
252 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
253 certextract with support for various fields. Bug 1358.
255 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
256 is requested by default, modifiable by smtp transport option
259 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
260 operate on certificate variables to give certificate fingerprints
261 Also new ${sha256:cert_variable}.
263 JH/23 The PRDR feature is moved from being Experimental into the mainline.
265 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
266 Christian Aistleitner.
268 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
270 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
271 file. Patch from Wolfgang Breyha.
273 JH/25 Expand the coverage of the delivery $host and $host_address to
274 client authenticators run in verify callout. Bug 1476.
276 JH/26 Port service names are now accepted for tls_on_connect_ports, to
277 align with daemon_smtp_ports. Bug 72.
279 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
280 support and error reporting did not work properly.
282 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
283 and is readable. Patch from Andrew Colin Kissa.
285 TL/14 Enhance documentation of ${run expansion and how it parses the
286 commandline after expansion, particularly in the case when an
287 unquoted variable expansion results in an empty value.
289 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
291 PP/02 Fix internal collision of T_APL on systems which support RFC3123
292 by renaming away from it. Addresses GH issue 15, reported by
295 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
297 TL/15 SECURITY: prevent double expansion in math comparison functions
298 (can expand unsanitized data). Not remotely exploitable.
305 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
307 PP/02 Make -n do something, by making it not do something.
308 When combined with -bP, the name of an option is not output.
310 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
313 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
314 $sender_host_name and config options to manage this, and basic check
317 PP/05 DSCP support for outbound connections and control modifier for inbound.
319 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
320 (Only plugin which currently uses this is kerberos4, which nobody should
321 be using, but we should make it available and other future plugins might
322 conceivably use it, even though it would break NAT; stuff *should* be
323 using channel bindings instead).
325 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
326 name; added for Sendmail compatibility; requires admin caller.
327 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
328 just ignore it); requires trusted caller.
329 Also parse but ignore: -Ac -Am -X<logfile>
332 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
334 TL/02 Add +smtp_confirmation as a default logging option.
336 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
337 Patch by Magnus Holmgren from 2007-02-20.
339 TL/04 Bugzilla 1281 - Spec typo.
340 Bugzilla 1283 - Spec typo.
341 Bugzilla 1290 - Spec grammar fixes.
343 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
345 TL/06 Add Experimental DMARC support using libopendmarc libraries.
347 TL/07 Fix an out of order global option causing a segfault. Reported to dev
348 mailing list by by Dmitry Isaikin.
350 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
352 JH/02 Support "G" suffix to numbers in ${if comparisons.
354 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
356 NM/01 Bugzilla 1197 - Spec typo
357 Bugzilla 1196 - Spec examples corrections
359 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
361 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
362 gnutls_enable_pkcs11, but renamed to more accurately indicate its
365 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
366 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
368 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
369 "acl {{name}{arg}...}", and optional args on acl condition
372 JH/05 Permit multiple router/transport headers_add/remove lines.
374 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
376 JH/07 Avoid using a waiting database for a single-message-only transport.
377 Performance patch from Paul Fisher. Bugzilla 1262.
379 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
382 JH/09 Add $headers_added variable, with content from use of ACL modifier
383 add_header (but not yet added to the message). Bugzilla 199.
385 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
386 Pulled from Bugzilla 817 by Wolfgang Breyha.
388 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
390 (nb: this is the same fix as in Exim 4.80.1)
392 JH/11 Add A= logging on delivery lines, and a client_set_id option on
395 JH/12 Add optional authenticated_sender logging to A= and a log_selector
398 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
400 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
401 advertise SMTP AUTH mechanism to us, instead of a generic
402 protocol violation error. Also, make Exim more robust to bad
403 data from the Dovecot auth socket.
405 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
407 When a queue runner is handling a message, Exim first routes the
408 recipient addresses, during which it prunes them based on the retry
409 hints database. After that it attempts to deliver the message to
410 any remaining recipients. It then updates the hints database using
413 So if a recipient address works intermittently, it can get repeatedly
414 deferred at routing time. The retry hints record remains fresh so the
415 address never reaches the final cutoff time.
417 This is a fairly common occurrence when a user is bumping up against
418 their storage quota. Exim had some logic in its local delivery code
419 to deal with this. However it did not apply to per-recipient defers
420 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
422 This change adds a proper retry rule check during routing so that the
423 final cutoff time is checked against the message's age. We only do
424 this check if there is an address retry record and there is not a
425 domain retry record; this implies that previous attempts to handle
426 the address had the retry_use_local_parts option turned on. We use
427 this as an approximation for the destination being like a local
428 delivery, as in LMTP.
430 I suspect this new check makes the old local delivery cutoff check
431 redundant, but I have not verified this so I left the code in place.
433 TF/02 Correct gecos expansion when From: is a prefix of the username.
435 Test 0254 submits a message to Exim with the header
439 When I ran the test suite under the user fanf2, Exim expanded
440 the header to contain my full name, whereas it should have added
441 a Resent-Sender: header. It erroneously treats any prefix of the
442 username as equal to the username.
444 This change corrects that bug.
446 GF/01 DCC debug and logging tidyup
447 Error conditions log to paniclog rather than rejectlog.
448 Debug lines prefixed by "DCC: " to remove any ambiguity.
450 TF/03 Avoid unnecessary rebuilds of lookup-related code.
452 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
453 Bug spotted by Jeremy Harris; was flawed since initial commit.
454 Would have resulted in OCSP responses post-SNI triggering an Exim
455 NULL dereference and crash.
457 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
459 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
460 Bug detection, analysis and fix by Samuel Thibault.
461 Bugzilla 1331, Debian bug #698092.
463 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
465 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
466 Server implementation by Todd Lyons, client by JH.
467 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
468 config variable "prdr_enable" controls whether the server
469 advertises the facility. If the client requests PRDR a new
470 acl_data_smtp_prdr ACL is called once for each recipient, after
471 the body content is received and before the acl_smtp_data ACL.
472 The client is controlled by bolth of: a hosts_try_prdr option
473 on the smtp transport, and the server advertisement.
474 Default client logging of deliveries and rejections involving
475 PRDR are flagged with the string "PRDR".
477 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
478 fclose(). Diagnosis by Todd Lyons.
480 PP/17 Update configure.default to handle IPv6 localhost better.
481 Patch by Alain Williams (plus minor tweaks).
484 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
485 This is now consistent with GnuTLS, and is now documented: the
486 previous undocumented portable approach to treating the option as
487 unset was to force an expansion failure. That still works, and
488 an empty string is now equivalent.
490 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
491 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
492 not performing validation itself.
494 PP/20 Added force_command boolean option to pipe transport.
495 Patch from Nick Koston, of cPanel Inc.
497 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
500 TF/04 Added udpsend ACL modifer and hexquote expansion operator
502 PP/21 Fix eximon continuous updating with timestamped log-files.
503 Broken in a format-string cleanup in 4.80, missed when I repaired the
504 other false fix of the same issue.
505 Report and fix from Heiko Schlichting.
508 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
509 Report from Prashanth Katuri.
511 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
512 It's SecureTransport, so affects any MacOS clients which use the
513 system-integrated TLS libraries, including email clients.
515 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
516 using a MIME ACL for non-SMTP local injection.
517 Report and assistance in diagnosis by Warren Baker.
519 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
521 JH/16 Fix comparisons for 64b. Bugzilla 1385.
523 TL/09 Add expansion variable $authenticated_fail_id to keep track of
524 last id that failed so it may be referenced in subsequent ACL's.
526 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
529 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
530 ldap library initialization, allowing self-signed CA's to be
531 used. Also properly sets require_cert option later in code by
532 using NULL (global ldap config) instead of ldap handle (per
533 session). Bug diagnosis and testing by alxgomz.
535 TL/12 Enhanced documentation in the ratelimit.pl script provided in
536 the src/util/ subdirectory.
538 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
539 renamed to Transport Post Delivery Action by Jeremy Harris, as
542 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
543 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
544 redis_servers = needs to be configured which will be used by the redis
545 lookup. Patch from Warren Baker, of The Packet Hub.
547 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
549 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
550 hostname or reverse DNS when processing a host list. Used suggestions
551 from multiple comments on this bug.
553 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
555 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
556 Missed a few lines, added it to make the runtest require no keyboard
559 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
560 contains upper case chars. Make router use caseful_local_part.
562 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
563 support when GnuTLS has been built with p11-kit.
569 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
571 This, or similar/improved, will also be change PP/11 of 4.82.
577 PP/01 Handle short writes when writing local log-files.
578 In practice, only affects FreeBSD (8 onwards).
579 Bugzilla 1053, with thanks to Dmitry Isaikin.
581 NM/01 Bugzilla 949 - Documentation tweak
583 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
586 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
588 PP/02 Implemented gsasl authenticator.
590 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
592 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
593 `pkg-config foo` for cflags/libs.
595 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
596 with rest of GSASL and with heimdal_gssapi.
598 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
599 `pkg-config foo` for cflags/libs for the TLS implementation.
601 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
602 properties get this fed in as external SSF. A number of robustness
603 and debugging improvements to the cyrus_sasl authenticator.
605 PP/08 cyrus_sasl server now expands the server_realm option.
607 PP/09 Bugzilla 1214 - Log authentication information in reject log.
608 Patch by Jeremy Harris.
610 PP/10 Added dbmjz lookup type.
612 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
614 PP/12 MAIL args handles TAB as well as SP, for better interop with
615 non-compliant senders.
616 Analysis and variant patch by Todd Lyons.
618 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
619 Bug report from Lars Müller <lars@samba.org> (via SUSE),
620 Patch from Dirk Mueller <dmueller@suse.com>
622 PP/13 tls_peerdn now print-escaped for spool files.
623 Observed some $tls_peerdn in wild which contained \n, which resulted
624 in spool file corruption.
626 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
627 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
628 or write after TLS renegotiation, which otherwise led to messages
631 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
632 as a tracking header (ie: a signed header comes before the signature).
633 Patch from Wolfgang Breyha.
635 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
636 comma-sep list; embedded commas doubled.
638 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
640 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
642 Report and patch from Dmitry Banschikov.
644 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
645 Removed SSL_clear() after SSL_new() which led to protocol negotiation
646 failures. We appear to now support TLS1.1+ with Exim.
648 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
649 lets Exim select keys and certificates based upon TLS SNI from client.
650 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
651 before an outbound SMTP session. New log_selector, +tls_sni.
653 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
654 NULL dereference. Report and patch from Alun Jones.
656 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
657 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
658 Not seeing resolver debug output on NetBSD, but suspect this is a
659 resolver implementation change.
661 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
662 Left warnings. Added "eximon gdb" invocation mode.
664 PP/21 Defaulting "accept_8bitmime" to true, not false.
666 PP/22 Added -bw for inetd wait mode support.
668 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
669 locate the relevant includes and libraries. Made this the default.
671 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
672 Bugzilla 1246, report and most of solution from Tomasz Kusy.
674 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
675 This may cause build issues on older platforms.
677 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
678 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
679 gnutls_require_mac & gnutls_require_protocols (no longer supported).
680 Added SNI support via GnuTLS too.
681 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
683 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
685 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
686 Applied second patch from Janne, implementing suggestion to default
687 multiple-strings-in-record handling to match SPF spec.
689 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
691 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
692 read-only, out of scope).
693 Patch from Wolfgang Breyha, report from Stuart Northfield.
695 PP/29 Fix three issues highlighted by clang analyser static analysis.
696 Only crash-plausible issue would require the Cambridge-specific
697 iplookup router and a misconfiguration.
698 Report from Marcin Mirosław.
700 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
702 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
703 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
704 As part of this, removing so much warning spew let me fix some minor
705 real issues in debug logging.
707 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
708 assignment on my part. Fixed.
710 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
711 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
712 Janne Snabb (who went above and beyond: thank you).
714 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
715 string otherwise requires a connection and a bunch more work and it's
716 relatively easy to get wrong. Should also expose TLS library linkage
719 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
720 64-bit ${eval} (JH/03).
722 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
723 GNU libc to support some of the 64-bit stuff, should not lead to
724 conflicts. Defined before os.h is pulled in, so if a given platform
725 needs to override this, it can.
727 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
728 protection layer was required, which is not implemented.
729 Bugzilla 1254, patch from Wolfgang Breyha.
731 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
732 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
733 tls_dhparam take prime identifiers. Also unbreak combination of
734 OpenSSL+DH_params+TLSSNI.
736 PP/39 Disable SSLv2 by default in OpenSSL support.
742 PP/01 Solaris build fix for Oracle's LDAP libraries.
743 Bugzilla 1109, patch from Stephen Usher.
745 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
747 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
750 TF/02 Fix a couple more cases where we did not log the error message
751 when unlink() failed. See also change 4.74-TF/03.
753 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
754 lock up or crash if it happened to be inside a call to libc when it
755 got a SIGUSR1 from exiwhat.
757 The SIGUSR1 handler appends the current process status to the process
758 log which is later printed by exiwhat. It used to use the general
759 purpose logging code to do this, but several functions it calls are
760 not safe for signals.
762 The new output code in the SIGUSR1 handler is specific to the process
763 log, and simple enough that it's easy to inspect for signal safety.
764 Removing some special cases also simplifies the general logging code.
765 Removing the spurious timestamps from the process log simplifies
768 TF/04 Improved ratelimit ACL condition.
770 The /noupdate option has been deprecated in favour of /readonly which
771 has clearer semantics. The /leaky, /strict, and /readonly update modes
772 are mutually exclusive. The update mode is no longer included in the
773 database key; it just determines when the database is updated. (This
774 means that when you upgrde Exim will forget old rate measurements.)
776 Exim now checks that the per_* options are used with an update mode that
777 makes sense for the current ACL. For example, when Exim is processing a
778 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
779 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
780 must specify per_mail/readonly. If you omit the update mode it defaults to
781 /leaky where that makes sense (as before) or /readonly where required.
783 The /noupdate option is now undocumented but still supported for
784 backwards compatibility. It is equivalent to /readonly except that in
785 ACLs where /readonly is required you may specify /leaky/noupdate or
786 /strict/noupdate which are treated the same as /readonly.
788 A useful new feature is the /count= option. This is a generalization
789 of the per_byte option, so that you can measure the throughput of other
790 aggregate values. For example, the per_byte option is now equivalent
791 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
793 The per_rcpt option has been generalized using the /count= mechanism
794 (though it's more complicated than the per_byte equivalence). When it is
795 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
796 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
797 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
798 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
799 Note that using per_rcpt with a non-readonly update mode in more than
800 one ACL will cause the recipients to be double-counted. (The per_mail
801 and per_byte options don't have this problem.)
803 The handling of very low rates has changed slightly. If the computed rate
804 is less than the event's count (usually one) then this event is the first
805 after a long gap. In this case the rate is set to the same as this event's
806 count, so that the first message of a spam run is counted properly.
808 The major new feature is a mechanism for counting the rate of unique
809 events. The new per_addr option counts the number of different
810 recipients that someone has sent messages to in the last time period. It
811 behaves like per_rcpt if all the recipient addresses are different, but
812 duplicate recipient addresses do not increase the measured rate. Like
813 the /count= option this is a general mechanism, so the per_addr option
814 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
815 example, measure the rate that a client uses different sender addresses
816 with the options per_mail/unique=$sender_address. There are further
817 details in the main documentation.
819 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
821 TF/06 Removed a few PCRE remnants.
823 TF/07 Automatically extract Exim's version number from tags in the git
824 repository when doing development or release builds.
826 PP/02 Raise smtp_cmd_buffer_size to 16kB.
827 Bugzilla 879. Patch from Paul Fisher.
829 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
830 Heavily based on revision 40f9a89a from Simon Arlott's tree.
833 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
835 PP/05 Variable $av_failed, true if the AV scanner deferred.
836 Bugzilla 1078. Patch from John Horne.
838 PP/06 Stop make process more reliably on build failure.
839 Bugzilla 1087. Patch from Heiko Schlittermann.
841 PP/07 Make maildir_use_size_file an _expandable_ boolean.
842 Bugzilla 1089. Patch from Heiko Schlittermann.
844 PP/08 Handle ${run} returning more data than OS pipe buffer size.
845 Bugzilla 1131. Patch from Holger Weiß.
847 PP/09 Handle IPv6 addresses with SPF.
848 Bugzilla 860. Patch from Wolfgang Breyha.
850 PP/10 GnuTLS: support TLS 1.2 & 1.1.
852 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
855 PP/11 match_* no longer expand right-hand-side by default.
856 New compile-time build option, EXPAND_LISTMATCH_RHS.
857 New expansion conditions, "inlist", "inlisti".
859 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
861 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
863 PP/14 fix log_write() format string regression from TF/03.
864 Bugzilla 1152. Patch from Dmitry Isaikin.
870 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
872 PP/02 Harmonised TLS library version reporting; only show if debugging.
873 Layout now matches that introduced for other libraries in 4.74 PP/03.
875 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
877 PP/04 New "dns_use_edns0" global option.
879 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
882 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
883 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
885 TK/01 Updated PolarSSL code to 0.14.2.
886 Bugzilla 1097. Patch from Andreas Metzler.
888 PP/07 Catch divide-by-zero in ${eval:...}.
891 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
894 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
895 format-string attack -- SECURITY: remote arbitrary code execution.
897 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
898 time unintentionally subject to list matching rules, letting the header
899 cause arbitrary Exim lookups (of items which can occur in lists, *not*
900 arbitrary string expansion). This allowed for information disclosure.
902 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
903 INT_MIN/-1 -- value coerced to INT_MAX.
909 NM/01 Workround for PCRE version dependency in version reporting
912 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
913 This fixes portability to compilers other than gcc, notably
914 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
916 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
917 makefiles for portability to HP-UX and POSIX correctness.
919 PP/01 Permit LOOKUP_foo enabling on the make command-line.
920 Also via indented variable definition in the Makefile.
921 (Debugging by Oliver Heesakkers).
923 PP/02 Restore caching of spamd results with expanded spamd_address.
924 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
926 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
927 Improves build reliability. Fix from: Frank Elsner
929 NM/02 Fix wide character breakage in the rfc2047 coding
930 Fixes bug 1064. Patch from Andrey N. Oktyabrski
932 NM/03 Allow underscore in dnslist lookups
933 Fixes bug 1026. Patch from Graeme Fowler
935 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
936 Code patches from Adam Ciarcinski of NetBSD.
938 NM/04 Fixed exiqgrep to cope with mailq missing size issue
941 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
942 is logged, to avoid truncation. Patch from John Horne.
944 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
945 Patch from Jakob Hirsch.
947 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
948 SQL string expansion failure details.
949 Patch from Andrey Oktyabrski.
951 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
952 Patch from Simon Arlott.
954 PP/09 New lookups functionality failed to compile on old gcc which rejects
955 extern declarations in function scope.
956 Patch from Oliver Fleischmann
958 PP/10 Use sig_atomic_t for flags set from signal handlers.
959 Check getgroups() return and improve debugging.
960 Fixed developed for diagnosis in bug 927 (which turned out to be
963 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
964 Patch from Mark Zealey.
966 PP/12 Bugzilla 1056: Improved spamd server selection.
967 Patch from Mark Zealey.
969 PP/13 Bugzilla 1086: Deal with maildir quota file races.
970 Based on patch from Heiko Schlittermann.
972 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
973 Patch from Uwe Doering, sign-off by Michael Haardt.
975 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
976 variable declaration deep within a block. Bug and patch from
979 PP/15 lookups-Makefile IRIX compatibilty coercion.
981 PP/16 Make DISABLE_DKIM build knob functional.
983 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
984 Patch by Simon Arlott
986 TF/03 Fix valgrind.h portability to C89 compilers that do not support
987 variable argument macros. Our copy now differs from upstream.
993 TF/01 Failure to get a lock on a hints database can have serious
994 consequences so log it to the panic log.
996 TF/02 Log LMTP confirmation messages in the same way as SMTP,
997 controlled using the smtp_confirmation log selector.
999 TF/03 Include the error message when we fail to unlink a spool file.
1001 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1002 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1003 for maintaining out-of-tree patches for some time.
1005 PP/01 Bugzilla 139: Documentation and portability issues.
1006 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1007 Handle per-OS dynamic-module compilation flags.
1009 PP/02 Let /dev/null have normal permissions.
1010 The 4.73 fixes were a little too stringent and complained about the
1011 permissions on /dev/null. Exempt it from some checks.
1012 Reported by Andreas M. Kirchwitz.
1014 PP/03 Report version information for many libraries, including
1015 Exim version information for dynamically loaded libraries. Created
1016 version.h, now support a version extension string for distributors
1017 who patch heavily. Dynamic module ABI change.
1019 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1020 privilege escalation vulnerability whereby the Exim run-time user
1021 can cause root to append content of the attacker's choosing to
1024 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1027 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1028 If dropping privileges for untrusted macros, we disabled normal logging
1029 on the basis that it would fail; for the Exim run-time user, this is not
1030 the case, and it resulted in successful deliveries going unlogged.
1031 Fixed. Reported by Andreas Metzler.
1037 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1038 only prepend for the Resent-* case. Fixes regression introduced in
1039 Exim 4.70 by NM/22 for Bugzilla 607.
1041 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1042 increasing numbers of administrators be bitten by this.
1044 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1046 PP/03 Bugzilla 994: added openssl_options main configuration option.
1048 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1050 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1052 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1054 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1055 without a peer certificate, leading to a segfault because of an
1056 assumption that peers always have certificates. Be a little more
1057 paranoid. Problem reported by Martin Tscholak.
1059 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1060 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1061 NB: ClamAV planning to remove STREAM in "middle of 2010".
1062 CL also introduces -bmalware, various -d+acl logging additions and
1063 more caution in buffer sizes.
1065 PP/09 Implemented reverse_ip expansion operator.
1067 PP/10 Bugzilla 937: provide a "debug" ACL control.
1069 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1071 PP/12 Bugzilla 973: Implement --version.
1073 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1075 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1077 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1079 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1080 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1081 ignore trailing whitespace.
1083 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1085 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1088 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1089 Notification from Dr Andrew Aitchison.
1091 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1092 ExtendedDetectionInfo response format.
1093 Notification from John Horne.
1095 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1098 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1099 XSL and documented dependency on system catalogs, with examples of how
1102 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1105 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1106 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1107 configuration file which is writeable by the Exim user or group.
1109 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1110 of configuration files to cover files specified with the -C option if
1111 they are going to be used with root privileges, not just the default
1114 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1115 option (effectively making it always true).
1117 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1118 files to be used while preserving root privileges.
1120 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1121 that rogue child processes cannot use them.
1123 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1124 run-time user, instead of root.
1126 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1127 Exim run-time user without dropping privileges.
1129 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1130 result string, instead of calling string_vformat() twice with the same
1133 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1134 for other users. Others should always drop root privileges if they use
1135 -C on the command line, even for a whitelisted configure file.
1137 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1139 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1145 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1146 $data_path, and $header_path variables; fixed documentation bugs and
1149 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1150 exipick to access non-standard spools, including the "frozen" queue
1153 NM/01 Bugzilla 965: Support mysql stored procedures.
1154 Patch from Alain Williams
1156 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1158 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1159 Patch from Andreas Metzler
1161 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1162 Patch from Kirill Miazine
1164 NM/05 Bugzilla 671: Added umask to procmail example.
1166 JJ/03 installed exipick 20100323.0, fixing doc bug
1168 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1169 directory. Notification and patch from Dan Rosenberg.
1171 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1173 TK/02 Improve log output when DKIM signing operation fails.
1175 MH/01 Treat the transport option dkim_domain as a colon separated
1176 list, not as a single string, and sign the message with each element,
1177 omitting multiple occurences of the same signer.
1179 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1180 Bugzilla 985, 986. Patch by Simon Arlott
1182 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1183 Patch by Simon Arlott
1185 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1186 MBX locking. Notification from Dan Rosenberg.
1192 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1194 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1196 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1198 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1200 NM/04 Bugzilla 915: Flag broken perl installation during build.
1206 TK/01 Added patch by Johannes Berg that expands the main option
1207 "spamd_address" if it starts with a dollar sign.
1209 TK/02 Write list of recipients to X-Envelope-Sender header when building
1210 the mbox-format spool file for content scanning (suggested by Jakob
1213 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1214 (http://www.dcc-servers.net/) support via dccifd. Activated by
1215 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1217 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1218 by Mark Daniel Reidel <mr@df.eu>.
1220 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1221 When building exim an external PCRE library is now needed -
1222 PCRE is a system library on the majority of modern systems.
1223 See entry on PCRE_LIBS in EDITME file.
1225 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1226 conversation. Added nologin parameter to request.
1227 Patch contributed by Kirill Miazine.
1229 TF/01 Do not log submission mode rewrites if they do not change the address.
1231 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1233 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1234 log files in place. Contributed by Roberto Lima.
1236 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1238 TF/03 Bugzilla 615: When checking the local_parts router precondition
1239 after a local_part_suffix or local_part_prefix option, Exim now
1240 does not use the address's named list lookup cache, since this
1241 contains cached lookups for the whole local part.
1243 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1244 Robert Millan. Documentation is in experimental-spec.txt.
1246 TF/04 Bugzilla 668: Fix parallel build (make -j).
1248 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1250 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1251 Patch provided by Jan Srzednicki.
1253 TF/05 Leading white space used to be stripped from $spam_report which
1254 wrecked the formatting. Now it is preserved.
1256 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1257 that they are available at delivery time.
1259 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1261 TF/08 TLS error reporting now respects the incoming_interface and
1262 incoming_port log selectors.
1264 TF/09 Produce a more useful error message if an SMTP transport's hosts
1265 setting expands to an empty string.
1267 NM/06 Bugzilla 744: EXPN did not work under TLS.
1268 Patch provided by Phil Pennock.
1270 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1271 Patch provided by Richard Godbee.
1273 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1274 acl_smtp_notquit, added index entry.
1276 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1277 Patch provided by Eugene Bujak.
1279 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1280 accept(). Patch provided by Maxim Dounin.
1282 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1283 Patch provided by Phil Pennock.
1285 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1287 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1288 Patch provided by Brad "anomie" Jorsch.
1290 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1291 Patch provided by Dean Brooks.
1293 TK/05 Add native DKIM support (does not depend on external libraries).
1295 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1296 Patch provided by Graeme Fowler.
1298 NM/16 Bugzilla 851: Documentation example syntax fix.
1300 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1302 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1305 NM/19 Bugzilla 745: TLS version reporting.
1306 Patch provided by Phil Pennock.
1308 NM/20 Bugzilla 167: bool: condition support.
1309 Patch provided by Phil Pennock.
1311 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1312 clients. Patch provided by Phil Pennock.
1314 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1315 Patch provided by Brad "anomie" Jorsch.
1317 NM/23 Bugzilla 687: Fix misparses in eximstats.
1318 Patch provided by Heiko Schlittermann.
1320 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1321 Patch provided by Heiko Schlittermann.
1323 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1324 plus update to original patch.
1326 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1328 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1329 Patch provided by David Brownlee.
1331 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1333 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1335 NM/30 Bugzilla 888: TLS documentation bugfixes.
1337 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1339 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1340 Unlike the original bugzilla I have changed all shell scripts in src tree.
1342 NM/33 Bugzilla 898: Transport filter timeout fix.
1343 Patch by Todd Rinaldo.
1345 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1346 Patch by Serge Demonchaux.
1348 NM/35 Bugzilla 39: Base64 decode bug fixes.
1349 Patch by Jakob Hirsch.
1351 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1353 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1355 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1357 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1363 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1364 ALT-N's libdkim that I have put here:
1365 http://duncanthrax.net/exim-experimental/
1367 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1368 were called 'true' and it seems that C99 defines that as a reserved
1369 keyword to be used with 'bool' variable types. That means you could
1370 not include C99-style headers which use bools without triggering
1371 build errors in sieve.c.
1373 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1374 as mailq or other aliases. Changed the --help handling significantly
1375 to do whats expected. exim_usage() emits usage/help information.
1377 SC/01 Added the -bylocaldomain option to eximstats.
1379 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1381 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1383 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1389 PH/01 Another patch from the Sieve maintainer.
1391 PH/02 When an IPv6 address is converted to a string for single-key lookup
1392 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1393 dots are used instead of colons so that keys in lsearch files need not
1394 contain colons. This was done some time before quoting was made available
1395 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1396 (notated using the quote facility) so as to distinguish them from IPv4
1397 keys. This meant that lookups for IP addresses in host lists did not work
1398 for iplsearch lookups.
1400 This has been fixed by arranging for IPv6 addresses to be expressed with
1401 colons if the lookup type is iplsearch. This is not incompatible, because
1402 previously such lookups could never work.
1404 The situation is now rather anomolous, since one *can* have colons in
1405 ordinary lsearch keys. However, making the change in all cases is
1406 incompatible and would probably break a number of configurations.
1408 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1411 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1412 conversion specification without a maximum field width, thereby enabling
1413 a rogue spamd server to cause a buffer overflow. While nobody in their
1414 right mind would setup Exim to query an untrusted spamd server, an
1415 attacker that gains access to a server running spamd could potentially
1416 exploit this vulnerability to run arbitrary code as the Exim user.
1418 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1419 $primary_hostname instead of what libspf2 thinks the hosts name is.
1421 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1422 a directory entry by the name of the lookup key. Previously, if a
1423 symlink pointed to a non-existing file or a file in a directory that
1424 Exim lacked permissions to read, a lookup for a key matching that
1425 symlink would fail. Now it is enough that a matching directory entry
1426 exists, symlink or not. (Bugzilla 503.)
1428 PH/03 The body_linecount and body_zerocount variables are now exported in the
1431 PH/04 Added the $dnslist_matched variable.
1433 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1434 This means they are set thereafter only if the connection becomes
1437 PH/06 Added the client_condition to authenticators so that some can be skipped
1438 by clients under certain conditions.
1440 PH/07 The error message for a badly-placed control=no_multiline_responses left
1441 "_responses" off the end of the name.
1443 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1445 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1446 (without spaces) instead of just copying the configuration text.
1448 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1450 PH/11 Added $max_received_linelength.
1452 PH/12 Added +ignore_defer and +include_defer to host lists.
1454 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1455 way in which PCRE > 7.0 is built.
1457 PH/14 Implemented queue_only_load_latch.
1459 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1460 MAIL command. The effect was to mangle the value on 64-bit systems.
1462 PH/16 Another patch from the Sieve maintainer.
1464 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1466 PH/18 If a system quota error occurred while trying to create the file for
1467 a maildir delivery, the message "Mailbox is full" was not appended to the
1468 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1469 only to a quota excession during the actual writing of the file.
1471 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1472 characters?) which causes problems in log lines. The DN values are now
1473 passed through string_printing() before being added to log lines.
1475 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1476 and InterBase are left for another time.)
1478 PH/21 Added message_body_newlines option.
1480 PH/22 Guard against possible overflow in moan_check_errorcopy().
1482 PH/23 POSIX allows open() to be a macro; guard against that.
1484 PH/24 If the recipient of an error message contained an @ in the local part
1485 (suitably quoted, of course), incorrect values were put in $domain and
1486 $local_part during the evaluation of errors_copy.
1492 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1493 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1496 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1497 issue a MAIL command.
1499 PH/02 In an ACL statement such as
1501 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1503 if a client was not listed at all, or was listed with a value other than
1504 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1505 the condition was not true (as it should be), so access was not denied.
1506 The bug was that the ! inversion was incorrectly passed on to the second
1507 item. This has been fixed.
1509 PH/03 Added additional dnslists conditions == and =& which are different from
1510 = and & when the dns lookup returns more than one IP address.
1512 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1513 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1515 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1516 FSYNC, which compiles an option called disable_fsync that allows for
1517 bypassing fsync(). The documentation is heavily laced with warnings.
1519 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1521 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1522 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1523 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1524 including adding "make clean"; (3) Added -fPIC when compiling the test
1525 dynamically loaded module, to get rid of a warning.
1527 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1528 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1529 = 0s. The bug is otherwise harmless.
1531 PH/07 There was a bug in the dovecot authenticator such that the value of
1532 $auth1 could be overwritten, and so not correctly preserved, after a
1533 successful authentication. This usually meant that the value preserved by
1534 the server_setid option was incorrect.
1536 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1538 PH/09 Installed PCRE release 7.0.
1540 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1541 run for batched SMTP input. It is now run at the start of every message
1542 in the batch. While fixing this I discovered that the process information
1543 (output by running exiwhat) was not always getting set for -bs and -bS
1544 input. This is fixed, and it now also says "batched" for BSMTP.
1546 PH/11 Added control=no_pipelining.
1548 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1549 patch, slightly modified), and move the expansion of helo_data till after
1550 the connection is made in the smtp transport (so it can use these
1553 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1555 PH/14 Added log_selector = +pid.
1557 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1559 PH/16 Add ${if forany and ${if forall.
1561 PH/17 Added dsn_from option to vary the From: line in DSNs.
1563 PH/18 Flush SMTP output before performing a callout, unless control =
1564 no_callout_flush is set.
1566 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1567 was true (the default) a successful delivery failed to delete the retry
1568 item, thus causing premature timeout of the address. The bug is now
1571 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1573 PH/21 Long custom messages for fakedefer and fakereject are now split up
1574 into multiline reponses in the same way that messages for "deny" and
1575 other ACL rejections are.
1577 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1578 with slight modification.
1580 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1581 draft, changing the syntax and factoring some duplicate code".
1583 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1584 for deliveries of the second and subsequent messages over the same SMTP
1587 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1588 ${reduce, with only minor "tidies".
1590 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1592 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1593 expansion side effects.
1595 PH/27 When a message times out after an over-quota error from an Exim-imposed
1596 quota, the bounce message says "mailbox is full". This message was not
1597 being given when it was a system quota that was exceeded. It now should
1600 MH/03 Made $recipients available in local_scan(). local_scan() already has
1601 better access to the recipient list through recipients_list[], but
1602 $recipients can be useful in postmaster-provided expansion strings.
1604 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1605 in the case of a MAIL command with additional options following the
1606 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1607 were accidentally chopped off.
1609 PH/29 SMTP synchronization checks are implemented when a command is read -
1610 there is a check that no more input is waiting when there shouldn't be
1611 any. However, for some commands, a delay in an ACL can mean that it is
1612 some time before the response is written. In this time, more input might
1613 arrive, invalidly. So now there are extra checks after an ACL has run for
1614 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1615 pipelining has not been advertised.
1617 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1619 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1620 returned in the SMTP response when smtp_return_error_details was false.
1621 This has been fixed.
1623 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1624 instead of the C I/O that was originally supplied, because problems were
1625 reported on Solaris.
1627 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1628 Exim which did not show up earlier: it was assuming that a call to
1629 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1630 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1631 was a macro that became an assignment, so it seemed to work. This has
1632 changed to a proper function call with a void return, hence the compile
1633 error. Exim's code has been fixed.
1635 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1638 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1640 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1642 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1643 because it turns out to be too big or there is a timeout) there is no
1644 "Completed" line in the log. When some messages of this type were
1645 selected by exigrep, they were listed as "not completed". Others were
1646 picked up by some special patterns. I have improved the selection
1647 criteria to be more general.
1649 PH/38 The host_find_failed option in the manualroute router can now be set
1650 to "ignore", to completely ignore a host whose IP address cannot be
1651 found. If all hosts are ignored, the behaviour is controlled by the new
1652 host_all_ignored option.
1654 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1655 homing or because of multiple MX records with /mx) generated more than
1656 one IP address, and the following item turned out to be the local host,
1657 all the secondary addresses of the first item were incorrectly removed
1658 from the list, along with the local host and any following hosts (which
1659 is what is supposed to happen).
1661 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1662 whoever called Exim into the -H file. In the case of the daemon it was
1663 behaving confusingly. When first started, it used values for whoever
1664 started the daemon, but after a SIGHUP it used the Exim user (because it
1665 calls itself on a restart). I have changed the code so that it now always
1668 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1669 message are rejected with the same error (e.g. no authentication or bad
1670 sender address), and a DATA command is nevertheless sent (as can happen
1671 with PIPELINING or a stupid MUA), the error message that was given to the
1672 RCPT commands is included in the rejection of the DATA command. This is
1673 intended to be helpful for MUAs that show only the final error to their
1676 PH/42 Another patch from the Sieve maintainer.
1678 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1679 Eximstats - Fixed some broken HTML links and added missing column headers
1681 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1682 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1684 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1686 PH/43 Yet another patch from the Sieve maintainer.
1688 PH/44 I found a way to check for a TCP/IP connection going away before sending
1689 the response to the final '.' that terminates a message, but only in the
1690 case where the client has not sent further data following the '.'
1691 (unfortunately, this is allowed). However, in many cases there won't be
1692 any further data because there won't be any more messages to send. A call
1693 to select() can be used: if it shows that the input is "ready", there is
1694 either input waiting, or the socket has been closed. An attempt to read
1695 the next input character can distinguish the two cases. Previously, Exim
1696 would have sent an OK response which the client would never have see.
1697 This could lead to message repetition. This fix should cure that, at
1698 least in a lot of common cases.
1700 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1701 advertised in response to EHLO.
1707 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1708 fixed by 4.65/MH/01 (is this a record?) are fixed:
1710 (i) An empty string was always treated as zero by the numeric comparison
1711 operators. This behaviour has been restored.
1713 (ii) It is documented that the numeric comparison operators always treat
1714 their arguments as decimal numbers. This was broken in that numbers
1715 starting with 0 were being interpreted as octal.
1717 While fixing these problems I realized that there was another issue that
1718 hadn't been noticed. Values of message_size_limit (both the global option
1719 and the transport option) were treated as octal if they started with 0.
1720 The documentation was vague. These values are now always treated as
1721 decimal, and I will make that clear in the documentation.
1727 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1728 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1731 MH/01 Don't check that the operands of numeric comparison operators are
1732 integers when their expansion is in "skipping" mode (fixes bug
1733 introduced by 4.64-PH/07).
1735 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1736 child addresses, Exim now panics and dies. Previously, because the count
1737 is held in a short int, deliveries were likely to be lost. As such a
1738 large number of recipients for a single message is ridiculous
1739 (performance will be very, very poor), I have chosen to impose a limit
1740 rather than extend the field.
1746 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1747 leftover -K file (the existence of which was triggered by #402).
1748 While we were at it, introduced process PID as part of the -K
1749 filename. This should rule out race conditions when creating
1752 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1753 processing considerably. Previous code took too long for large mails,
1754 triggering a timeout which in turn triggers #401.
1756 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1757 in the DK code in transports.c. sendfile() is not really portable,
1758 hence the _LINUX specificness.
1760 TF/01 In the add_headers option to the mail command in an Exim filter,
1761 there was a bug that Exim would claim a syntax error in any
1762 header after the first one which had an odd number of characters
1765 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1766 callout verification, Exim cached a "reject" for the entire domain. This
1767 is correct for most verifications, but it is not correct for a recipient
1768 verification with use_sender or use_postmaster set, because in that case
1769 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1770 case of MAIL FROM:<> rejection from other early rejections (e.g.
1771 rejection of HELO). When verifying a recipient using a non-null MAIL
1772 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1773 Whatever the result of the callout, the value of the domain cache is
1774 left unchanged (for any other kind of callout, getting as far as trying
1775 RCPT means that the domain itself is ok).
1777 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1780 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1781 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1782 session, the daemon ignores ECONNECT errors and logs others; it now
1783 ignores EPIPE as well.
1785 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1786 (quoted-printable decoding).
1788 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1789 later the small subsequent patch to fix an introduced bug.
1791 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1793 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1795 PH/08 An error is now given if message_size_limit is specified negative.
1797 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1798 to be given (somewhat) arbitrary names.
1800 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1803 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1804 miscellaneous code fixes
1806 PH/10 Added the log_reject_target ACL modifier to specify where to log
1809 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1810 hostname. This is wrong, because it relates to the incoming message (and
1811 probably the interface on which it is arriving) and not to the outgoing
1812 callout (which could be using a different interface). This has been
1813 changed to use the value of the helo_data option from the smtp transport
1814 instead - this is what is used when a message is actually being sent. If
1815 there is no remote transport (possible with a router that sets up host
1816 addresses), $smtp_active_hostname is used.
1818 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1819 tweaks were necessary in order to get it to work (see also 21 below):
1820 (a) The code assumed that strncpy() returns a negative number on buffer
1821 overflow, which isn't the case. Replaced with Exim's string_format()
1823 (b) There were several signed/unsigned issues. I just did the minimum
1824 hacking in of casts. There is scope for a larger refactoring.
1825 (c) The code used strcasecmp() which is not a standard C function.
1826 Replaced with Exim's strcmpic() function.
1827 (d) The code set only $1; it now sets $auth1 as well.
1828 (e) A simple test gave the error "authentication client didn't specify
1829 service in request". It would seem that Dovecot has changed its
1830 interface. Fortunately there's a specification; I followed it and
1831 changed what the client sends and it appears to be working now.
1833 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1836 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1837 address A is aliased to B and C, where B exists and C does not. Without
1838 -v the output is "A verified" because verification stops after a
1839 successful redirection if more than one address is generated. However,
1840 with -v the child addresses are also verified. Exim was outputting "A
1841 failed to verify" and then showing the successful verification for C,
1842 with its parentage. It now outputs "B failed to verify", showing B's
1843 parentage before showing the successful verification of C.
1845 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1846 look up a TXT record in a specific list after matching in a combined
1849 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1850 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1851 they consult the DNS. I had assumed they would set it the way they
1852 wanted; and indeed my experiments on Linux seem to show that in some
1853 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1854 To be on the safe side, however, I have now made the interface to
1855 host_find_byname() similar to host_find_bydns(), with an argument
1856 containing the DNS resolver options. The host_find_byname() function now
1857 sets these options at its start, just as host_find_bydns() does. The smtp
1858 transport options dns_qualify_single and dns_search_parents are passed to
1859 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1860 of host_find_byname() use the default settings of RES_DEFNAMES
1861 (qualify_single) but not RES_DNSRCH (search_parents).
1863 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1864 spool_read_header() do less string testing, by means of a preliminary
1865 switch on the second character of optional "-foo" lines. (This is
1866 overdue, caused by the large number of possibilities that now exist.
1867 Originally there were few.) While I was there, I also converted the
1868 str(n)cmp tests so they don't re-test the leading "-" and the first
1869 character, in the hope this might squeeze out yet more improvement.
1871 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1872 flag allowing group syntax was set by the header_syntax check but not
1873 turned off, possible causing trouble later; (2) The flag was not being
1874 set at all for the header_verify test, causing "group"-style headers to
1875 be rejected. I have now set it in this case, and also caused header_
1876 verify to ignore an empty address taken from a group. While doing this, I
1877 came across some other cases where the code for allowing group syntax
1878 while scanning a header line wasn't quite right (mostly, not resetting
1879 the flag correctly in the right place). These bugs could have caused
1880 trouble for malformed header lines. I hope it is now all correct.
1882 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1883 with the "reply" argument non-NULL. The code, however (which originally
1884 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1885 but it didn't always do it. This confused somebody who was copying the
1886 code for some other use. I have removed all the tests.
1888 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1889 feature that was used to support insecure browsers during the U.S. crypto
1890 embargo. It requires special client support, and Exim is probably the
1891 only MTA that supported it -- and would never use it because real RSA is
1892 always available. This code has been removed, because it had the bad
1893 effect of slowing Exim down by computing (never used) parameters for the
1894 RSA_EXPORT functionality.
1896 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1897 authenticator to fail if there's a tab character in the incoming data
1898 (there should never be unless someone is messing about, as it's supposed
1899 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1900 if the connection is using TLS or if the remote IP is the same as the
1901 local IP, and the "valid-client-cert option" if a client certificate has
1904 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1905 authenticators. This can be used for authorization after authentication
1906 succeeds. (In the case of plaintext, it servers for both authentication
1909 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1910 if any retry times were supplied.
1912 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1913 connection, where there is no client IP address to check. In this
1914 situation, the verify now always succeeds.
1916 PH/25 Applied John Jetmore's -Mset patch.
1918 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1920 PH/27 In a string expansion for a processed (not raw) header when multiple
1921 headers of the same name were present, leading whitespace was being
1922 removed from all of them, but trailing whitespace was being removed only
1923 from the last one. Now trailing whitespace is removed from each header
1924 before concatenation. Completely empty headers in a concatenation (as
1925 before) are ignored.
1927 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1928 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1930 PH/29 [Removed. This was a change that I later backed out, and forgot to
1931 correct the ChangeLog entry (that I had efficiently created) before
1932 committing the later change.]
1934 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1935 address errors (4xx response to RCPT) over the same connection as other
1936 messages routed to the same hosts. Such deliveries are always "forced",
1937 so retry times are not inspected. This resulted in far too many retries
1938 for the affected addresses. The effect occurred only when there were more
1939 hosts than the hosts_max_try setting in the smtp transport when it had
1940 the 4xx errors. Those hosts that it had tried were not added to the list
1941 of hosts for which the message was waiting, so if all were tried, there
1942 was no problem. Two fixes have been applied:
1944 (i) If there are any address or message errors in an SMTP delivery, none
1945 of the hosts (tried or untried) are now added to the list of hosts
1946 for which the message is waiting, so the message should not be a
1947 candidate for sending over the same connection that was used for a
1948 successful delivery of some other message. This seems entirely
1949 reasonable: after all the message is NOT "waiting for some host".
1950 This is so "obvious" that I'm not sure why it wasn't done
1951 previously. Hope I haven't missed anything, but it can't do any
1952 harm, as the worst effect is to miss an optimization.
1954 (ii) If, despite (i), such a delivery is accidentally attempted, the
1955 routing retry time is respected, so at least it doesn't keep
1956 hammering the server.
1958 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1959 in ${readsocket because some servers need this prod.
1961 PH/32 Added some extra debug output when updating a wait-xxx database.
1963 PH/33 The hint "could be header name not terminated by colon", which has been
1964 given for certain expansion errors for a long time, was not being given
1965 for the ${if def:h_colon_omitted{... case.
1967 PH/34 The spec says: "With one important exception, whenever a domain list is
1968 being scanned, $domain contains the subject domain." There was at least
1969 one case where this was not true.
1971 PH/35 The error "getsockname() failed: connection reset by peer" was being
1972 written to the panic log as well as the main log, but it isn't really
1973 panic-worthy as it just means the connection died rather early on. I have
1974 removed the panic log writing for the ECONNRESET error when getsockname()
1977 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1978 runs only) independently of the message's sender address. This meant
1979 that, if the 4xx error was in fact related to the sender, a different
1980 message to the same recipient with a different sender could confuse
1981 things. In particualar, this can happen when sending to a greylisting
1982 server, but other circumstances could also provoke similar problems.
1983 I have changed the default so that the retry time for these errors is now
1984 based a combination of the sender and recipient addresses. This change
1985 can be overridden by setting address_retry_include_sender=false in the
1988 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1989 remote server are returned as part of bounce messages. This was not
1990 happening for LMTP over a pipe (the lmtp transport), but now it is the
1991 same for both kinds of LMTP.
1993 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1994 in header lines that were in fact CNAMEs. This is no longer the case.
1996 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1997 and queue runs started by the daemon processed all messages. This has
1998 been fixed so that -R and -S can now usefully be given with -q<time>.
2000 PH/40 Import PCRE release 6.7 (fixes some bugs).
2002 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2004 PH/42 Give an error if -q is specified more than once.
2006 PH/43 Renamed the variables $interface_address and $interface_port as
2007 $received_ip_address and $received_port, to make it clear that these
2008 values apply to message reception, and not to the outgoing interface when
2009 a message is delivered. (The old names remain recognized, of course.)
2011 PH/44 There was no timeout on the connect() call when using a Unix domain
2012 socket in the ${readsocket expansion. There now is.
2014 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2015 be meaningful with "accept".
2017 SC/01 Eximstats V1.43
2018 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2020 SC/02 Eximstats V1.44
2021 Use a glob alias rather than an array ref in the generated
2022 parser. This improves both readability and performance.
2024 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2025 Collect SpamAssassin and rejection statistics.
2026 Don't display local sender or destination tables unless
2027 there is data to show.
2028 Added average volumes into the top table text output.
2030 SC/04 Eximstats V1.46
2031 Collect data on the number of addresses (recipients)
2032 as well as the number of messages.
2034 SC/05 Eximstats V1.47
2035 Added 'Message too big' to the list of mail rejection
2036 reasons (thanks to Marco Gaiarin).
2038 SC/06 Eximstats V1.48
2039 Mainlog lines which have GMT offsets and are too short to
2040 have a flag are now skipped.
2042 SC/07 Eximstats V1.49 (Alain Williams)
2043 Added the -emptyok flag.
2045 SC/08 Eximstats V1.50
2046 Fixes for obtaining the IP address from reject messages.
2048 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2049 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2050 whitesspace changes from 4.64-PH/27
2052 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2055 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2056 are found, allow negative numbers in numeric criteria)
2058 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2060 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2061 to match changes made in 4.64-PH/43
2063 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2065 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2066 "domain" is actually an IP address, and force a failure. This locks out
2067 those revolvers/nameservers that support "A-for-A" lookups, in
2068 contravention of the specifications.
2070 PH/48 When a host name was looked up from an IP address, and the subsequent
2071 forward lookup of the name timed out, the host name was left in
2072 $sender_host_name, contrary to the specification.
2074 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2075 restricted to single-key lookups, Exim was not diagnosing an error if
2076 * or *@ was used with a query-style lookup.
2078 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2080 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2081 long ago, but noone interested enough thought of it. Let's just say that
2082 the "1.1" means that there are some new functions that weren't there at
2083 some point in the past.
2085 PH/51 Error processing for expansion failure of helo_data from an smtp
2086 transport during callout processing was broken.
2088 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2089 tested/used via the -bh/-bhc/-bs options.
2091 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2092 bug, fixed in subsequent PCRE releases).
2094 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2095 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2097 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2103 SC/01 Use a glob alias rather than an array ref in eximstats generated
2104 parser. This improves both readability and performance.
2106 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2107 Don't display local sender or destination tables in eximstats unless
2108 there is data to show.
2109 Added average volumes into the eximstats top table text output.
2111 SC/03 Collect data on the number of addresses (recipients) as well
2112 as the number of messages in eximstats.
2114 TF/01 Correct an error in the documentation for the redirect router. Exim
2115 does (usually) call initgroups() when daemonizing.
2117 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2118 with consistent privilege compared to when running as a daemon.
2120 TF/03 Note in the spec that $authenticated_id is not set for local
2121 submissions from trusted users.
2123 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2124 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2126 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2127 by adding some example configuration directives to the default
2128 configuration file. A little bit of work is required to uncomment the
2129 directives and define how usernames and passwords are checked, but
2130 there is now a framework to start from.
2132 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2133 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2134 without this. I don't know how relevant this is to other LDAP libraries.
2136 PH/02 Add the verb name to the "unknown ACL verb" error.
2138 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2140 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2142 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2143 directory not expanded when it should be if an expanded home directory
2144 was set for the address (which is overridden by the transport).
2146 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2149 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2150 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2151 because it is too late at that time, and has no effect.
2153 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2154 security issue with \' (bugzilla #107). I could not use the
2155 PQescapeStringConn() function, because it needs a PGconn value as one of
2158 PH/08 When testing addresses using -bt, indicate those final addresses that
2159 are duplicates that would not cause an additional delivery. At least one
2160 person was confused, thinking that -bt output corresponded to deliveries.
2161 (Suppressing duplicates isn't a good idea as you lose the information
2162 about possibly different redirections that led to the duplicates.)
2164 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2165 systems where poll() doesn't work, in particular OS X.
2167 PH/10 Added more information to debugging output for retry time not reached.
2169 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2170 operations in malware.c.
2172 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2175 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2176 syslog_duplication set false, log lines that would normally be written
2177 both the the main log and to the reject log were not written to syslog at
2180 PH/14 In the default configuration, change the use of "message" in ACL warn
2181 statements to "add_header".
2183 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2184 not followed by a command (e.g. "seen endif").
2186 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2187 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2190 PH/17 Added extra conditions to the default value of delay_warning_condition
2194 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2195 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2196 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2199 The Auto-Submitted: and various List- headers are standardised, whereas I
2200 don't think Precedence: ever was.
2202 PH/18 Refactored debugging code in route_finduser() to show more information,
2203 in particular, the error code if getpwnam() issues one.
2205 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2206 This is apparently needed in addition to the PH/07 change above to avoid
2207 any possible encoding problems.
2209 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2210 but not after initializing Perl.
2212 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2213 output them only if debugging. By default they are written stderr,
2214 apparently, which is not desirable.
2216 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2219 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2222 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2224 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2225 authenticated or an ident call has been made. Suppress the default
2226 values for $authenticated_id and $authenticated_sender (but permit -oMai
2227 and -oMas) when testing with -bh.
2229 PH/24 Re-jigged the order of the tests in the default configuration so that the
2230 tests for valid domains and recipients precede the DNS black list and CSA
2231 tests, on the grounds that those ones are more expensive.
2233 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2234 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2235 This bug exists in every version of Exim that I still have, right back to
2238 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2239 However, an attempt to turn on case-sensitivity in a regex key by
2240 including (?-i) didn't work because the subject string was already
2241 lowercased, and the effects were non-intuitive. It turns out that a
2242 one-line patch can be used to allow (?-i) to work as expected.
2248 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2249 other effects) broke the use of negated acl sub-conditions.
2251 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2254 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2255 "Deny" causes Exim to reject the incoming connection with a 554 error.
2256 Unfortunately, if there is a major crisis, such as a disk failure,
2257 tcp-wrappers gives "deny", whereas what one would like would be some
2258 kind of temporary error. A kludge has been added to help with this.
2259 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2260 554 error is used if errno is still zero or contains ENOENT (which occurs
2261 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2264 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2266 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2267 errors. Otherwise a message that provokes a temporary error (when other
2268 messages do not) can cause a whole host to time out.
2270 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2271 addresses were routed directly to files or pipes from a redirect router.
2272 File deliveries just didn't batch; pipe deliveries might have suffered
2275 PH/06 A failure to get a lock for a hints database would erroneously always say
2276 "Failed to get write lock", even when it was really a read lock.
2278 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2279 of 0600. This has been changed to use the value of the lockfile_mode
2280 option (which defaults to 0600).
2282 PH/08 Applied small patch from the Sieve maintainer.
2284 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2285 folder from quota calculations, a direct delivery into this folder messed
2286 up the contents of the maildirsize file. This was because the regex was
2287 used only to exclude .Trash (or whatever) when the size of the mailbox
2288 was calculated. There was no check that a delivery was happening into an
2289 excluded directory. This bug has been fixed by ignoring all quota
2290 processing for deliveries into excluded directories.
2292 PH/10 Added the maildirfolder_create_regex option to appendfile.
2298 PH/01 The code for finding all the local interface addresses on a FreeBSD
2299 system running IPv6 was broken. This may well have applied to all BSD
2300 systems, as well as to others that have similar system calls. The broken
2301 code found IPv4 interfaces correctly, but gave incorrect values for the
2302 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2303 that it would not match correctly against @[] and not recognize the IPv6
2306 PH/02 The ipliteral router was not recognizing addresses of the form user@
2307 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2309 PH/03 Added disable_ipv6.
2311 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2312 decoded form, because it is most often used to construct To: headers
2313 lines in autoreplies, and the decoded form may well be syntactically
2314 invalid. However, $reply_address has leading white space removed, and all
2315 newlines turned into spaces so that the autoreply transport does not
2318 PH/05 If group was specified without a user on a router, and no group or user
2319 was specified on a transport, the group from the router was ignored.
2321 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2322 for visible compile-time settings that can be used to change these
2323 numbers, for those that want even more. Backwards compatibility with old
2324 spool files has been maintained. However, going back to a previous Exim
2325 release will lost any variables that are in spool files.
2327 PH/07 Two small changes when running in the test harness: increase delay when
2328 passing a TCP/IP connection to a new process, in case the original
2329 process has to generate a bounce, and remove special handling of
2330 127.0.0.2 (sic), which is no longer necessary.
2332 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2333 be the same on different OS.
2335 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2338 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2339 whether --show-vars was specified or not
2341 JJ/02 exipick: Added support for new ACL variable spool format introduced
2344 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2345 syntactically invalid From: or Reply-to: line, and a filter used this to
2346 generate an autoreply, and therefore failed to obtain an address for the
2347 autoreply, Exim could try to deliver to a non-existent relative file
2348 name, causing unrelated and misleading errors. What now happens is that
2349 it logs this as a hard delivery error, but does not attempt to create a
2352 PH/11 The exinext utility has a -C option for testing purposes, but although
2353 the given file was scanned by exinext itself; it wasn't being passed on
2354 when Exim was called.
2356 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2357 an end-of-file indication when reading a command response.
2359 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2360 compiled. In many other places in Exim, IPv6 addresses are always
2361 recognized, so I have changed this. It also means that IPv4 domain
2362 literals of the form [IPV4:n.n.n.n] are now always recognized.
2364 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2365 used if the router is not running as root, for example, when verifying at
2366 ACL time, or when using -bh. The debugging output from this situation was
2367 non-existent - all you got was a failure to exec. I have made two
2370 (a) Failures to set uid/gid, the current directory, or a process leader
2371 in a subprocess such as that created by queryprogram now generate
2372 suitable debugging ouput when -d is set.
2374 (b) The queryprogram router detects when it is not running as root,
2375 outputs suitable debugging information if -d is set, and then runs
2376 the subprocess without attempting to change uid/gid.
2378 PH/15 Minor change to Makefile for building test_host (undocumented testing
2381 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2382 additional section of a DNS packet that returns MX or SRV records.
2383 Instead, it always explicitly searches for A/AAAA records. This avoids
2384 major problems that occur when a DNS server includes only records of one
2385 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2386 fixed another bug: if SRV records were looked up and the corresponding
2387 address records were *not* found in the additional section, the port
2388 values from the SRV records were lost.
2390 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2391 using the correct key (the original address) when searching the retry
2392 rules in order to find which one to use for generating the retry hint.
2394 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2395 adding the default one. Similarly, if it contains a Reply-To: header, the
2396 errors_reply_to option, if set, is not used.
2398 PH/19 When calculating a retry time, Exim used to measure the "time since
2399 failure" by looking at the "first failed" field in the retry record. Now
2400 it does not use this if it is later than than the arrival time of the
2401 message. Instead it uses the arrival time. This makes for better
2402 behaviour in cases where some deliveries succeed, thus re-setting the
2403 "first failed" field. An example is a quota failure for a huge message
2404 when small messages continue to be delivered. Without this change, the
2405 "time since failure" will always be short, possible causing more frequent
2406 delivery attempts for the huge message than are intended.
2407 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2409 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2410 $1, $2, $3) because the numerical variables can be reset during some
2411 expansion items (e.g. "match"), thereby losing the authentication data.
2413 PH/21 Make -bV show the size of off_t variables so that the test suite can
2414 decide whether to run tests for quotas > 2G.
2416 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2417 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2418 filecount value is greater than 2G or if a quota value is greater than 2G
2419 on a system where the size of off_t is not greater than 4, a panic error
2422 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2423 never match. The debug and -bh output now contains an explicit error
2424 message indicating a malformed IPv4 address or mask.
2426 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2427 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2428 PH/23 above applies.
2430 PH/25 Do not write to syslog when running in the test harness. The only
2431 occasion when this arises is a failure to open the main or panic logs
2432 (for which there is an explicit test).
2434 PH/26 Added the /no_tell option to "control=freeze".
2436 PH/27 If a host name lookup failed very early in a connection, for example, if
2437 the IP address matched host_lookup and the reverse lookup yielded a name
2438 that did not have a forward lookup, an error message of the form "no IP
2439 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2440 could be logged. Now it outputs the IP address instead of "NULL".
2442 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2443 allows the sender and the authenticated sender to be set when
2444 submitting a message from within Exim. Since child_open_exim() is
2445 documented for local_scan(), the new function should be too.
2447 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2448 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2449 results in an empty string is now treated as unset.
2451 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2453 PH/31 Added sender_verify_fail logging option.
2455 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2456 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2457 tidied the source and removed it altogether.
2459 PH/33 When a queue run was abandoned because the load average was too high, a
2460 log line was always written; now it is written only if the queue_run log
2461 selector is set. In addition, the log line for abandonment now contains
2462 information about the queue run such as the pid. This is always present
2463 in "start" and "stop" lines but was omitted from the "abandon" line.
2465 PH/34 Omit spaces between a header name and the colon in the error message that
2466 is given when verify = headers_syntax fails (if there are lots of them,
2467 the message gets confusing).
2469 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2470 names, as there are now some PTR records that contain slashes. This check
2471 is only to protect against broken name servers that fall over on strange
2472 characters, so the fact that it applies to all lookups doesn't matter.
2474 PH/36 Now that the new test suite is complete, we can remove some of the
2475 special code in Exim that was needed for the old test suite. For example,
2476 sorting DNS records because real resolvers return them in an arbitrary
2477 order. The new test suite's fake resolver always returns records in the
2480 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2481 bounces) except when queue_only is set, to avoid logging races between
2482 the different processes.
2484 PH/38 Panic-die if .include specifies a non-absolute path.
2486 PH/39 A tweak to the "H" retry rule from its user.
2488 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2489 a label. They prevented compilation on older perls.
2491 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2492 a warning to be raised on newish perls.
2494 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2495 on queue. Changes to match documented behaviour of showing count of
2496 messages matching specified criteria.
2498 PH/40 Changed the default ident timeout from 30s to 5s.
2500 PH/41 Added support for the use of login_cap features, on those BSD systems
2501 that have them, for controlling the resources used by pipe deliveries.
2503 PH/42 The content-scanning code uses fopen() to create files in which to put
2504 message data. Previously it was not paying any attention to the mode of
2505 the files. Exim runs with umask(0) because the rest of the code creates
2506 files with open(), and sets the required mode explicitly. Thus, these
2507 files were ending up world-writeable. This was not a big issue, because,
2508 being within the spool directory, they were not world-accessible. I have
2509 created a function called modefopen, which takes an additional mode
2510 argument. It sets umask(777), creates the file, chmods it to the required
2511 mode, then resets the umask. All the relevant calls to fopen() in the
2512 content scanning code have been changed to use this function.
2514 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2515 to 24 hours. This avoids potential overflow problems when processing G
2516 and H retry rules. I suspect nobody ever tinkers with this value.
2518 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2520 PH/45 When the plaintext authenticator is running as a client, the server's
2521 challenges are checked to ensure they are valid base64 strings. By
2522 default, the authentication attempt is cancelled if an invalid string is
2523 received. Setting client_ignore_invalid_base64 true ignores these errors.
2524 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2525 they are received. Thus, the responses can be made to depend on the
2526 challenges. If an invalid string is ignored, an empty string is placed in
2529 PH/46 Messages that are created by the autoreply transport now contains a
2530 References: header, in accordance with RFCs 2822 and 3834.
2532 PH/47 Added authenticated_sender_force to the smtp transport.
2534 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2536 PH/49 Installed latest patch from the Sieve maintainer.
2538 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2539 was also set, the appendfile transport was unnecessarily scanning a
2540 directory of message files (e.g. for maildir delivery) to find the count
2541 of files (along with the size), even though it did not need this
2542 information. It now does the scan only if it needs to find either the
2543 size of the count of files.
2545 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2547 PH/52 Two bugs concerned with error handling when the smtp transport is
2550 (i) Exim was not creating retry information for temporary errors given
2551 for individual recipients after the DATA command when the smtp transport
2552 was used in LMTP mode. This meant that they could be retried too
2553 frequently, and not timed out correctly.
2555 (ii) Exim was setting the flag that allows error details to be returned
2556 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2557 recipients that were returned after the DATA command.
2559 PH/53 This is related to PH/52, but is more general: for any failing address,
2560 when detailed error information was permitted to be returned to the
2561 sender, but the error was temporary, then after the final timeout, only
2562 "retry timeout exceeded" was returned. Now it returns the full error as
2563 well as "retry timeout exceeded".
2565 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2566 do this, and (what is worse) MTAs that accept it.
2568 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2569 will now be deprecated.
2571 PH/56 New os.c-cygwin from the Cygwin maintainer.
2573 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2574 formats (previously only available in exim formats via -bpr, -bpru,
2575 and -bpra. Now also available in native and exiqgrep formats)
2577 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2578 with very large, slow to parse queues
2580 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2582 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2584 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2585 responses to authentication challenges, though it was showing the
2586 challenges; (ii) I've removed the CR characters from the debug output for
2589 PH/58 Allow for the insertion of a newline as well as a space when a string
2590 is turned into more than one encoded-word during RFC 2047 encoding. The
2591 Sieve code now uses this.
2593 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2594 data_4xx, lost_connection, tls_required.
2596 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2597 message was being sent as an SMTP response.
2599 PH/61 Add -l and -k options to exicyclog.
2601 PH/62 When verifying, if an address was redirected to one new address, so that
2602 verification continued, and the new address failed or deferred after
2603 having set something in $address_data, the value of $address_data was not
2604 passed back to the ACL. This was different to the case when no
2605 redirection occurred. The value is now passed back in both cases.
2607 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2608 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2609 use login_cap.h, so on its own it isn't the distinguishing feature. The
2610 new name refers directly to the setclassresources() function.
2612 PH/65 Added configuration files for NetBSD3.
2614 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2616 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2617 is preferred over IPv4.
2619 PH/68 The bounce_return_message and bounce_return_body options were not being
2620 honoured for bounces generated during the reception of non-SMTP messages.
2621 In particular, this applied to messages rejected by the ACL. This bug has
2622 been fixed. However, if bounce_return_message is true and bounce_return_
2623 body is false, the headers that are returned for a non-SMTP message
2624 include only those that have been read before the error was detected.
2625 (In the case of an ACL rejection, they have all been read.)
2627 PH/69 The HTML version of the specification is now built in a directory called
2628 spec_html instead of spec.html, because the latter looks like a path with
2629 a MIME-type, and this confuses some software.
2631 PH/70 Catch two compiler warnings in sieve.c.
2633 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2634 function verify_get_ident() calls ip_connect() to connect a socket, but
2635 if the "connect()" function timed out, ip_connect() used to close the
2636 socket. However, verify_get_ident() also closes the socket later, and in
2637 between Exim writes to the log, which may get opened at this point. When
2638 the socket was closed in ip_connect(), the log could get the same file
2639 descriptor number as the socket. This naturally causes chaos. The fix is
2640 not to close the socket in ip_connect(); the socket should be closed by
2641 the function that creates it. There was only one place in the code where
2642 this was missing, in the iplookup router, which I don't think anybody now
2643 uses, but I've fixed it anyway.
2645 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2646 well as to direct DNS lookups. Otherwise the handling of names in host
2647 lists is inconsistent and therefore confusing.
2653 PH/01 Two changes to the default runtime configuration:
2655 (1) Move the checks for relay_from_hosts and authenticated clients from
2656 after to before the (commented out) DNS black list checks.
2658 (2) Add control=submission to the relay_from_hosts and authenticated
2659 clients checks, on the grounds that messages accepted by these
2660 statements are most likely to be submissions.
2662 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2664 (1) Generate an error if the third argument for the ${prvs expansion is
2667 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2670 (3) Reset the variables that are obtained from the first argument of
2671 ${prvscheck and used in the second argument before leaving the code,
2672 because their memory is reclaimed, so using them afterwards may do
2675 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2676 one (it's much easier than Tom thought :-).
2678 (5) Because of (4), we can now allow for the use of $prvscheck_result
2679 inside the third argument.
2681 PH/03 For some reason, the default setting of PATH when running a command from
2682 a pipe transport was just "/usr/bin". I have changed it to
2685 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2686 anything to be listed in the output from -bV.
2688 PH/05 When a filter generated an autoreply, the entire To: header line was
2689 quoted in the delivery log line, like this:
2691 => >A.N.Other <ano@some.domain> <original@ddress> ...
2693 This has been changed so that it extracts the operative address. There
2694 may be more than one such address. If so, they are comma-separated, like
2697 => >ano@some.domain,ona@other.domain <original@ddress> ...
2699 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2700 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2701 not being looked up in the rDNS to get a host name, Exim was showing the
2702 IP address twice in Received: lines, even though the IP addresses were
2703 identical. For example:
2705 Received: from [1.2.3.4] (helo=[1.2.3.4])
2707 However, if the real host name was known, it was omitting the HELO data
2708 if it matched the actual IP address. This has been tidied up so that it
2709 doesn't show the same IP address twice.
2711 PH/07 When both +timestamp and +memory debugging was on, the value given by
2712 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2713 called by the debug printing, thereby overwriting the timestamp buffer.
2714 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2716 PH/08 When the original message was included in an autoreply transport, it
2717 always said "this is a copy of the message, including all the headers",
2718 even if body_only or headers_only was set. It now gives an appropriate
2721 PH/09 Applied a patch from the Sieve maintainer which:
2723 o fixes some comments
2724 o adds the (disabled) notify extension core
2725 o adds some debug output for the result of if/elsif tests
2726 o points to the current vacation draft in the documentation
2727 and documents the missing references header update
2731 o fixes a bug in processing the envelope test (when testing
2732 multiple envelope elements, the last element determinted the
2735 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2736 Electronic Mail") by including:
2738 Auto-submitted: auto-generated
2740 in the messages that it generates (bounce messages and others, such as
2741 warnings). In the case of bounce messages for non-SMTP mesages, there was
2742 also a typo: it was using "Auto_submitted" (underscore instead of
2743 hyphen). Since every message generated by Exim is necessarily in response
2744 to another message, thes have all been changed to:
2746 Auto-Submitted: auto-replied
2748 in accordance with these statements in the RFC:
2750 The auto-replied keyword:
2752 - SHOULD be used on messages sent in direct response to another
2753 message by an automatic process,
2755 - MUST NOT be used on manually-generated messages,
2757 - MAY be used on Delivery Status Notifications (DSNs) and Message
2758 Disposition Notifications (MDNs),
2760 - MUST NOT be used on messages generated by automatic or periodic
2761 processes, except for messages which are automatic responses to
2764 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2765 to the default Received: header definition.
2767 PH/12 Added log selector acl_warn_skipped (default on).
2769 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2770 variables because (a) they are in the wrong storage pool and (b) even if
2771 they were copied, it wouldn't work properly because of the caching.
2773 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2774 checking when decoding. Apparently there are clients that generate
2775 overlong encoded strings. Why am I not surprised?
2777 PH/15 If the first argument of "${if match_address" was not empty, but did not
2778 contain an "@" character, Exim crashed. Now it writes a panic log message
2779 and treats the condition as false.
2781 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2783 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2784 extension "envelope-auth". The code is finished and in agreement with
2785 other implementations, but there is no documentation so far and in fact,
2786 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2787 not changing the active code.
2789 Print executed "if" and "elsif" statements when debugging is used. This
2790 helps a great deal to understand what a filter does.
2792 Document more things not specified clearly in RFC3028. I had all this
2793 sorted out, when out of a sudden new issues came to my mind. Oops."
2795 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2798 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2799 canonical form (as documented). However, after a host name lookup from
2800 the IP address, check_host() was doing a simple string comparison with
2801 addresses acquired from the DNS when checking that the found name did
2802 have the original IP as one of its addresses. Since any found IPv6
2803 addresses are likely to be in abbreviated form, the comparison could
2804 fail. Luckily, there already exists a function for doing the comparison
2805 by converting both addresses to binary, so now that is used instead of
2806 the text comparison.
2808 PH/20 There was another similar case to PH/19, when a complete host name was
2809 given in a host list; looking up its IP address could give an abbreviated
2810 form, whereas the current host's name might or might not be abbreviated.
2811 The same fix has been applied.
2817 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2818 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2821 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2822 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2824 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2826 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2827 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2828 Also, for "Auto-Submitted", treat anything other than "no" as
2829 non-personal, in accordance with RFC 3834. (Previously it treated
2830 anything starting "auto-" as non-personal.)
2832 TF/01 The control=submission/name=... option had a problem with syntax
2833 errors if the name included a slash character. The /name= option
2834 now slurps the rest of the string, so it can include any characters
2835 but it must come last in the list of options (after /sender_retain
2838 PH/05 Some modifications to the interface to the fake nameserver for the new
2846 TK/01 Added the "success_on_redirect" address verification option. See
2847 NewStuff for rationale and an example.
2849 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2851 PH/02 Patch to exigrep to allow it to work on syslog lines.
2853 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2854 fread() to skip over the body file's header line, because in Cygwin the
2855 header line is locked and is inaccessible.
2857 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2858 co-exist for some time) to make it clear that it is the Exim ID that is
2859 referenced, not the Message-ID: header line.
2861 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2862 string_format() function, because snprintf() does not exist on all
2865 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2866 ${stat: expansion item.
2868 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2869 protocol synchronization error", to keep the pedants happy.
2871 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2872 well as for IRIX systems, when gcc is being used. See the host.c source
2875 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2877 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2880 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2881 transport and to the smtp transport in LMTP mode.
2883 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2885 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2886 run a filter in a subprocess. This could lead to confusion in subsequent
2887 lookups in the parent process. There should also be a search_tidyup() at
2888 the end of the subprocess.
2890 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2891 only if the host matched helo_try_verify_hosts, which caused the
2892 verification to occur when the EHLO/HELO command was issued. The ACL just
2893 tested the remembered result. Now, if a previous verification attempt has
2894 not happened, "verify = helo" does it there and then.
2896 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2898 TK/03 Fix log output including CR from clamd.
2900 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2901 exist provoked a memory error which could cause a segfault.
2903 PH/15 Installed PCRE 6.2
2905 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2907 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2908 of the problem. Specifically, suggested +O2 rather than +O1 for the
2911 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2913 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2914 record, the port was not correctly logged when the outgoing_port log
2915 selector was set (it logged the transort's default port).
2917 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2918 fallback_hosts, and "hosts" in the smtp transport.
2920 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2921 host errors such as "Connection refused".
2923 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2924 authentication with radiusclient 0.4.9:
2926 - Error returned from rc_read_config was caught wrongly
2927 - Username/password not passed on to radius server due to wrong length.
2929 The presumption is that some radiusclient API changes for 4.51/PH/17
2930 were not taken care of correctly. The code is still untested by me (my
2931 Linux distribution still has 0.3.2 of radiusclient), but it was
2932 contributed by a Radius user.
2934 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2935 expanding the "port" option of the smtp transport.
2937 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2938 while reading a MIME header. Thanks to Tom Hughes for a patch.
2940 PH/24 Include config.h inside local_scan.h so that configuration settings are
2943 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2944 that in an ACL for RCPT (for example), you can examine exactly what was
2947 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2948 commands, but it was not correctly comparing the address with the actual
2949 client host address. Thus, it would show the EHLO address in Received:
2950 header lines when this was not necessary.
2952 PH/27 Added the % operator to ${eval:}.
2954 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2955 it should be ignoring failures (because with -C, for example, it has lost
2956 privilege). It wasn't ignoring creation failures other than "already
2959 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2962 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2963 because an input line was too long, either on its own, or by virtue of
2964 too many continuations, the temporary file was not being removed, and the
2965 return code was incorrect.
2967 PH/31 Missing "BOOL" in function definition in filtertest.c.
2969 PH/32 Applied Sieve patches from the maintainer.
2971 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2973 PH/33 Added "verify = not_blind".
2975 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2976 Local/Makefile (with some defaults set). These are used in built scripts
2977 such as exicyclog, but they have never been used in the exim_install
2978 script (though there are many overriding facilities there). I have
2979 arranged that the exim_install script now takes note of these two
2982 PH/35 Installed configuration files for Dragonfly.
2984 PH/36 When a locally submitted message by a trusted user did not contain a
2985 From: header, and the sender address was obtained from -f or from an SMTP
2986 MAIL command, and the trusted user did not use -F to supply a sender
2987 name, $originator_name was incorrectly used when constructing a From:
2988 header. Furthermore, $originator_name was used for submission mode
2989 messages from external hosts without From: headers in a similar way,
2990 which is clearly wrong.
2992 PH/37 Added control=suppress_local_fixups.
2994 PH/38 When log_selector = +received_sender was set, and the addition of the
2995 sender made the log line's construction buffer exactly full, or one byte
2996 less than full, an overflow happened when the terminating "\n" was
2999 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3000 when the result of a list match is failure because a DNS lookup failed.
3002 PH/40 RM_COMMAND is now used in the building process.
3004 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3005 the "build-* directories that it finds.
3007 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3008 address in a domain literal was a prefix of an interface address.
3010 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3011 when verifying a sender address, unless rewrite_headers is false.
3013 PH/44 Wrote a long comment about why errors_to addresses are verified as
3014 recipients, not senders.
3016 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3017 the ratelimit ACL was added.
3019 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3021 PH/46 Added extra information about PostgreSQL errors to the error string.
3023 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3024 avoiding the need to install special zones in a real server. This is
3025 backwards compatible; if it can't find the fake resolver, it drops back.
3026 Thus, both old and new test suites can be run.
3028 TF/02 Added util/ratelimit.pl
3030 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3031 clock is set back in time.
3033 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3034 Candler <B.Candler@pobox.com>.
3036 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3037 OK for addresses that are the result of redirections.
3039 PH/48 A number of further additions for the benefit of the new test suite,
3040 including a fake gethostbyname() that interfaces to the fake DNS resolver
3043 TF/06 The fix for widen_domains has also been applied to qualify_single and
3044 search_parents which are the other dnslookup options that can cause
3047 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3050 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3052 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3053 service" if it encountered an MX record with an empty target hostname.
3054 The message is now "an MX or SRV record indicated no SMTP service".
3056 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3057 if the DNS of the sending site is misconfigured. This is quite a
3058 common situation. This change restores the behaviour of treating a
3059 helo verification defer as a failure.
3061 PH/51 If self=fail was set on a router, the bounce message did not include the
3062 actual error message.
3068 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3070 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3071 command itself ended in error, the underlying message about the transport
3072 filter timeout was being overwritten with the pipe command error. Now the
3073 underlying error message should be appended to the second error message.
3075 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3077 PH/02 Reduce the amount of output that "make" produces by default. Full output
3078 can still be requested.
3080 PH/03 The warning log line about a condition test deferring for a "warn" verb
3081 was being output only once per connection, rather than after each
3082 occurrence (because it was using the same function as for successful
3083 "warn" verbs). This seems wrong, so I have changed it.
3085 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3086 it should not have, which might have caused a crash in the right
3087 circumstances, but probably never did.
3089 PH/04 Installed a modified version of Tony Finch's patch to make submission
3090 mode fix the return path as well as the Sender: header line, and to
3091 add a /name= option so that you can make the user's friendly name appear
3094 TF/03 Added the control = fakedefer ACL modifier.
3096 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3097 Mark Lowes for thorough testing.
3099 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3101 TK/03 Merged latest SRS patch from Miles Wilton.
3103 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3104 with the definition in sysexits.h (which is #included earlier).
3105 Fortunately, Exim does not actually use EX_OK. The code used to try to
3106 preserve the sysexits.h value, by assumimg that macro definitions were
3107 scanned for macro replacements. I have been disabused of this notion,
3108 so now the code just undefines EX_OK before #including unistd.h.
3110 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3111 in the smtp transport. When a block could not be written in a single
3112 write() function, the timeout was being re-applied to each part-write.
3113 This seems wrong - if the receiver was accepting one byte at a time it
3114 would take for ever. The timeout is now adjusted when this happens. It
3115 doesn't have to be particularly precise.
3117 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3118 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3120 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3121 without a domain if the check to <postmaster@domain> fails.
3123 SC/01 Eximstats: added -xls and the ability to specify output files
3124 (patch written by Frank Heydlauf).
3126 SC/02 Eximstats: use FileHandles for outputing results.
3128 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3130 SC/04 Eximstats: fixed display of large numbers with -nvr option
3132 SC/05 Eximstats: fixed merging of reports with empty tables.
3134 SC/06 Eximstats: added the -include_original_destination flag
3136 SC/07 Eximstats: removed tabs and trailing whitespace.
3138 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3140 TK/06 MBOX spool code: Add real "From " MBOX separator line
3141 so the .eml file is really in mbox format (even though
3142 most programs do not really care). Patch from Alex Miller.
3144 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3145 The latter is generated from $received_to and is only set if the
3146 message has one envelope recipient. SA can use these headers,
3147 obviously out-of-the-box. Patch from Alex Miller.
3149 PH/08 The ${def test on a variable was returning false if the variable's
3150 value was "0", contrary to what the specification has always said!
3151 The result should be true unless the variable is empty.
3153 PH/09 The syntax error of a character other than { following "${if
3154 def:variable_name" (after optional whitespace) was not being diagnosed.
3155 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3156 accidental colon was present, for example, could give incorrect results.
3158 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3159 result is used (not including appendfile, where other changes are about
3162 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3163 This involved changing a lot of size variables from int to off_t. It
3164 should work with maildirs and everything.
3166 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3167 spamd dying while we are connected to it.
3169 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3172 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3173 function for BDB 4.3.
3175 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3177 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3178 the sources for PCRE have been re-arranged and more files are now
3181 PH/15 The code I had for printing potentially long long variables in PH/11
3182 above was not the best (it lost precision). The length of off_t variables
3183 is now inspected at build time, and an appropriate printing format (%ld
3184 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3185 to be "long long int" or "long int". This is needed for the internal
3186 formatting function string_vformat().
3188 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3189 the configuration file to be ":syslog", then the script "guesses" where
3190 the logs files are, rather than using the compiled in default. In our
3191 case the guess is not the same as the compiled default, so the script
3192 suddenly stopped working when I started to use syslog. The patch checks
3193 to see if log_file_path is "". If so, it attempts to read it from exim
3194 with no configuration file to get the compiled in version, before it
3195 falls back to the previous guessing code."
3197 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3198 implementing BATV in an Exim configuration. See NewStuff for the gory
3201 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3202 Makefile that are specific to HP-UX.
3204 PH/18 If the "use_postmaster" option was set for a recipient callout together
3205 with the "random" option, the postmaster address was used as the MAIL
3206 FROM address for the random test, but not for the subsequent recipient
3207 test. It is now used for both.
3209 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3210 patch removes a few documentation additions to RFC 3028, because the
3211 latest draft now contains them. It adds the new en;ascii-case comparator
3212 and a new error check for 8bit text in MIME parts. Comparator and
3213 require names are now matched exactly. I enabled the subaddress
3214 extension, but it is not well tested yet (read: it works for me)."
3216 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3217 rework some of the code of TK/09 above to avoid the hardwired use of
3218 "%lld" and "long long". Replaced the call to snprintf() with a call to
3221 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3222 records point to non-existent hosts", "retry timeout exceeded", and
3223 "retry time not reached for any host after a long failure period".
3225 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3226 experimental DomainKeys support:
3228 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3229 (2) On an error such as an illegally used "control", the wrong name for
3230 the control was given.
3232 These problems did NOT occur unless DomainKeys support was compiled.
3234 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3236 PH/24 Added ${if match_ip condition.
3238 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3239 databases so that it will be absolutely obvious if a crash occurs in the
3240 DB library. This is a regular occurrence (often caused by mis-matched
3243 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3244 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3245 picked up on a user's system that detects such things. There doesn't seem
3246 to be a gcc warning option for this - only an attribute that has to be
3247 put on the function's prototype. It seems that in Fedora Core 4 they have
3248 set this on a number of new functions. No doubt there will be more in due
3251 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3252 specify a transport. However, if an address that was verified by such a
3253 router was the subject of a callout, Exim crashed because it tried to
3254 read the rcpt_include_affixes from the non-existent transport. Now it
3255 just assumes that the setting of that option is false. This bug was
3256 introduced by 4.51/PH/31.
3258 PH/28 Changed -d+all to exclude +memory, because that information is very
3259 rarely of interest, but it makes the output a lot bigger. People tend to
3260 do -d+all out of habit.
3262 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3263 code in os-type was giving problems when libc.so lives in lib64, like on
3266 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3267 aren't the modern standard, and it seems that some systems' include files
3268 don't always have them. Exim was already checking for some of the newer
3269 ones like T_AAAA, and defining it itself. I've added checks for all the
3270 record types that Exim uses.
3272 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3273 not automatically generating a new one, as it is supposed to. This
3274 prevented TLS from working. If the file did exist, but contained invalid
3275 data, a new version was generated, as expected. It was only the case of a
3276 non-existent file that was broken.
3278 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3279 with a change in libdomainkeys > 0.64.
3281 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3282 from DNS. If the selector record carries the flag, it now has
3283 precedence over the domain-wide flag.
3285 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3287 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3288 the use of a port name that isn't defined in /etc/services) occurred, the
3289 message was deferred as in a normal delivery, and thus remained on the
3290 spool, instead of being failed because of the mua_wrapper setting. This
3291 is now fixed, and I tidied up some of the mua_wrapper messages at the
3294 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3295 the messages in a hash of arrays rather than using individual hashes.
3296 This is a bit cleaner and results in dramatic memory savings, albeit
3297 at a slight CPU cost.
3299 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3300 as requested by Marc Sherman.
3302 SC/10 Eximstats: added histograms for user specified patterns as requested
3305 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3307 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3308 fopen() in the content-scanning modules that did not already have it.
3314 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3315 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3317 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3319 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3321 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3322 to test Sieve filters that use "vacation".
3324 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3325 that changes the way the GnuTLS parameters are stored in the cache file.
3326 The new format can be generated externally. For backward compatibility,
3327 if the data in the cache doesn't make sense, Exim assumes it has read an
3328 old-format file, and it generates new data and writes a new file. This
3329 means that you can't go back to an older release without removing the
3332 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3333 work if there are any delivery delays because "one_time" forces the
3334 parent to be marked "delivered", so its unseen clone is never tried
3335 again. For this reason, Exim now forbids the simultaneous setting of
3338 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3339 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3340 entry if you want to know the details. The fix, however, neglected to
3341 consider the case when local delivery batching is involved. The test for
3342 "previously delivered" was not happening when checking to see if an
3343 address could be batched with a previous (undelivered) one; under
3344 certain circumstances this could lead to multiple deliveries to the same
3347 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3348 in its include files, and this causes problems building Exim.
3350 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3351 header_syntax) but Exim was just ignoring anything given after a slash.
3352 In particular, this caused confusion with an attempt to use "verify =
3353 reverse_host_lookup/defer_ok". An error is now given when options are
3354 supplied for verify items that do not have them. (Maybe reverse_host_
3355 lookup should have a defer_ok option, but that's a different point.)
3357 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3358 defined by RFC 821) to 2048, because there were problems with some AUTH
3359 commands, and RFC 1869 says the size should be increased for extended
3360 SMTP commands that take arguments.
3362 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3365 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3366 "unknown" error; now it says that the functionality isn't in the binary.
3368 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3369 an address' error message when a string expansion fails (syntax or
3370 whatever). Otherwise the password may appear in the log. Following change
3371 PH/42 below, there is no longer a chance of it appearing in a bounce
3374 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3376 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3377 of its addresses was ever tried. (Bugzilla bug #2.)
3379 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3380 the result incorrectly in the debug output. (It correctly added a newline
3381 to what was transported.)
3383 TF/01 Added $received_time.
3385 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3386 commented out examples of how to interface to a virus scanner and to
3387 SpamAssassin. Also added commented examples of av_scanner and
3388 spamd_address settings.
3390 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3391 and controls are allowed in which ACLs. There were a couple of minor
3392 errors. Some of the entries in the conditions table (which is a table of
3393 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3394 negation of where the condition IS allowed.
3396 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3398 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3399 header file does not have a version number, so I've had to invent a new
3400 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3401 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3402 radiusclient), but it was contributed by a Radius user.
3404 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3405 files or directories, for OpenSSL.
3407 PH/19 When an Exim process that is running as root has to create an Exim log
3408 file, it does so in a subprocess that runs as exim:exim so as to get the
3409 ownership right at creation (otherwise, other Exim processes might see
3410 the file with the wrong ownership). There was no test for failure of this
3411 fork() call, which would lead to the process getting stuck as it waited
3412 for a non-existent subprocess. Forks do occasionally fail when resources
3413 run out. I reviewed all the other calls to fork(); they all seem to check
3416 PH/20 When checking for unexpected SMTP input at connect time (before writing
3417 the banner), Exim was not dealing correctly with a non-positive return
3418 from the read() function. If the client had disconnected by this time,
3419 the result was a log entry for a synchronization error with an empty
3420 string after "input=" when read() returned zero. If read() returned -1
3421 (an event I could not check), uninitialized data bytes were printed.
3422 There were reports of junk text (parts of files, etc) appearing after
3425 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3427 PH/22 Added support for macro redefinition, and (re)definition in between
3428 driver and ACL definitions.
3430 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3431 forgetting to use the resulting value; it was using the unexpanded value.
3433 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3434 hadn't been configured. The fix is from Juergen Kreileder, who
3435 understands it better than I do:
3437 "Here's what I see happening with three configured cyrus_sasl
3438 authenticators configured (plain, login, cram-md5):
3440 On startup auth_cyrus_sasl_init() gets called for each of these.
3441 This means three calls to sasl_listmech() without a specified mech_list.
3442 => SASL tests which mechs of all available mechs actually work
3443 => three warnings about OTP not working
3444 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3446 With the patch, sasl_listmech() also gets called three times. But now
3447 SASL's mech_list option is set to the server_mech specified in the the
3448 authenticator. Or in other words, the answer from sasl_listmech()
3449 gets limited to just the mech you're testing for (which is different
3451 => the return list contains just 'plain' or 'login', 'cram-md5' or
3452 nothing depending on the value of ob->server_mech.
3454 I've just tested the patch: Authentication still works fine,
3455 unavailable mechs specified in the exim configuration are still
3456 caught, and the auth.log warnings about OTP are gone."
3458 PH/25 When debugging is enabled, the contents of the command line are added
3459 to the debugging output, even when log_selector=+arguments is not
3462 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3463 answer is "GNU", and only if the return is "GNU/something" is the answer
3466 PH/27 $acl_verify_message is now set immediately after the failure of a
3467 verification in an ACL, and so is available in subsequent modifiers. In
3468 particular, the message can be preserved by coding like this:
3470 warn !verify = sender
3471 set acl_m0 = $acl_verify_message
3473 Previously, $acl_verify_message was set only while expanding "message"
3474 and "log_message" when a very denied access.
3476 PH/28 Modified OS/os.c-Linux with
3478 -#ifndef OS_LOAD_AVERAGE
3479 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3481 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3482 nomenclature these days.)
3484 PH/29 Installed patch from the Sieve maintainer that adds the options
3485 sieve_useraddress and sieve_subaddress to the redirect router.
3487 PH/30 In these circumstances:
3488 . Two addresses routed to the same list of hosts;
3489 . First host does not offer TLS;
3490 . First host accepts first address;
3491 . First host gives temporary error to second address;
3492 . Second host offers TLS and a TLS session is established;
3493 . Second host accepts second address.
3494 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3495 and peerdn, if requested) that were in fact used only for the second
3498 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3499 attention to any local part prefix or suffix that was matched by the
3500 router that accepted the address. It now behaves in the same way as it
3501 does for delivery: the affixes are removed from the local part unless
3502 rcpt_include_affixes is set on the transport.
3504 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3505 timeout during the DATA phase of an incoming message.
3507 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3508 applied a patch sanctioned by the Sieve maintainer.
3510 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3511 the uid or gid is negative. A case of a negative gid caused this to be
3512 noticed. The fix allows for either to be negative.
3514 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3515 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3518 PH/36 The change PH/12 above was broken. Fixed it.
3520 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3521 the grounds that routing the same address twice would always produce the
3522 same answer. This might have been true once, but it is certainly no
3523 longer true now. Routing a child address may depend on the previous
3524 routing that produced that child. Some complicated redirection strategies
3525 went wrong when messages had multiple recipients, and made Exim's
3526 behaviour dependent on the order in which the addresses were given.
3528 I have moved the duplicate checking until after the routing is complete.
3529 Exim scans the addresses that are assigned to local and remote
3530 transports, and removes any duplicates. This means that more work will be
3531 done, as duplicates will always all be routed, but duplicates are
3532 presumably rare, so I don't expect this is of any significance.
3534 For deliveries to pipes, files, and autoreplies, the duplicate checking
3535 still happens during the routing process, since they are not going to be
3538 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3539 It corrects a timeout issue with spamd. This is Ian's comment: "The
3540 background is that sometimes spamd either never reads data from a
3541 connection it has accepted, or it never writes response data. The exiscan
3542 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3543 blindly assumes that writes won't block so it may never time out."
3545 PH/39 Allow G after quota size as well as K and M.
3547 PH/40 The value set for $authenticated_id in an authenticator may not contain
3548 binary zeroes or newlines because the value is written to log lines and
3549 to spool files. There was no check on this. Now the value is run through
3550 the string_printing() function so that such characters are converted to
3551 printable escape sequences.
3553 PH/41 $message_linecount is a new variable that contains the total number of
3554 lines in the message. Compare $body_linecount, which is the count for the
3557 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3558 bounce and delay warning messages, except in certain special cases, which
3561 (a) An SMTP error message from a remote host;
3562 (b) A message specified in a :fail: redirection;
3563 (c) A message specified in a "fail" command in a system filter;
3564 (d) A message specified in a FAIL return from the queryprogram router;
3565 (e) A message specified by the cannot_route_message router option.
3567 In these cases only, Exim does include the error details in bounce and
3568 warning messages. There are also a few cases where bland messages such
3569 as "unrouteable address" or "local delivery error" are given.
3571 PH/43 $value is now also set for the "else" part of a ${run expansion.
3573 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3574 being worked on, but at least Exim now implements the latest version to
3577 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3578 process to complete was treated as a delivery failure, a timeout while
3579 writing the message to the pipe was logged, but erroneously treated as a
3580 successful delivery. Such timeouts include transport filter timeouts. For
3581 consistency with the overall process timeout, these timeouts are now
3582 treated as errors, giving rise to delivery failures by default. However,
3583 there is now a new Boolean option for the pipe transport called
3584 timeout_defer, which, if set TRUE, converts the failures into defers for
3585 both kinds of timeout. A transport filter timeout is now identified in
3588 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3589 systems where "make" and "gmake" are different, calling "gmake" at top
3590 level broke things. I've arranged for the value of $(MAKE) to be passed
3591 from the Makefile to this script so that it can call the same version of
3595 A note about Exim versions 4.44 and 4.50
3596 ----------------------------------------
3598 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3599 changes of various kinds. As a consequence, a big documentation update was
3600 needed. This delayed the release for rather longer than seemed good, especially
3601 in the light of a couple of (minor) security issues. Therefore, the changes
3602 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3603 release. So 4.44 and 4.50 are in effect two different branches that both start
3606 I have left the 4.50 change log unchanged; it contains all the changes since
3607 4.43. The change log for 4.44 is below; many of its items are identical to
3608 those for 4.50. This seems to be the most sensible way to preserve the
3609 historical information.
3615 1. Minor wording change to the doc/README.SIEVE file.
3617 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3618 computation of the current number of files was incorrect.
3620 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3621 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3624 4. Give more explanation in the error message when the command for a transport
3625 filter fails to execute.
3627 5. There are several places where Exim runs a non-Exim command in a
3628 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3629 was being done only for the command run by the queryprogram router. It is
3630 now done for all such subprocesses. The other cases are: ${run, transport
3631 filters, and the commands run by the lmtp and pipe transports.
3633 6. Added CONFIGURE_GROUP build-time option.
3635 7. Some older OS have a limit of 256 on the maximum number of file
3636 descriptors. Exim was using setrlimit() to set 1000 as a large value
3637 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3638 systems. I've change it so that if it can't get 1000, it tries for 256.
3640 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3641 was an oversight, and furthermore, ever since the addition of extra
3642 controls (e.g. 4.43/32), the checks on when to allow different forms of
3643 "control" were broken. There should now be diagnostics for all cases when a
3644 control that does not make sense is encountered.
3646 9. Added the /retain_sender option to "control=submission".
3648 10. $recipients is now available in the predata ACL (oversight).
3650 11. Tidy the search cache before the fork to do a delivery from a message
3651 received from the command line. Otherwise the child will trigger a lookup
3652 failure and thereby defer the delivery if it tries to use (for example) a
3653 cached ldap connection that the parent has called unbind on.
3655 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3656 of $address_data from the recipient verification was clobbered by the
3657 sender verification.
3659 13. The value of address_data from a sender verification is now available in
3660 $sender_address_data in subsequent conditions in the ACL statement.
3662 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3664 15. Added a new option "connect=<time>" to callout options, to set a different
3667 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3668 was its contents. (It was OK if the option was not defined at all.)
3670 17. A "Completed" log line is now written for messages that are removed from
3671 the spool by the -Mrm option.
3673 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3674 information about exactly what failed.
3676 19. Added -dd to debug only the daemon process.
3678 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3679 handles timeouts, both on the server side and network timeouts. Renamed the
3680 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3682 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3683 It is now set to "smtps".
3685 22. $host_address is now set to the target address during the checking of
3686 ignore_target_hosts.
3688 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3689 being passed; this would have caused $sender_host_name to have been used if
3690 matching the list had actually called for a host name (not very likely,
3691 since this list is usually IP addresses). A host name is now passed as
3694 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3695 code that specifies a non-restarting handler (typically sigaction() in
3696 modern systems) in an attempt to fix a rare and obscure crash bug.
3698 25. Narrowed the window for a race in the daemon that could cause it to ignore
3699 SIGCHLD signals. This is not a major problem, because they are used only to
3700 wake it up if nothing else does.
3702 26. A malformed maildirsize file could cause Exim to calculate negative values
3703 for the mailbox size or file count. Odd effects could occur as a result.
3704 The maildirsize information is now recalculated if the size or filecount
3707 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3708 support for a long time. Removed HAVE_SYS_VFS_H.
3710 28. Installed the latest version of exipick from John Jetmore.
3712 29. In an address list, if the pattern was not a regular expression, an empty
3713 subject address (from a bounce message) matched only if the pattern was an
3714 empty string. Non-empty patterns were not even tested. This was the wrong
3715 because it is perfectly reasonable to use an empty address as part of a
3716 database query. An empty address is now tested by patterns that are
3717 lookups. However, all the other forms of pattern expect the subject to
3718 contain a local part and a domain, and therefore, for them, an empty
3719 address still always fails if the pattern is not itself empty.
3721 30. Exim went into a mad DNS loop when attempting to do a callout where the
3722 host was specified on an smtp transport, and looking it up yielded more
3723 than one IP address.
3725 31. Re-factored the code for checking spool and log partition space into a
3726 function that finds that data and another that does the check. The former
3727 is then used to implement four new variables: $spool_space, $log_space,
3728 $spool_inodes, and $log_inodes.
3730 32. The RFC2047 encoding function was originally intended for short strings
3731 such as real names; it was not keeping to the 75-character limit for
3732 encoded words that the RFC imposes. It now respects the limit, and
3733 generates multiple encoded words if necessary. To be on the safe side, I
3734 have increased the buffer size for the ${rfc2047: expansion operator from
3737 33. It is now permitted to omit both strings after an "if" condition; if the
3738 condition is true, the result is "true". As before, when the second string
3739 is omitted, a false condition yields an empty string. This makes it less
3740 cumbersome to write custom ACL and router conditions.
3742 34. Failure to deliver a bounce message always caused it to be frozen, even if
3743 there was an errors_to setting on the router. The errors_to setting is now
3746 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3747 canonical form (fully expanded) before being placed in
3748 $sender_host_address.
3750 36. The table in the code that translates DNS record types into text (T_A to
3751 "A" for instance) was missing entries for NS and CNAME. It is just possible
3752 that this could have caused confusion if both these types were looked up
3753 for the same domain, because the text type is used as part of Exim's
3754 per-process caching. But the chance of anyone hitting this buglet seems
3757 37. The dnsdb lookup has been extended in a number of ways.
3759 (1) There is a new type, "zns", which walks up the domain tree until it
3760 finds some nameserver records. It should be used with care.
3762 (2) There is a new type, "mxh", which is like "mx" except that it returns
3763 just the host names, not the priorities.
3765 (3) It is now possible to give a list of domains (or IP addresses) to be
3766 looked up. The behaviour when one of the lookups defers can be
3767 controlled by a keyword.
3769 (4) It is now possible to specify the separator character for use when
3770 multiple records are returned.
3772 38. The dnslists ACL condition has been extended: it is now possible to supply
3773 a list of IP addresses and/or domains to be looked up in a particular DNS
3776 39. Added log_selector=+queue_time_overall.
3778 40. When running the queue in the test harness, wait just a tad after forking a
3779 delivery process, to get repeatability of debugging output.
3781 41. Include certificate and key file names in error message when GnuTLS fails
3782 to set them up, because the GnuTLS error message doesn't include the name
3783 of the failing file when there is a problem reading it.
3785 42. Allow both -bf and -bF in the same test run.
3787 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3789 44. The "Exiscan patch" is now merged into the mainline Exim source.
3791 45. Sometimes the final signoff response after QUIT could fail to get
3792 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3793 < 0 before doing a fflush(). This bug looks as though it goes back to the
3794 introduction of TLS in release 3.20, but "sometimes" must have been rare
3795 because the tests only now provoked it.
3797 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3798 (this can affect the format of dates).
3800 47. exim_tidydb, when checking for the continued existence of a message for
3801 which it has found a message-specific retry record, was not finding
3802 messages that were in split spool directories. Consequently, it was
3803 deleting retry records that should have stayed in existence.
3805 48. Steve fixed some bugs in eximstats.
3807 49. The SPA authentication driver was not abandoning authentication and moving
3808 on to the next authenticator when an expansion was forced to fail,
3809 contradicting the general specification for all authenticators. Instead it
3810 was generating a temporary error. It now behaves as specified.
3812 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3813 (the order specifies the preference for clients). The order is now AES256,
3814 AES128, 3DES, ARCFOUR128.
3816 51. Small patch to Sieve code - explicitly set From: when generating an
3819 52. Exim crashed if a remote delivery caused a very long error message to be
3820 recorded - for instance if somebody sent an entire SpamAssassin report back
3821 as a large number of 550 error lines. This bug was coincidentally fixed by
3822 increasing the size of one of Exim's internal buffers (big_buffer) that
3823 happened as part of the Exiscan merge. However, to be on the safe side, I
3824 have made the code more robust (and fixed the comments that describe what
3827 53. Now that there can be additional text after "Completed" in log lines (if
3828 the queue_time_overall log selector is set), a one-byte patch to exigrep
3829 was needed to allow it to recognize "Completed" as not the last thing in
3832 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3833 patch that reportedly fixes this has been added. I am not expert enough to
3834 create a test for it. This is what the patch creator wrote:
3836 "I found a little strange behaviour of ldap code when working with
3837 Windows 2003 AD Domain, where users was placed in more than one
3838 Organization Units. When I tried to give exim partial DN, the exit code
3839 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3840 But simultaneously result of request was absolutely normal ldap result,
3841 so I produce this patch..."
3843 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3844 REFERENCE, so I have modified the code to exclude the patch when that macro
3847 55. Some experimental protocols are using DNS PTR records for new purposes. The
3848 keys for these records are domain names, not reversed IP addresses. The
3849 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3850 leaves it alone. Component reversal etc. now happens only for IP addresses.
3853 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3855 57. Double the size of the debug message buffer (to 2048) so that more of very
3856 long debug lines gets shown.
3858 58. The exicyclog utility now does better if the number of log files to keep
3859 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3861 59. Two changes related to the smtp_active_hostname option:
3863 (1) $smtp_active_hostname is now available as a variable.
3864 (2) The default for smtp_banner uses $smtp_active_hostname instead
3865 of $primary_hostname.
3867 60. The host_aton() function is supposed to be passed a string that is known
3868 to be a valid IP address. However, in the case of IPv6 addresses, it was
3869 not checking this. This is a hostage to fortune. Exim now panics and dies
3870 if the condition is not met. A case was found where this could be provoked
3871 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3872 components; fortuitously, this particular loophole had already been fixed
3873 by change 4.50/55 above.
3875 If there are any other similar loopholes, the new check in host_aton()
3876 itself should stop them being exploited. The report I received stated that
3877 data on the command line could provoke the exploit when Exim was running as
3878 exim, but did not say which command line option was involved. All I could
3879 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3880 running as the user.
3883 61. There was a buffer overflow vulnerability in the SPA authentication code
3884 (which came originally from the Samba project). I have added a test to the
3885 spa_base64_to_bits() function which I hope fixes it.
3888 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3889 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3891 63. The daemon start-up calls getloadavg() while still root for those OS that
3892 need the first call to be done as root, but it missed one case: when
3893 deliver_queue_load_max is set with deliver_drop_privilege. This is
3894 necessary for the benefit of the queue runner, because there is no re-exec
3895 when deliver_drop_privilege is set.
3897 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3898 This has been fixed.
3900 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3901 was in use, was not putting the data itself into the right store pool;
3902 consequently, it could be overwritten for a subsequent message in the same
3903 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3906 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3908 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3909 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3910 the function were treating the return as a boolean value, which happened to
3911 work because 0=false and not-0=true, but is not correct code.
3913 68. The host_aton() function was not handling scoped IPv6 addresses (those
3914 with, for example, "%eth0" on the end) correctly.
3916 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3917 negated items (that is, ~something) in unsigned ints. Some compilers
3918 apparently mutter when there is no cast.
3920 70. If an address verification called from an ACL failed, and did not produce a
3921 user-specific message (i.e. there was only a "system" message), nothing was
3922 put in $acl_verify_message. In this situation, it now puts the system
3925 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3926 session; change 4.31/43 added the unwanted input to the log line - except
3927 that it did not do this in the start of session case. It now does.
3929 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3930 This is wrong and can cause the other end to generate a synchronization
3931 error if it is another Exim or anything else that does the synchronization
3932 check. A QUIT command is no longer sent after a timeout.
3934 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3935 during host lookups.
3937 74. The defer_ok option of callout verification was not working if it was used
3938 when verifying addresses in header lines, that is, for this case:
3940 verify = header_sender/callout=defer_ok
3942 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3943 those file descriptors could be used for SMTP connections. If anything
3944 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3945 could be sent to the SMTP client, causing chaos. The daemon now opens
3946 stdin, stdout, and stderr to /dev/null when it puts itself into the
3949 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3950 log by default. The user can override this with suitable Perl magic.
3952 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3953 the log message when discard triggers, was not working for the DATA ACL or
3954 for the non-SMTP ACL.
3956 78. Error message wording change in sieve.c.
3958 79. If smtp_accept_max_per_host was set, the number of connections could be
3959 restricted to fewer than expected, because the daemon was trying to set up
3960 a new connection before checking whether the processes handling previous
3961 connections had finished. The check for completed processes is now done
3962 earlier. On busy systems, this bug wouldn't be noticed because something
3963 else would have woken the daemon, and it would have reaped the completed
3966 80. If a message was submitted locally by a user whose login name contained one
3967 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3968 It caused a spool format error. I have fixed the spool reading code. A
3969 related problem was that the "from" clause in the Received: line became
3970 illegal because of the space(s). It is now covered by ${quote_local_part.
3972 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3975 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3977 83. Patch from Sieve maintainer to fix unterminated string problem in
3978 "vacation" handling.
3980 84. Some minor changes to the Linux configuration files to help with other
3981 OS variants using glibc.
3983 85. One more patch for Sieve to update vacation handling to latest spec.
3986 ----------------------------------------------------
3987 See the note above about the 4.44 and 4.50 releases.
3988 ----------------------------------------------------
3994 1. Change 4.43/35 introduced a bug that caused file counts to be
3995 incorrectly computed when quota_filecount was set in an appendfile
3998 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3999 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4002 3. Give more explanation in the error message when the command for a transport
4003 filter fails to execute.
4005 4. There are several places where Exim runs a non-Exim command in a
4006 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4007 was being done only for the command run by the queryprogram router. It is
4008 now done for all such subprocesses. The other cases are: ${run, transport
4009 filters, and the commands run by the lmtp and pipe transports.
4011 5. Some older OS have a limit of 256 on the maximum number of file
4012 descriptors. Exim was using setrlimit() to set 1000 as a large value
4013 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4014 systems. I've change it so that if it can't get 1000, it tries for 256.
4016 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4017 was an oversight, and furthermore, ever since the addition of extra
4018 controls (e.g. 4.43/32), the checks on when to allow different forms of
4019 "control" were broken. There should now be diagnostics for all cases when a
4020 control that does not make sense is encountered.
4022 7. $recipients is now available in the predata ACL (oversight).
4024 8. Tidy the search cache before the fork to do a delivery from a message
4025 received from the command line. Otherwise the child will trigger a lookup
4026 failure and thereby defer the delivery if it tries to use (for example) a
4027 cached ldap connection that the parent has called unbind on.
4029 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4030 of $address_data from the recipient verification was clobbered by the
4031 sender verification.
4033 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4034 was its contents. (It was OK if the option was not defined at all.)
4036 11. A "Completed" log line is now written for messages that are removed from
4037 the spool by the -Mrm option.
4039 12. $host_address is now set to the target address during the checking of
4040 ignore_target_hosts.
4042 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4043 being passed; this would have caused $sender_host_name to have been used if
4044 matching the list had actually called for a host name (not very likely,
4045 since this list is usually IP addresses). A host name is now passed as
4048 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4049 code that specifies a non-restarting handler (typically sigaction() in
4050 modern systems) in an attempt to fix a rare and obscure crash bug.
4052 15. Narrowed the window for a race in the daemon that could cause it to ignore
4053 SIGCHLD signals. This is not a major problem, because they are used only to
4054 wake it up if nothing else does.
4056 16. A malformed maildirsize file could cause Exim to calculate negative values
4057 for the mailbox size or file count. Odd effects could occur as a result.
4058 The maildirsize information is now recalculated if the size or filecount
4061 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4062 support for a long time. Removed HAVE_SYS_VFS_H.
4064 18. Updated exipick to current release from John Jetmore.
4066 19. Allow an empty sender to be matched against a lookup in an address list.
4067 Previously the only cases considered were a regular expression, or an
4070 20. Exim went into a mad DNS lookup loop when doing a callout where the
4071 host was specified on the transport, if the DNS lookup yielded more than
4074 21. The RFC2047 encoding function was originally intended for short strings
4075 such as real names; it was not keeping to the 75-character limit for
4076 encoded words that the RFC imposes. It now respects the limit, and
4077 generates multiple encoded words if necessary. To be on the safe side, I
4078 have increased the buffer size for the ${rfc2047: expansion operator from
4081 22. Failure to deliver a bounce message always caused it to be frozen, even if
4082 there was an errors_to setting on the router. The errors_to setting is now
4085 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4086 canonical form (fully expanded) before being placed in
4087 $sender_host_address.
4089 24. Updated eximstats to version 1.33
4091 25. Include certificate and key file names in error message when GnuTLS fails
4092 to set them up, because the GnuTLS error message doesn't include the name
4093 of the failing file when there is a problem reading it.
4095 26. Expand error message when OpenSSL has problems setting up cert/key files.
4098 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4099 (this can affect the format of dates).
4101 28. exim_tidydb, when checking for the continued existence of a message for
4102 which it has found a message-specific retry record, was not finding
4103 messages that were in split spool directories. Consequently, it was
4104 deleting retry records that should have stayed in existence.
4106 29. eximstats updated to version 1.35
4107 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4108 1.35 - bugfix such that pie charts by volume are generated correctly
4110 30. The SPA authentication driver was not abandoning authentication and moving
4111 on to the next authenticator when an expansion was forced to fail,
4112 contradicting the general specification for all authenticators. Instead it
4113 was generating a temporary error. It now behaves as specified.
4115 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4116 (the order specifies the preference for clients). The order is now AES256,
4117 AES128, 3DES, ARCFOUR128.
4119 31. Small patch to Sieve code - explicitly set From: when generating an
4122 32. Exim crashed if a remote delivery caused a very long error message to be
4123 recorded - for instance if somebody sent an entire SpamAssassin report back
4124 as a large number of 550 error lines. This bug was coincidentally fixed by
4125 increasing the size of one of Exim's internal buffers (big_buffer) that
4126 happened as part of the Exiscan merge. However, to be on the safe side, I
4127 have made the code more robust (and fixed the comments that describe what
4130 33. Some experimental protocols are using DNS PTR records for new purposes. The
4131 keys for these records are domain names, not reversed IP addresses. The
4132 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4133 leaves it alone. Component reversal etc. now happens only for IP addresses.
4136 34. The host_aton() function is supposed to be passed a string that is known
4137 to be a valid IP address. However, in the case of IPv6 addresses, it was
4138 not checking this. This is a hostage to fortune. Exim now panics and dies
4139 if the condition is not met. A case was found where this could be provoked
4140 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4141 components; fortuitously, this particular loophole had already been fixed
4142 by change 4.50/55 or 4.44/33 above.
4144 If there are any other similar loopholes, the new check in host_aton()
4145 itself should stop them being exploited. The report I received stated that
4146 data on the command line could provoke the exploit when Exim was running as
4147 exim, but did not say which command line option was involved. All I could
4148 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4149 running as the user.
4152 35. There was a buffer overflow vulnerability in the SPA authentication code
4153 (which came originally from the Samba project). I have added a test to the
4154 spa_base64_to_bits() function which I hope fixes it.
4157 36. The daemon start-up calls getloadavg() while still root for those OS that
4158 need the first call to be done as root, but it missed one case: when
4159 deliver_queue_load_max is set with deliver_drop_privilege. This is
4160 necessary for the benefit of the queue runner, because there is no re-exec
4161 when deliver_drop_privilege is set.
4163 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4164 was in use, was not putting the data itself into the right store pool;
4165 consequently, it could be overwritten for a subsequent message in the same
4166 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4169 38. Sometimes the final signoff response after QUIT could fail to get
4170 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4171 < 0 before doing a fflush(). This bug looks as though it goes back to the
4172 introduction of TLS in release 3.20, but "sometimes" must have been rare
4173 because the tests only now provoked it.
4179 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4180 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4181 Now it returns an integer. A number of calls were still expecting a T/F
4182 return. Fortuitously, in all cases, the tests worked in OK situations,
4183 which is the norm. However, things would have gone wrong on any write
4184 failures on the smtp file descriptor. This function is used when sending
4185 messages over SMTP and also when doing verify callouts.
4187 2. When Exim is called to do synchronous delivery of a locally submitted
4188 message (the -odf or -odi options), it no longer closes stderr before doing
4191 3. Implemented the mua_wrapper option.
4193 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4195 5. Implemented the functions header_remove(), header_testname(),
4196 header_add_at_position(), and receive_remove_recipient(), and exported them
4199 6. If an ACL "warn" statement specified the addition of headers, Exim already
4200 inserted X-ACL-Warn: at the start if there was no header name. However, it
4201 was not making this test for the second and subsequent header lines if
4202 there were newlines in the string. This meant that an invalid header could
4203 be inserted if Exim was badly configured.
4205 7. Allow an ACL "warn" statement to add header lines at the start or after all
4206 the Received: headers, as well as at the end.
4208 8. Added the rcpt_4xx retry error code.
4210 9. Added postmaster_mailfrom=xxx to callout verification option.
4212 10. Added mailfrom=xxxx to the callout verification option, for verify=
4215 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4216 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4218 12. Inserted some casts to stop certain compilers warning when using pointer
4219 differences as field lengths or precisions in printf-type calls (mostly
4220 affecting debugging statements).
4222 13. Added optional readline() support for -be (dynamically loaded).
4224 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4225 same clock tick as a message's arrival, so that its received time was the
4226 same as the "first fail" time on the retry record, and that message
4227 remained on the queue past the ultimate address timeout, every queue runner
4228 would try a delivery (because it was past the ultimate address timeout) but
4229 after another failure, the ultimate address timeout, which should have then
4230 bounced the address, did not kick in. This was a "< instead of <=" error;
4231 in most cases the first failure would have been in the next clock tick
4232 after the received time, and all would be well.
4234 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4235 being recognized when the domain list was tested by the match_domain
4236 condition in an expansion string.
4238 16. Added the ${str2b64: operator.
4240 17. Exim was always calling setrlimit() to set a large limit for the number of
4241 processes, without checking whether the existing limit was already
4242 adequate. (It did check for the limit on file descriptors.) Furthermore,
4243 errors from getrlimit() and setrlimit() were being ignored. Now they are
4244 logged to the main and panic logs, but Exim does carry on, to try to do its
4245 job under whatever limits there are.
4247 18. Imported PCRE 5.0.
4249 19. Trivial typo in log message " temporarily refused connection" (the leading
4252 20. If the log selector return_path_on_delivery was set and an address was
4253 redirected to /dev/null, the delivery process crashed because it assumed
4254 that a return path would always be set for a "successful" delivery. In this
4255 case, the whole delivery is bypassed as an optimization, and therefore no
4258 21. Internal re-arrangement: the function for sending a challenge and reading
4259 a response while authentication was assuming a zero-terminated challenge
4260 string. It's now changed to take a pointer and a length, to allow for
4261 binary data in such strings.
4263 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4265 23. Exim was not respecting finduser_retries when seeking the login of the
4266 uid under which it was called; it was always trying 10 times. (The default
4267 setting of finduser_retries is zero.) Also, it was sleeping after the final
4268 failure, which is pointless.
4270 24. Implemented tls_on_connect_ports.
4272 25. Implemented acl_smtp_predata.
4274 26. If the domain in control=submission is set empty, Exim assumes that the
4275 authenticated id is a complete email address when it generates From: or
4276 Sender: header lines.
4278 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4279 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4280 chown and chgrp in /bin and hostname in /usr/bin.
4282 28. Exim was keeping the "process log" file open after each use, just as it
4283 does for the main log. This opens the possibility of it remaining open for
4284 long periods when the USR1 signal hits a daemon. Occasional processlog
4285 errors were reported, that could have been caused by this. Anyway, it seems
4286 much more sensible not to leave this file open at all, so that is what now
4289 29. The long-running daemon process does not normally write to the log once it
4290 has entered its main loop, and it closes the log before doing so. This is
4291 so that log files can straightforwardly be renamed and moved. However,
4292 there are a couple of unusual error situations where the daemon does write
4293 log entries, and I had neglected to close the log afterwards.
4295 30. The text of an SMTP error response that was received during a remote
4296 delivery was being truncated at 512 bytes. This is too short for some of
4297 the long messages that one sometimes sees. I've increased the limit to
4300 31. It is now possible to make retry rules that apply only when a message has a
4301 specific sender, in particular, an empty sender.
4303 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4304 it possible to be selective about when SMTP synchronization is enforced.
4306 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4308 32. Implemented hosts_connection_nolog.
4310 33. Added an ACL for QUIT.
4312 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4315 35. Added mailbox_size and mailbox_filecount to appendfile.
4317 36. Added control = no_multiline_responses to ACLs.
4319 37. There was a bug in the logic of the code that waits for the clock to tick
4320 in the case where the clock went backwards by a substantial amount such
4321 that the microsecond fraction of "now" was more than the microsecond
4322 fraction of "then" (but the whole seconds number was less).
4324 38. Added support for the libradius Radius client library this is found on
4325 FreeBSD (previously only the radiusclient library was supported).
4331 1. When certain lookups returned multiple values in the form name=value, the
4332 quoting of the values was not always being done properly. Specifically:
4333 (a) If the value started with a double quote, but contained no whitespace,
4335 (b) If the value contained whitespace other than a space character (i.e.
4336 tabs or newlines or carriage returns) it was not quoted.
4337 This fix has been applied to the mysql and pgsql lookups by writing a
4338 separate quoting function and calling it from the lookup code. The fix
4339 should probably also be applied to nisplus, ibase and oracle lookups, but
4340 since I cannot test any of those, I have not disturbed their existing code.
4342 2. A hit in the callout cache for a specific address caused a log line with no
4343 reason for rejecting RCPT. Now it says "Previous (cached) callout
4344 verification failure".
4346 3. There was an off-by-one bug in the queryprogram router. An over-long
4347 return line was truncated at 256 instead of 255 characters, thereby
4348 overflowing its buffer with the terminating zero. As well as fixing this, I
4349 have increased the buffer size to 1024 (and made a note to document this).
4351 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4352 when Exim is waiting for an SMTP response from a remote server, Exim
4353 restarts its select() call on the socket, thereby resetting its timeout.
4354 This is not a problem when such interrupts are rare. Somebody set up a cron
4355 job to run exiwhat every 2 minutes, which is less than the normal select()
4356 timeout (5 or 10 minutes). This meant that the select() timeout never
4357 kicked in because it was always reset. I have fixed this by comparing the
4358 time when an interrupt arrives with the time at the start of the first call
4359 to select(). If more time than the timeout has elapsed, the interrupt is
4360 treated as a timeout.
4362 5. Some internal re-factoring in preparation for the addition of Sieve
4363 extensions (by MH). In particular, the "personal" test is moved to a
4364 separate function, and given an option for scanning Cc: and Bcc: (which is
4365 not set for Exim filters).
4367 6. When Exim created an email address using the login of the caller as the
4368 local part (e.g. when creating a From: or Sender: header line), it was not
4369 quoting the local part when it contained special characters such as @.
4371 7. Installed new OpenBSD configuration files.
4373 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4374 try to make them clearer.
4376 9. Callout options, other than the timeout value, were being ignored when
4377 verifying sender addresses in header lines. For example, when using
4379 verify = header_sender/callout=no_cache
4381 the cache was (incorrectly) being used.
4383 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4384 only the Cygwin environment.
4386 11. When return_path_on_delivery was set as a log selector, if different remote
4387 addresses in the same message used different return paths and parallel
4388 remote delivery occurred, the wrong values would sometimes be logged.
4389 (Whenever a remote delivery process finished, the return path value from
4390 the most recently started remote delivery process was logged.)
4392 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4393 lines when AUTH and/or TLS are in use. This is the "received protocol"
4394 field. Exim used to use "asmtp" for authenticated SMTP, without any
4395 indication (in the protocol name) for TLS use. Now it follows the RFC and
4396 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4397 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4398 names appear in log lines as well as in Received: header lines.
4400 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4401 extensions, and comparison tests, and to fix some bugs.
4403 14. Changes to the "personal" filter test:
4405 (1) The test was buggy in that it was just doing the equivalent of
4406 "contains" tests on header lines. For example, if a user's address was
4407 anne@some.where, the "personal" test would incorrectly be true for
4409 To: susanne@some.where
4411 This test is now done by extracting each address from the header in turn,
4412 and checking the entire address. Other tests that are part of "personal"
4413 are now done using regular expressions (for example, to check local parts
4414 of addresses in From: header lines).
4416 (2) The list of non-personal local parts in From: addresses has been
4417 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4418 taken from the Sieve specification recommendations.
4420 (3) If the message contains any header line starting with "List-" it is
4421 treated as non-personal.
4423 (4) The test for "circular" in the Subject: header line has been removed
4424 because it now seems ill-conceived.
4426 15. Minor typos in src/EDITME comments corrected.
4428 16. Installed latest exipick from John Jetmore.
4430 17. If headers_add on a router specified a text string that was too long for
4431 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4432 of string_sprintf() is now avoided.
4434 18. $message_body_size was not set (it was always zero) when running the DATA
4435 ACL and the local_scan() function.
4437 19. For the "mail" command in an Exim filter, no default was being set for
4438 the once_repeat time, causing a random time value to be used if "once" was
4439 specified. (If the value happened to be <= 0, no repeat happened.) The
4440 default is now 0s, meaning "never repeat". The "vacation" command was OK
4441 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4442 (I found it when inspecting the code).
4444 20. There is now an overall timeout for performing a callout verification. It
4445 defaults to 4 times the callout timeout, which applies to individual SMTP
4446 commands during the callout. The overall timeout applies when there is more
4447 than one host that can be tried. The timeout is checked before trying the
4448 next host. This prevents very long delays if there are a large number of
4449 hosts and all are timing out (e.g. when the network connections are timing
4450 out). The value of the overall timeout can be changed by specifying an
4451 additional sub-option for "callout", called "maxwait". For example:
4453 verify = sender/callout=5s,maxwait=20s
4455 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4456 to the end before writing, but this should make it even safer).
4458 22. Exim was forgetting that it had advertised PIPELINING for the second and
4459 subsequent messages on an SMTP connection. It was also not resetting its
4460 memory on STARTTLS and an internal HELO.
4462 23. When Exim logs an SMTP synchronization error within a session, it now
4463 records whether PIPELINING has been advertised or not.
4465 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4466 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4467 rather than long int.
4469 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4471 26. Added the never_mail option to autoreply.
4477 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4478 crash if the getsockname() call failed; this can happen if a connection is
4479 closed very soon after it is established. The problem was simply in the
4480 order in which certain operations were done, causing Exim to try to write
4481 to the SMTP stream before it had set up the file descriptor. The bug has
4482 been fixed by making things happen in the correct order.
4488 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4489 before the connection was closed, thus losing the rejection response.
4491 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4492 some early Solaris releases, but causes trouble in current releases where
4493 socklen_t is defined.
4495 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4498 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4501 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4502 that releasing the top of it at the end releases what was used for sub-
4503 expansions (unless the block got too big). However, discard this block if
4504 the first thing is a variable or header, so that we can use its block when
4505 it is dynamic (useful for very large $message_headers, for example).
4507 6. Lookups now cache *every* query, not just the most recent. A new, separate
4508 store pool is used for this. It can be recovered when all lookup caches are
4509 flushed. Lookups now release memory at the end of their result strings.
4510 This has involved some general refactoring of the lookup sources.
4512 7. Some code has been added to the store_xxx() functions to reduce the amount
4513 of flapping under certain conditions.
4515 8. log_incoming_interface used to affect only the <= reception log lines. Now
4516 it causes the local interface and port to be added to several more SMTP log
4517 lines, for example "SMTP connection from", and rejection lines.
4519 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4521 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4523 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4524 could be overwritten at the end of the current message (or the start of a
4525 new message if it was set in a HELO ACL). The value is now preserved for
4526 the duration of the SMTP connection.
4528 12. If a transport had a headers_rewrite setting, and a matching header line
4529 contained an unqualified address, that address was qualified, even if it
4530 did not match any rewriting rules. The underlying bug was that the values
4531 of the flags that permit the existence of unqualified sender and recipient
4532 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4533 non-local messages, and by -bnq for local messages) were not being
4534 preserved with the message after it was received.
4536 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4537 "next input=" as part of the text comprising the host identity instead of
4538 the correct text. The code was using the same buffer for two different
4539 strings. However, depending on which order the printing function evaluated
4540 its arguments, the bug did not always show up. Under Linux, for example, my
4541 test suite worked just fine.
4543 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4544 doesn't work with some older versions of Perl. It has been changed to "my",
4545 which in any case is probably the better facility to use.
4547 15. A really picky compiler found some instances of statements for creating
4548 error messages that either had too many or two few arguments for the format
4551 16. The size of the buffer for calls to the DNS resolver has been increased
4552 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4553 for addresses that have a lot of PTR records. This alleviates a problem; it
4554 does not fully solve it.
4556 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4557 buffer now truncates the list and logs the incident, which is the same
4558 action as happens when Exim is looking up a host name and its aliases.
4559 Previously in this situation something unpredictable would happen;
4560 sometimes it was "internal error: store_reset failed".
4562 18. If a server dropped the connection unexpectedly when an Exim client was
4563 using GnuTLS and trying to read a response, the client delivery process
4564 crashed while trying to generate an error log message.
4566 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4567 string, for example:
4569 warn message = H1: something\nH2: something
4571 the text was added as a single header line from Exim's point of view
4572 though it ended up OK in the delivered message. However, searching for the
4573 second and subsequent header lines using $h_h2: did not work. This has been
4574 fixed. Similarly, if a system filter added multiple headers in this way,
4575 the routers could not see them.
4577 20. Expanded the error message when iplsearch is called with an invalid key to
4578 suggest using net-iplsearch in a host list.
4580 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4581 ACLs are no longer actually imposed (and a message to that effect is
4584 22. If a "gecos" field in a passwd entry contained escaped characters, in
4585 particular, if it contained a \" sequence, Exim got it wrong when building
4586 a From: or a Sender: header from that name. A second bug also caused
4587 incorrect handling when an unquoted " was present following a character
4588 that needed quoting.
4590 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4591 was not being matched caselessly.
4593 24. Arranged for all hyphens in the exim.8 source to be escaped with
4596 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4597 properly. Recipient callout cache records were still being keyed to include
4598 the sender, even when use_sender was set false. This led to far more
4599 callouts that were necessary. The sender is no longer included in the key
4600 when use_sender is false.
4602 26. Added "control = submission" modifier to ACLs.
4604 27. Added the ${base62d: operator to decode base 62 numbers.
4606 28. dnsdb lookups can now access SRV records.
4608 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4609 the configuration file.
4611 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4612 (-v) mode. This makes the output for a verbose queue run more intelligible.
4614 31. Added a use_postmaster feature to recipient callouts.
4616 32. Added the $body_zerocount variable, containing the number of binary zero
4617 bytes in the message body.
4619 33. The time of last modification of the "new" subdirectory is now used as the
4620 "mailbox time last read" when there is a quota error for a maildir
4623 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4625 35. Added +ignore_unknown as a special item in host lists.
4627 36. Code for decoding IPv6 addresses in host lists is now included, even if
4628 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4629 address was recognized as an IP address, but was then not correctly decoded
4630 into binary, causing unexpected and incorrect effects when compared with
4637 1. Very minor rewording of debugging text in manualroute to say "list of
4638 hosts" instead of "hostlist".
4640 2. If verify=header_syntax was set, and a header line with an unqualified
4641 address (no domain) and a large number of spaces between the end of the
4642 name and the colon was received, the reception process suffered a buffer
4643 overflow, and (when I tested it) crashed. This was caused by some obsolete
4644 code that should have been removed. The fix is to remove it!
4646 3. When running in the test harness, delay a bit after writing a bounce
4647 message to get a bit more predictability in the log output.
4649 4. Added a call to search_tidyup() just before forking a reception process. In
4650 theory, someone could use a lookup in the expansion of smtp_accept_max_
4651 per_host which, without the tidyup, could leave open a database connection.
4653 5. Added the variables $recipient_data and $sender_data which get set from a
4654 lookup success in an ACL "recipients" or "senders" condition, or a router
4655 "senders" option, similar to $domain_data and $local_part_data.
4657 6. Moved the writing of debug_print from before to after the "senders" test
4660 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4661 problems for message scanning, either using a data ACL, or using
4662 local_scan() because the Received: header was not generated till after they
4663 were called (in order to set the time as the time of reception completion).
4664 I have revised the way this works. The header is now generated after the
4665 body is received, but before the ACL or local_scan() are called. After they
4666 are run, the timestamp in the header is updated.
4672 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4673 before starting a queue runner without re-exec. This happened only when
4674 deliver_drop_privilege was set or when the Exim user was set to root. The
4675 effect of the bug was that timeouts during subsequent deliveries caused
4676 crashes instead of being properly handled. The handler is now left at its
4677 default (and expected) setting.
4679 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4680 message, again when deliver_drop_privilege is set or Exim is run as root.
4681 The bug described in (1) was not present in this case, but the tidying up
4682 of the other signals was missing. I have made the two cases consistent.
4684 3. The ignore_target_hosts setting on a manualroute router was being ignored
4685 for hosts that were looked up using the /MX notation.
4687 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4690 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4691 operated on the sender address. After changing the $sender_address to <>
4692 for the sender address verify, Exim was re-instated it as the original
4693 (before rewriting) address, but remembering that it had rewritten it, so it
4694 wasn't rewriting it again. This bug also had the effect of breaking the
4695 sender address verification caching when the sender address was rewritten.
4697 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4698 This has been changed so that if the ip literal address matches
4699 ignore_target_hosts, the router declines.
4701 7. Added expansion conditions match_domain, match_address, and match_local_
4702 part (NOT match_host).
4704 8. The placeholder for the Received: header didn't have a length field set.
4706 9. Added code to Exim itself and to exim_lock to test for a specific race
4707 condition that could lead to file corruption when using MBX delivery. The
4708 issue is with the lockfile that is created in /tmp. If this file is removed
4709 after a process has opened it but before that process has acquired a lock,
4710 there is the potential for a second process to recreate the file and also
4711 acquire a lock. This could lead to two Exim processes writing to the file
4712 at the same time. The added code performs the same test as UW imapd; it
4713 checks after acquiring the lock that its file descriptor still refers to
4714 the same named file.
4716 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4717 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4720 11. Added the smtp_active_hostname option. If used, this will typically be made
4721 to depend on the incoming interface address. Because $interface_address is
4722 not set up until the daemon has forked a reception process, error responses
4723 that can happen earlier (such as "too many connections") no longer contain
4726 12. If an expansion in a condition on a "warn" statement fails because a lookup
4727 defers, the "warn" statement is abandoned, and the next ACL statement is
4728 processed. Previously this caused the whole ACL to be aborted.
4730 13. Added the iplsearch lookup type.
4732 14. Added ident_timeout as a log selector.
4734 15. Added tls_certificate_verified as a log selector.
4736 16. Added a global option tls_require_ciphers (compare the smtp transport
4737 option of the same name). This controls incoming TLS connections.
4739 17. I finally figured out how to make tls_require_ciphers do a similar thing
4740 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4741 before starting the TLS session.
4743 18. Tabs are now shown as \t in -bP output.
4745 19. If the log selector return_path_on_delivery was set, Exim crashed when
4746 bouncing a message because it had too many Received: header lines.
4748 20. If two routers both had headers_remove settings, and the first one included
4749 a superfluous trailing colon, the final name in the first list and the
4750 first name in the second list were incorrectly joined into one item (with a
4751 colon in the middle).
4757 1. Added -C and -D options to the exinext utility, mainly to make it easier
4758 to include in the automated testing, but these could be helpful when
4759 multiple configurations are in use.
4761 2. The exinext utility was not formatting the output nicely when there was
4762 an alternate port involved in the retry record key, nor when there was a
4763 message id as well (for retries that were specific to a specific message
4764 and a specific host). It was also confused by IPv6 addresses, because of
4765 the additional colons they contain. I have fixed the IPv4 problem, and
4766 patched it up to do a reasonable job for IPv6.
4768 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4769 delivery, the log line now contains "pipelined" if PIPELINING was used.
4771 4. An SMTP transport process used to panic and die if the bind() call to set
4772 an explicit outgoing interface failed. This has been changed; it is now
4773 treated in the same way as a connect() failure.
4775 5. A reference to $sender_host_name in the part of a conditional expansion
4776 that was being skipped was still causing a DNS lookup. This no longer
4779 6. The def: expansion condition was not recognizing references to header lines
4780 that used bh_ and bheader_.
4782 7. Added the _cache feature to named lists.
4784 8. The code for checking quota_filecount in the appendfile transport was
4785 allowing one more file than it should have been.
4787 9. For compatibility with Sendmail, the command line option
4795 and sets the incoming protocol and host name (for trusted callers). The
4796 host name and its colon can be omitted when only the protocol is to be set.
4797 Note the Exim already has two private options, -pd and -ps, that refer to
4798 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4799 "s", but I don't think that's a major issue.
4801 10. A number of refactoring changes to the code, none of which should affect
4804 (a) The number of logging options was getting close to filling up the
4805 32-bit word that was used as a bit map. I have split them into two classes:
4806 those that are passed in the argument to log_write(), and those that are
4807 only ever tested independently outside of that function. These are now in
4808 separate 32-bit words, so there is plenty of room for expansion again.
4809 There is no change in the user interface or the logging behaviour.
4811 (b) When building, for example, log lines, the code previously used a
4812 macro that called string_cat() twice, in order to add two strings. This is
4813 not really sufficiently general. Furthermore, there was one instance where
4814 it was actually wrong because one of the argument was used twice, and in
4815 one call a function was used. (As it happened, calling the function twice
4816 did not affect the overall behaviour.) The macro has been replaced by a
4817 function that can join an arbitrary number of extra strings onto a growing
4820 (c) The code for expansion conditions now uses a table and a binary chop
4821 instead of a serial search (which was left over from when there were very
4822 few conditions). Also, it now recognizes conditions like "pam" even when
4823 the relevant support is not compiled in: a suitably worded error message is
4824 given if an attempt is made to use such a condition.
4826 11. Added ${time_interval:xxxxx}.
4828 12. A bug was causing one of the ddress fields not to be passed back correctly
4829 from remote delivery subprocesses. The field in question was not being
4830 subsequently used, so this caused to problems in practice.
4832 13. Added new log selectors queue_time and deliver_time.
4834 14. Might have fixed a bug in maildirsizefile handling that threw up
4835 "unexpected character" debug warnings, and recalculated the data
4836 unnecessarily. In any case, I expanded the warning message to give more
4839 15. Added the message "Restricted characters in address" to the statements in
4840 the default ACL that block characters like @ and % in local parts.
4842 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4843 Three changes have been made:
4845 (a) There was a serious bug; a negative response to MAIL caused the whole
4846 recipient domain to be cached as invalid, thereby blocking all messages
4847 to all local parts at the same domain, from all senders. This bug has
4848 been fixed. The domain is no longer cached after a negative response to
4849 MAIL if the sender used is not empty.
4851 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4854 (c) A new callout option, "use_sender" has been added for people who want
4855 the modified behaviour.
4861 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4864 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4865 indeed breaks things for older releases.
4867 3. Added additional logging to the case where there is a problem reading data
4868 from a filter that is running in a subprocess using a pipe, in order to
4869 try to track down a specific problem.
4871 4. Testing facility fudge: when running in the test harness and attempting
4872 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4873 getting "No route to host". Convert this to a timeout.
4875 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4878 6. Some OS don't have socklen_t but use size_t instead. This affects the
4879 fifth argument of getsockopt() amongst other things. This is now
4880 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4881 can be set for individual OS. I have set it for SunOS5, OSF1, and
4882 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4883 some earlier ones do not.
4885 7. Change 4.30/15 was not doing the test caselessly.
4887 8. The standard form for an IPv6 address literal was being rejected by address
4888 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4889 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4890 this, as well as the form without the "IPv6" on the front (but only when
4891 address literals are enabled, of course).
4893 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4895 10. Exim crashed if a message with an empty sender address specified by -f
4896 encountered a router with an errors_to setting. This could be provoked only
4897 by a command such as
4901 where an empty string was supplied; "<>" did not hit this bug.
4903 11. Installed PCRE release 4.5.
4905 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4906 remained set. It is now erased.
4908 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4909 times from message ids (which are base 36 rather than the normal 62).
4911 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4912 were being counted as actual protocol errors, and logged if the log
4913 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4914 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4915 and DATA following a set of rejected RCPTs do not count as protocol errors.
4916 In other words, Exim assumes they were pipelined, though this may not
4917 actually be the case. Of course, in all cases the client gets an
4918 appropriate error code.
4920 15. If a lookup fails in an ACL condition, a message about the failure may
4921 be available; it is used if testing the ACL cannot continue, because most
4922 such messages specify what the cause of the deferral is. However, some
4923 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4924 that caused an old message to be retained and used if a later statement
4925 caused a defer, replacing the real cause of the deferral.
4927 16. If an IP address had so many PTR records that the DNS lookup buffer
4928 was not large enough to hold them, Exim could crash while trying to process
4929 the truncated data. It now detects and logs this case.
4931 17. Further to 4.21/58, another change has been made: if (and only if) the
4932 first line of a message (the first header line) ends with CRLF, a bare LF
4933 in a subsequent header line has a space inserted after it, so as not to
4934 terminate the header.
4936 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4937 unnecessarily, used atoi() instead of strtol(), and didn't check the
4938 termination when getting file sizes from file names by regex.
4940 19. Completely re-implemented the support for maildirsize files, in the light
4941 of a number of problems with the previous contributed implementation
4942 (4.30/29). In particular:
4944 . If the quota is zero, the maildirsize file is maintained, but no quota is
4947 . If the maildir directory does not exist, it is created before any attempt
4948 to write a maildirsize file.
4950 . The quota value in the file is just a cache; if the quota is changed in
4951 the transport, the new value overrides.
4953 . A regular expression is available for excluding directories from the
4956 20. The autoreply transport checks the characters in options that define the
4957 message's headers; it allows continued headers, but it was checking with
4958 isspace() after an embedded newline instead of explicitly looking for a
4961 21. If all the "regular" hosts to which an address was routed had passed their
4962 expiry times, and had not reached their retry times, the address was
4963 bounced, even if fallback hosts were defined. Now Exim should go on to try
4966 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4967 equivalent code in the SMTP transport. Some hosts send humungous responses
4968 to HELO/EHLO, more than 1024 it seems.
4970 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4971 gives compiler warnings in some environments. I've now done it "properly",
4974 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4975 (because of problems with the built-in one) was declared to return uschar *
4976 instead of char *, causing compiler failure.
4978 25. Fixed a file descriptor leak when processing alias/forward files.
4980 26. Fixed a minor format string issue in dbfn.c.
4982 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4984 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4985 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4988 29. When a sender address is verified, it is cached, to save repeating the test
4989 when there is more than one recipient in a message. However, when the
4990 verification involves a callout, it is possible for different callout
4991 options to be set for different recipients. It is too complicated to keep
4992 track of this in the cache, so now Exim always runs a verification when a
4993 callout is required, relying on the callout cache for the optimization.
4994 The overhead is duplication of the address routing, but this should not be
4997 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4998 to be verified with callout=postmaster, and the main callout worked but the
4999 postmaster check failed, the verification correctly failed. However, if a
5000 subsequent RCPT command asked for sender verification *without* the
5001 postmaster check, incorrect caching caused this verification also to fail,
5004 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5005 it was not caching the DNS options (qualify_single, search_parents) that
5006 were used when the lookup failed. A subsequent lookup with different
5007 options therefore always gave the same answer, though there were cases
5008 where it should not have. (Example: a "domains = !$mx_any" option on a
5009 dnslookup router: the "domains" option is always processed without any
5010 widening, but the router might have qualify_single set.) Now Exim uses the
5011 cached value only when the same options are set.
5013 32. Added John Jetmore's "exipick" utility to the distribution.
5015 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5016 than a timeout (e.g. a certificate is required, and is not provided), an
5017 Exim server now closes the connection immediately. Previously it waited for
5018 the client to close - but if the client is SSL, it seems that they each
5019 wait for each other, leading to a delay before one of them times out.
5021 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5022 maintained 0.8.x compatibility because I don't think many are using it, and
5023 it is clearly obsolete.
5025 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5028 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5029 client certificate was expired. A simple patch fixes this, though I don't
5030 understand the full logic of why the verify callback is called multiple
5033 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5034 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5035 which causes problems with some clients (such as the Certicom SSL Plus
5036 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5037 disables the coutermeasure allowing Eudora to connect."
5039 38. Exim was not checking that a write() to a log file succeeded. This could
5040 lead to Bad Things if a log got too big, in particular if it hit a file
5041 size limit. Exim now panics and dies if it cannot write to a log file, just
5042 as it does if it cannot open a log file.
5044 39. Modified OS/Makefile-Linux so that it now contains
5046 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5048 The two -D definitions ensure that Exim is compiled with large file
5049 support, which makes it possible to handle log files that are bigger than
5052 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5053 instance) a domain was checked against a named list that involved a lookup,
5054 causing $domain_data to be set, then another domain was checked against the
5055 same list, then the first domain was re-checked, the value of $domain_data
5056 after the final check could be wrong. In particular, if the second check
5057 failed, it could be set empty. This bug probably also applied to
5060 41. The strip_trailing_dot option was not being applied to the address given
5061 with the -f command-line option.
5063 42. The code for reading a message's header from the spool was incrementing
5064 $received_count, but never initializing it. This meant that the value was
5065 incorrect (doubled) while delivering a message in the same process in which
5066 it was received. In the most common configuration of Exim, this never
5067 happens - a fresh exec is done - but it can happen when
5068 deliver_drop_privilege is set.
5070 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5071 it now includes up to 150 characters of the unexpected data in the log
5074 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5075 and building data strings. The size of both of these buffers was 10 000
5076 bytes - far larger than anybody would *ever* want, thought I. Needless to
5077 say, somebody hit the limit. I have increased the maximum line length to
5078 20 000 and the maximum data length of concatenated lines to 100 000. I have
5079 also fixed two bugs, because there was no checking on these buffers. Tsk,
5080 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5081 buffer is too small.
5083 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5084 lsearch lookups. Now it does.
5086 46. When parsing a route_list item in a manualroute router, a fixed-length
5087 buffer was used for the list of hosts. I made this 1024 bytes long,
5088 thinking that nobody would ever have a list of hosts that long. Wrong.
5089 Somebody had a whole pile of complicated expansion conditions, and the
5090 string was silently truncated, leading to an expansion error. It turns out
5091 that it is easier to change to an unlimited length (owing to other changes
5092 that have happened since this code was originally written) than to build
5093 structure for giving a limitation error. The length of the item that
5094 expands into the list of hosts is now unlimited.
5096 47. The lsearch lookup could not handle data where the length of text line was
5097 more than 4095 characters. Such lines were truncated, leading to shortened
5098 data being returned. It should now handle lines of any length.
5100 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5101 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5104 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5105 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5106 no longer happens in comment lines. A list of replacements is now placed
5107 at the head of all of the source files, except those whose only change is
5108 to replace PERL_COMMAND in the very first #! line.
5110 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5111 messages on the queue, with a bottom-up merge sort, using code contributed
5112 by Michael Haardt. This should make operations like -bp somewhat faster on
5113 large queues. It won't affect queue runners, except when queue_run_in_order
5116 51. Installed eximstats 1.31 in the distribution.
5118 52. Added support for SRV lookups to the dnslookup router.
5120 53. If an ACL referred to $message_body or $message_body_end, the value was not
5121 reset for any messages that followed in the same SMTP session.
5123 54. The store-handling optimization for building very long strings was not
5124 differentiating between the different store pools. I don't think this
5125 actually made any difference in practice, but I've tidied it.
5127 55. While running the routers to verify a sender address, $sender_address
5128 was still set to the sender address. This is wrong, because when routing to
5129 send a bounce to the sender, it would be empty. Therefore, I have changed
5130 it so that, while verifying a sender address, $sender_address is set to <>.
5131 (There is no change to what happens when verifying a recipient address.)
5133 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5134 target A or AAAA records (if not already returned) without resetting the
5135 qualify_single or search_parents options of the DNS resolver. These are
5136 inappropriate in this case because the targets of MX and SRV records must
5137 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5138 target that, when qualified, matched something in the local domain. These
5139 two options are now turned off when doing these lookups.
5141 57. It seems that at least some releases of Reiserfs (which does not have the
5142 concept of a fixed number of inodes) returns zero and not -1 for the
5143 number of available inodes. This interacted badly with check_spool_inodes,
5144 which assumed that -1 was the "no such thing" setting. What I have done is
5145 to check that the total number of inodes is greater than zero before doing
5146 the test of how many are available.
5148 58. When a "warn" ACL statement has a log_message modifier, the message is
5149 remembered, and not repeated. This is to avoid a lot of repetition when a
5150 message has many recipients that cause the same warning to be written.
5151 Howewer, Exim was preserving the list of already written lines for an
5152 entire SMTP session, which doesn't seem right. The memory is now reset if a
5153 new message is started.
5155 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5156 debugging output unless log_rewrite was also set.
5158 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5159 of (int)(handle) when we know that handle contains (void *)(-1).
5161 61. The Exim daemon panic-logs an error return when it closes the incoming
5162 connection. However "connection reset by peer" seems to be common, and
5163 isn't really an error worthy of noting specially, so that particular error
5166 62. When Exim is trying to find all the local interfaces, it used to panic and
5167 die if the ioctl to get the interface flags failed. However, it seems that
5168 on at least one OS (Solaris 9) it is possible to have an interface that is
5169 included in the list of interfaces, but for which you get a failure error
5170 for this call. This happens when the interface is not "plumbed" into a
5171 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5172 failure of the "get flags" call assumes that the interface is down.
5174 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5175 makes life easier for people who are doing arithmetic on fields extracted
5176 from dates, where you often get leading zeros that should not be
5177 interpreted as octal.
5179 64. Added qualify_domain to the redirect router, to override the global
5182 65. If a pathologically long header line contained very many addresses (the
5183 report of this problem mentioned 10 000) and each of them was rewritten,
5184 Exim could use up a very large amount of memory. (It kept on making new
5185 copies of the header line as it rewrote, and never released the old ones.)
5186 At the expense of a bit more processing, the header rewriting function has
5187 been changed so that it no longer eats memory in this way.
5189 66. The generation of the Received: header has been moved from the time that a
5190 message starts to be received, to the time that it finishes. The timestamp
5191 in the Received: header should now be very close to that of the <= log
5192 line. There are two side-effects of this change:
5194 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5195 logged header lines no longer include the local Received: line, because
5196 it has not yet been created. The same applies to a copy of the message
5197 that is returned to a non-SMTP sender when a message is rejected.
5199 (b) When a filter file is tested using -bf, no additional Received: header
5200 is added to the test message. After some thought, I decided that this
5203 This change does not affect the value of $received_for. It is still set
5204 after address rewriting, but before local_scan() is called.
5206 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5208 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5209 gave an unhelpful panic error message, and a defer error. I have managed to
5210 change this behaviour so that it now rejects any supplied certificate,
5211 which seems right, as the list of acceptable certificates is empty.
5213 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5214 gave an unhelpful defer error. I have not managed to make this reject any
5215 supplied certificates, but the error message it gives is "no certificate
5216 supplied", which is not helpful.
5218 70. exigrep's output now also includes lines that are not associated with any
5219 message, but which match the given pattern. Implemented by a patch from
5220 Martin Sluka, which also tidied up the Perl a bit.
5222 71. Recipient callout verification, like sender verification, was using <> in
5223 the MAIL FROM command. This isn't really the right thing, since the actual
5224 sender may affect whether the remote host accepts the recipient or not. I
5225 have changed it to use the actual sender in the callout; this means that
5226 the cache record is now keyed on a recipient/sender pair, not just the
5227 recipient address. There doesn't seem to be a real danger of callout loops,
5228 since a callout by the remote host to check the sender would use <>.
5229 [SEE ABOVE: changed after hitting problems.]
5231 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5232 temporary errors. However, in the case of such a code being given after
5233 the end of a data transmission (i.e. after ".") Exim was failing to write
5234 a retry record for the message. (Yes, there was some broken host that was
5235 actually sending 8xx at this point.)
5237 73. An unknown lookup type in a host list could cause Exim to panic-die when
5238 the list was checked. (An example that provoked this was putting <; in the
5239 middle of a list instead of at the start.) If this happened during a DATA
5240 ACL check, a -D file could be left lying around. This kind of configuration
5241 error no longer causes Exim to die; instead it causes a defer errror. The
5242 incident is still logged to the main and panic logs.
5244 74. Buglet left over from Exim 3 conversion. The message "too many messages
5245 in one connection" was written to the rejectlog but not the mainlog, except
5246 when address rewriting (yes!) was being logged.
5248 75. Added write_rejectlog option.
5250 76. When a system filter was run not as root (that is, when system_filter_user
5251 was set), the values of the $n variables were not being returned to the
5252 main process; thus, they were not subsequently available in the $sn
5255 77. Added +return_path_on_delivery log selector.
5257 78. A connection timeout was being treated differently from recipients deferred
5258 when testing hosts_max_try with a message that was older than the host's
5259 retry timeout. (The host should not be counted, thus allowing all hosts to
5260 be tried at least once before bouncing.) This may have been the cause of an
5261 occasionally reported bug whereby a message would remain on the queue
5262 longer than the retry timeout, but would be bounced if a delivery was
5263 forced. I say "may" because I never totally pinned down the problem;
5264 setting up timeout/retry tests is difficult. See also the next item.
5266 79. The ultimate address timeout was not being applied to errors that involved
5267 a combination of host plus message (for example, a timeout on a MAIL
5268 command). When an address resolved to a number of possible hosts, and they
5269 were not all tried for each delivery (e.g. because of hosts_max_try), a
5270 message could remain on the queue longer than the retry timeout.
5272 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5275 81. Fixed an obscure SMTP outgoing bug which required at least the following
5276 conditions: (a) there was another message waiting for the same server;
5277 (b) the server returned 5xx to all RCPT commands in the first message so
5278 that the message was not completed; (c) the server dropped the connection
5279 or gave a negative response to the RSET that Exim sends to abort the
5280 transaction. The observed case was a dropped connection after DATA that had
5281 been sent in pipelining mode. That is, the server had advertised PIPELINING
5282 but was not implementing it correctly. The effect of the bug was incorrect
5283 behaviour, such as trying another host, and this could lead to a crash.
5289 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5290 and daemon.c were passed as pointers to ints; they should have been
5291 pointers to socklen_t variables (which are typically unsigned ints).
5293 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5296 3. Fixed a really odd bug that affected only the testing scheme; patching a
5297 certain fixed string in the binary changed the value of another string that
5298 happened to be identical to the end of the original first string.
5300 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5301 name", it returns that address as the IP address. On some operating
5302 systems (e.g. Solaris), it also passes back the IP address string as the
5303 "host name". However, on others (e.g. Linux), it passes back an empty
5304 string. Exim wasn't checking for this, and was changing the host name to an
5305 empty string, assuming it had been canonicized.
5307 5. Although rare, it is permitted to have more than one PTR record for a given
5308 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5309 all the names associated with an address, because they do in Solaris.
5310 However, it seems that they do not in Linux for data that comes from the
5311 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5312 I found this out when I moved to a new Linux workstation and tried to run
5313 the Exim test suite.
5315 To get round this problem I have changed the code so that it now does its
5316 own call to the DNS to look up PTR records when searching for a host name.
5317 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5318 addresses that are only in /etc/hosts are still found.
5320 This behaviour is, however, controlled by an option called host_lookup_
5321 order, which defaults to "bydns:byaddr". If people want to use the other
5322 order, or indeed, just use one or the other means of lookup, they can
5323 specify it in this variable.
5325 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5326 some operating systems, this comes back from gethostbyaddr() as an empty
5327 string, and this is what Exim used to test for. However, it seems that in
5328 other systems, "." is yielded. Exim now tests for this case too.
5330 7. The values of check_spool_space and check_log_space are now held internally
5331 as a number of kilobytes instead of an absolute number of bytes. If a
5332 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5333 kilobyte. This means that much larger values can be stored.
5335 8. Exim monitor: an attempt to get the action menu when not actually pointing
5336 at a message produces an empty menu entitled "No message selected". This
5337 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5338 no entries in it ("Shell widget menu has zero width and/or height"). So I
5339 have added a single, blank menu entry in this case.
5341 9. Added ${quote_local_part.
5343 10. MIME decoding is now applied to the contents of Subject: header lines when
5346 11. Now that a reference to $sender_host_address automatically causes a reverse
5347 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5348 host lookup before query-style lookups in lists that might use this
5349 variable. This has therefore been abolished, and the "net-" prefix is no
5350 longer necessary for query-style lookups.
5352 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5353 have been a typo for LFLAGS, so it has been changed.
5355 13. The install script calls Exim with "-C /dev/null" in order to find the
5356 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5357 to be output. Howeve, since Exim outputs its version number before the
5358 error, it didn't break the script. It just looked ugly. I fixed this by
5359 always allowing "-C /dev/null" if the caller is root.
5361 14. Ignore overlarge ACL variable number when reading spool file - insurance
5362 against a later release with more variables having written the file.
5364 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5365 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5366 this, as well as the form without the "IPv6" on the front.
5368 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5369 OS/Makefile-Darwin file.
5371 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5372 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5374 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5375 variable. However, some OS use ints for inodes. I've added cast to long int
5376 to get rid of the compiler warning.
5378 19. I had forgotten to lock out "/../" in configuration file names when
5379 ALT_CONFIG_PREFIX was set.
5381 20. Routers used for verification do not need to specify transports. However,
5382 if such a router generated a host list, and callout was configured, Exim
5383 crashed, because it could not find a port number from the (non-existent)
5384 transport. It now assumes port 25 in this circumstance.
5386 21. Added the -t option to exigrep.
5388 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5389 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5390 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5391 them is set, LOOKUP_LSEARCH is forced.
5393 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5395 24. Added sender and host information to the "rejected by local_scan()" log
5396 line; previously there was no indication of these.
5398 25. Added .include_if_exists.
5400 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5401 Linux. It turns out that not all file systems support this. Apparently some
5402 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5403 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5404 not supported on the file descriptor, is now ignored when Exim is trying to
5405 sync a directory. This applies only to Linux.
5407 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5409 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5410 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5413 29. Added support for maildirsize files from supplied patch (modified a bit).
5415 30. The use of :fail: followed by an empty string could lead Exim to respond to
5416 sender verification failures with (e.g.):
5418 550 Verification failed for <xxx>
5419 550 Sender verify failed
5421 where the first response line was missing the '-' that indicates it is not
5422 the final line of the response.
5424 31. The loop for finding the name of the user that called Exim had a hardwired
5425 limit of 10; it now uses the value of finduser_retries, which is used for
5426 all other user lookups.
5428 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5431 33. Exim was neglecting to zero errno before one call of strtol() when
5432 expanding a string and expecting an integer value. On some systems this
5433 resulted in spurious "integer overflow" errors. Also, it was casting the
5434 result into an int without checking.
5436 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5437 did not work. The code looks as if it has *never* worked, though it appears
5438 to have been documented since at least releast 1.62. I have made it work.
5440 35. The "timeout_DNS" error in retry rules, also documented since at least
5441 1.62, also never worked. As it isn't clear exactly what this means, and
5442 clearly it isn't a major issue, I have abolished the feature by treating it
5443 as "timeout", and writing a warning to the main and panic logs.
5445 36. The display of retry rules for -brt wasn't always showing the error code
5448 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5449 timeout_connect_A, timeout_connect_MX.
5451 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5452 to the empty sender.
5454 39. The daemon was not analysing the content of -oX till after it had closed
5455 stderr and disconnected from the controlling terminal. This meant that any
5456 syntax errors were only noted on the panic log, and the return code from
5457 the command was 0. By re-arranging the code a little, I've made the
5458 decoding happen first, so such errors now appear on stderr, and the return
5459 code is 1. However, the actual setting up of the sockets still happens in
5460 the disconnected process, so errors there are still only recorded on the
5463 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5464 connections (as happens on some IP stacks) was logged at start up time as
5465 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5466 it from "IPv6 and IPv4", which means that two separate sockets are being
5469 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5470 says whether AF_INET or AF_INET6 was passed as an argument.
5472 42. Exiwhat output was messed up when time zones were included in log
5475 43. Exiwhat now gives more information about the daemon's listening ports,
5476 and whether -tls-on-connect was used.
5478 44. The "port" option of the smtp transport is now expanded.
5480 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5481 silently ignored. Now an error message is written to the main and panic
5484 46. There's a new ACL modifier called "logwrite" which writes to a log file
5485 as soon as it is encountered.
5487 47. Added $local_user_uid and $local_user_gid at routing time.
5489 48. Exim crashed when trying to verify a sender address that was being
5492 49. Exim was recognizing only a space character after ".include". It now also
5493 recognizes a tab character.
5495 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5496 extracting the relevant information from the specification. The man page no
5497 longer contains scrambled data for the -d option, and I've added a section
5498 at the front about calling Exim under different names.
5500 51. Added "extra_headers" argument to the "mail" command in filter files.
5502 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5505 53. Installed eximstats 1.29.
5507 54. Added transport_filter_timeout as a generic transport option.
5509 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5510 Cc: header lines. This was required by RFC 822, but it not required by RFC
5513 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5514 incoming messages that did not have them. Now it does so only if the
5515 message originates locally, that is, if there is no associated remote host
5516 address. When Resent- header lines are present, this applies to the Resent-
5517 lines rather than the non-Resent- lines.
5519 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5520 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5522 58. Messages for configuration errors now include the name of the main
5523 configuration file - useful now that there may be more than one file in a
5524 list (.included file names were always shown).
5526 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5527 for those rare installations that do not start the daemon as root or run it
5528 setuid root. I've cut out the call to initgroups() if the daemon is not
5531 60. The Exim user and group can now be bound into the binary as text strings
5532 that are looked up at the start of Exim's processing.
5534 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5536 62. Added $mailstore_basename variable.
5538 63. Installed patch to sieve.c from Michael Haardt.
5540 64. When Exim failed to open the panic log after failing to open the main log,
5541 the original message it was trying to log was written to stderr and debug
5542 output, but if they were not available (the usual case in production), it
5543 was lost. Now it is written to syslog before the two lines that record the
5544 failures to open the logs.
5546 65. Users' Exim filters run in subprocesses under the user's uid. It is
5547 possible for a "deliver" command or an alias in a "personal" command to
5548 provoke an address rewrite. If logging of address rewriting is configured,
5549 this fails because the process is not running as root or exim. There may be
5550 a better way of dealing with this, but for the moment (because 4.30 needs
5551 to be released), I have disabled address rewrite logging when running a
5552 filter in a non-root, non-exim process.
5558 1. The buildconfig auxiliary program wasn't quoting the value set for
5559 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5560 not defined. This bug was masked in 4.22 by the effect that was fixed in
5563 2. Some messages that were rejected after a message id was allocated were
5564 shown as "incomplete" by exigrep. It no longer does this for messages that
5565 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5567 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5568 have allow_domain_literals set, the ID did not get logged in the <= line.
5569 Domain literals are now always recognized in Message-ID: header lines.
5571 4. The first argument for a ${extract expansion item is the key name or field
5572 number. Leading and trailing spaces in this item were not being ignored,
5573 causing some misleading effects.
5575 5. When deliver_drop_privilege was set, single queue runner processes started
5576 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5577 same command in the process it spins off) were not dropping privilege.
5579 6. When the daemon running as "exim" started a queue runner, it always
5580 re-executed Exim in the spun-off process. This is a waste of effort when
5581 deliver_drop_privilege is set. The new process now just calls the
5582 queue-runner function directly.
5588 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5591 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5592 ignored. Though the use of -oP was forcing the writing of a pid file, it
5593 was always written to the default place.
5595 3. If the message "no IP address found for host xxxx" is generated during
5596 incoming verification, it is now followed by identification of the incoming
5597 connection (so you can more easily find what provoked it).
5599 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5601 5. Added some features to "harden" Exim a bit more against certain attacks:
5603 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5604 be put in Local/Makefile. This is like the never_users runtime option,
5605 but it cannot be overridden. The default setting is "root".
5607 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5608 prefix string with which any file named in a -C command line option
5611 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5612 is retained for -C and -D only if the caller of Exim is root. Without
5613 it, the exim user may also use -C and -D and retain privilege.
5615 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5616 command line option is disabled.
5618 6. Macro names set by the -D option must start with an upper case letter, just
5619 like macro names defined in the configuration file.
5621 7. Added "dereference=" facility to LDAP.
5623 8. Two instances of the typo "uknown" in the source files are fixed.
5625 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5626 the Configure-Makefile script screwed up while processing it.
5628 10. Incorporated PCRE 4.4.
5630 11. The SMTP synchronization check was not operating right at the start of an
5631 SMTP session. For example, it could not catch a HELO sent before the client
5632 waited for the greeting. There is now a check for outstanding input at the
5633 point when the greeting is written. Because of the duplex, asynchronous
5634 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5635 way, but not yet received, when the check is performed.
5637 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5638 on TCP/IP sockets, because this apparently causes some broken clients to
5641 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5642 unchanged) from the Cygwin maintainer.
5644 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5645 was supported instead of testing for mbx. Effectively a typo.
5647 15. The spa authenticator server code was not checking that the input it
5648 received was valid base64.
5650 16. The debug output line for the "set" modifier in ACLs was not showing the
5651 name of the variable that was being set.
5653 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5655 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5656 lookup on its own. Something else was needed to trigger the lookup. For
5657 example, a match in host_lookup or the need for a host name in a host list.
5658 Now, if $sender_host_name is referenced and the host name has not yet been
5659 looked up, a lookup is performed. If the lookup fails, the variable remains
5660 empty, and $host_lookup_failed is set to "1".
5662 19. Added "eqi" as a case-independent comparison operator.
5664 20. The saslauthd authentication condition could segfault if neither service
5665 nor realm was specified.
5667 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5668 error message that was logged was misleading, and incoming SMTP
5669 connections were dropped. The message is now more accurate, and temporary
5670 errors are given to SMTP connections.
5672 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5673 (except RSET) with a 421 error, until QUIT is received. However, it was
5674 failing to send a response to QUIT.
5676 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5677 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5678 ACL is now run after the helo_try_verify_hosts code.
5680 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5681 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5682 case-independent, so other case variants are also recognized). Apparently
5683 some systems use these upper case variants.
5685 25. If more than two messages were waiting for the same host, and a transport
5686 filter was specified for the transport, Exim sent two messages over the
5687 same TCP/IP connection, and then failed with "socket operation on non-
5688 socket" when it tried to send the third.
5690 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5692 27. The extern definition of crypt16() in expand.c was not being excluded when
5693 the OS had its own crypt16() function.
5695 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5696 as a preferred synonym for return_size_limit, both as an option and as an
5699 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5701 30. Changed the default configuration ACL to relax the local part checking rule
5702 for addresses that are not in any local domains. For these addresses,
5703 slashes and pipe symbols are allowed within local parts, but the sequence
5704 /../ is explicitly forbidden.
5706 31. SPA server authentication was not clearing the challenge buffer before
5709 32. log_message in a "warn" ACL statement was writing to the reject log as
5710 well as to the main log, which contradicts the documentation and doesn't
5711 seem right (because no rejection is happening). So I have stopped it.
5713 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5714 However, I am unable to do any testing of this.
5716 34. Fixed an infelicity in the appendfile transport. When checking directories
5717 for a mailbox, to see if any needed to be created, it was accidentally
5718 using path names with one or more superfluous leading slashes; tracing
5719 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5721 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5722 contents are added to the log line that is written for every discarded
5723 recipient. (Previously a log_message setting was ignored.)
5725 36. The ${quote: operator now quotes the string if it is empty.
5727 37. The install script runs exim in order to find its version number. If for
5728 some reason other than non-existence or emptiness, which it checks, it
5729 could not run './exim', it was installing it with an empty version number,
5730 i.e. as "exim-". This error state is now caught, and the installation is
5733 38. An argument was missing from the function that creates an error message
5734 when Exim fails to connect to the socket for saslauthd authentication.
5735 This could cause Exim to crash, or give a corrupted message.
5737 39. Added isip, isip4, and isip6 to ${if conditions.
5739 40. The ACL variables $acl_xx are now saved with the message, and can be
5740 accessed later in routers, transports, and filters.
5742 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5743 strings in the file are not string-expanded.
5745 42. If a MAIL command specified a SIZE value that was too large to fit into an
5746 int variable, the check against message_size_limit failed. Such values are
5747 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5748 day this will have to be increased, but I don't think I want to be around
5749 when emails are that large.
5756 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5757 iconv() is not standard in FreeBSD.
5759 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5760 IPv6 enabled. The observed symptom was a segmentation fault on return from
5761 the function os_common_find_running_interfaces() in src/os.c.
5763 3. In the check_special_case() function in daemon.c I had used "errno" as an
5764 argument name, which causes warnings on some systems. This was basically a
5765 typo, since it was named "eno" in the comments!
5767 4. The code that waits for the clock to tick (at a resolution of some fraction
5768 of a second) so as to ensure message-id uniqueness was always waiting for
5769 at least one whole tick, when it could have waited for less. [This is
5770 almost certainly not relevant at current processor speeds, where it is
5771 unlikely to ever wait at all. But we try to future-proof.]
5773 5. The function that sleeps for a time interval that includes fractions of a
5774 second contained a race. It did not block SIGALRM between setting the
5775 timer, and suspending (a couple of lines later). If the interval was short
5776 and the sigsuspend() was delayed until after it had expired, the suspension
5777 never ended. On busy systems this could lead to processes getting stuck for
5780 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5781 process, before it forks any delivery processes. The open lookup caching
5782 mechanism meant that the open file or database connection was passed into
5783 the delivery process. The problem was that delivery processes always tidy
5784 up cached lookup data. This could cause a problem for the next delivery
5785 process started by the queue runner, because the external queue runner
5786 process does not know about the closure. So the next delivery process
5787 still has data in the lookup cache. In the case of a file lookup, there was
5788 no problem because closing a file descriptor in a subprocess doesn't affect
5789 the parent. However, if the lookup was caching a connection to a database,
5790 the connection was closed, and the second delivery process was likely to
5791 see errors such as "PGSQL: query failed: server closed the connection
5792 unexpectedly". The problem has been fixed by closing all cached lookups
5793 in a queue runner before running a delivery process.
5795 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5796 seem to have the "const" qualifier which it has on other OS. I've
5799 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5800 *of the same type* (client or server) with the same public name that an
5801 error should be diagnosed.
5803 9. When Exim looked up a host name for an IP address, but failed to find the
5804 original IP address when looking up the host name (a safety check), it
5805 output the message "<ip address> does not match any IP for NULL", which was
5806 confusing, to say the least. The bug was that the host name should have
5807 appeared instead of "NULL".
5809 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5810 user that is built into the binary, and the -C or -D options is used, root
5811 privilege is dropped before the configuration file is read. In addition,
5812 logging is switched to stderr instead of the normal log files. If the
5813 configuration then re-defines the Exim user, the unprivileged environment
5814 is probably not what is expected, so Exim logs a panic warning message (but
5817 However, if deliver_drop_privilege is set, the unprivileged state may well
5818 be exactly what is intended, so the warning has been cut out in that case,
5819 and Exim is allowed to try to write to its normal log files.
5825 1. smtp_return_error_details was not giving details for temporary sender
5826 or receiver verification errors.
5828 2. Diagnose a configuration error if two authenticators have the same public
5831 3. Exim used not to create the message log file for a message until the first
5832 delivery attempt. This could be confusing when incoming messages were held
5833 for policy or load reasons. The message log file is now created at the time
5834 the message is received, and an initial "Received" line is written to it.
5836 4. The automatically generated man page for command line options had a minor
5837 bug that caused no ill effects; however, a more serious problem was that
5838 the procedure for building the man page automatically didn't always
5839 operate. Consequently, release 4.20 contains an out-of-date version. This
5840 shouldn't happen again.
5842 5. When building Exim with embedded Perl support, the script that builds the
5843 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5844 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5846 6. The freeze_tell option was not being used for messages that were frozen on
5847 arrival, either by an ACL or by local_scan().
5849 7. Added the smtp_incomplete_transaction log selector.
5851 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5852 was accepting AUTH without a new EHLO.
5854 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5855 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5858 10. Logging of TCP/IP connections (when configured) now happens in the main
5859 daemon process instead of the child process, so that the TCP/IP connection
5860 count is more accurate (but it can never be perfect).
5862 11. The use of "drop" in a nested ACL was not being handled correctly in the
5863 outer ACL. Now, if condition failure induced by the nested "drop" causes
5864 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5865 or "require"), the connection is dropped.
5867 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5868 that yield "discard" can now be used with an "accept" or a "discard" verb,
5869 but an error is generated for any others (because I can't see a useful way
5870 to define what should happen).
5872 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5873 are now processed in the same way as lines in the Exim configuration file.
5874 In particular, continuation lines are supported.
5876 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5878 15. Added -ti meaning -t -i.
5880 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5881 domains, and warn by logging if others are found.
5883 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5884 returned from GIFCONF when Exim is trying to find the list of interfaces on
5885 a host. The code in os.c has been modified to copy each ifreq to an aligned
5886 structure in all cases.
5888 Also, in some cases, the returned ifreq's were being copied to a 'struct
5889 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5890 means the last couple of bytes of an IPv6 address could be chopped if the
5891 ifreq contained only a normal sockaddr (14 bytes storage).
5893 18. Named domain lists were not supported in the hosts_treat_as_local option.
5894 An entry such as +xxxx was not recognized, and was treated as a literal
5897 19. Ensure that header lines added by a DATA ACL are included in the reject log
5898 if the ACL subsequently rejects the message.
5900 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5901 MD5 (which is deprecated).
5903 21. When testing a filter file using -bf, Exim was writing a message when it
5904 took the sender from a "From " line in the message, but it was not doing so
5905 when it took $return_path from a Return-Path: header line. It now does.
5907 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5908 with a valid header line field name (a series of printing characters
5909 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5911 23. Changed "disc" in the source to "disk" to conform to the documentation and
5912 the book and for uniformity.
5914 24. Ignore Sendmail's -Ooption=value command line item.
5916 25. When execve() failed while trying to run a command in a pipe transport,
5917 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5918 could be confused with a return value of 69 from the command itself. This
5919 has been changed to 127, the value the shell returns if it is asked to run
5920 a non-existent command. The wording for the related log line suggests a
5921 non-existent command as the problem.
5923 26. If received_header_text expands to an empty string, do not add a Received:
5924 header line to the message. (Well, it adds a token one on the spool, but
5925 marks it "old" so that it doesn't get used or transmitted.)
5927 27. Installed eximstats 1.28 (addition of -nt option).
5929 28. There was no check for failure on the call to getsockname() in the daemon
5930 code. This can fail if there is a shortage of resources on the system, with
5931 ENOMEM, for example. A temporary error is now given on failure.
5933 29. Contrary to the C standard, it seems that in some environments, the
5934 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5935 program. Exim now does this explicitly; it affects the formatting of
5936 timestamps using strftime().
5938 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5939 complaints. I've now initialized all the variables, to avoid this.
5941 32. Header lines added by a system filter were not being "seen" during
5942 transport-time rewrites.
5944 33. The info_callback() function passed to OpenSSL is set up with type void
5945 (*)(SSL *, int, int), as described somewhere. However, when calling the
5946 function (actually a macro) that sets it up, the type void(*)() is
5947 expected. I've put in a cast to prevent warnings from picky compilers.
5949 34. If a DNS black list lookup found a CNAME record, but there were no A
5950 records associated with the domain it pointed at, Exim crashed.
5952 35. If a DNS black list lookup returned more than one A record, Exim ignored
5953 all but the first. It now scans all returned addresses if a particular IP
5954 value is being sought. In this situation, the contents of the
5955 $dnslist_value variable are a list of all the addresses, separated by a
5958 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5959 to accept a host name and all its aliases if the forward lookup for any of
5960 them yielded the IP address of the incoming connection. Now it accepts only
5961 those names whose forward lookup yields the correct IP address. Any other
5962 names are discarded. This closes a loophole whereby a rogue DNS
5963 administrator could create reverse DNS records to break through a
5964 wildcarded host restriction in an ACL.
5966 37. If a user filter or a system filter that ran in a subprocess used any of
5967 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5968 the wrong values were passed to the pipe command ($thisaddress had the
5969 value of $0, $0 had the value of $1, etc). This bug was introduced by
5970 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5972 38. Improved the line breaking for long SMTP error messages from ACLs.
5973 Previously, if there was no break point between 40 and 75 characters, Exim
5974 left the rest of the message alone. Two changes have been made: (a) I've
5975 reduced the minimum length to 35 characters; (b) if it can't find a break
5976 point between 35 and 75 characters, it looks ahead and uses the first one
5977 that it finds. This may give the occasional overlong line, but at least the
5978 remaining text gets split now.
5980 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5981 file descriptors might be low, and that setting 1000 would always raise it.
5982 It turns out that in some environments, the limit is already over 1000 and
5983 that lowering it causes trouble. So now Exim takes care not to decrease it.
5985 40. When delivering a message, the value of $return_path is set to $sender_
5986 address at the start of routing (routers may change the value). By an
5987 oversight, this default was not being set up when an address was tested by
5988 -bt or -bv, which affected the outcome if any router or filter referred to
5991 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5992 log only when "message" or "log_message" are set. However, if one of the
5993 conditions was an address verification, or a call to a nested ACL, the
5994 messages generated by the underlying test were being passed through. This
5995 no longer happens. The underlying message is available in $acl_verify_
5996 message for both "message" and "log_message" expansions, so it can be
5997 passed through if needed.
5999 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6000 new expansion $bh_ to give the encoded byte string without charset
6001 translation. Translation happens only if iconv() is available; HAVE_ICONV
6002 indicates this at build time. HEADERS_CHARSET gives the charset to
6003 translate to; headers_charset can change it in the configuration, and
6004 "headers charset" can change it in an individual filter file.
6006 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6007 that creates RFC 2047 encoded "words" labels them as that charset instead
6008 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6009 expansion operator; (ii) when Exim creates a From: line for a local
6010 message; (iii) when a header line is rewritten to include a "phrase" part.
6012 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6013 buggy, causing it to skip the first lines of messages whose message ID
6014 ended in 'D'. This would not have bitten before Exim release 4.14, because
6015 message IDs were unlikely to end in 'D' before then. The effect was to have
6016 incorrect size information for certain domains.
6018 45. #include "config.h" was missing at the start of the crypt16.c module. This
6019 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6022 46. If there was a timeout during a "random" callout check, Exim treated it as
6023 a failure of the random address, and carried on sending RSET and the real
6024 address. If the delay was just some slowness somewhere, the response to the
6025 original RCPT would be taken as a response to RSET and so on, causing
6026 mayhem of various kinds.
6028 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6029 when I implemented it. It didn't allow for the fact that some option values
6030 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6031 the right test for positive values.
6033 48. Domain names in DNS records are case-independent. Exim always looks them up
6034 in lower case. Some resolvers return domain names in exactly the case they
6035 appear in the zone file, that is, they may contain uppercase letters. Not
6036 all resolvers do this - some return always lower case. Exim was treating a
6037 change of case by a resolver as a change of domain, similar to a widening
6038 of a domain abbreviation. This triggered its re-routing code and so it was
6039 trying to route what was effectively the same domain again. This normally
6040 caused routing to fail (because the router wouldn't handle the domain
6041 twice). Now Exim checks for this case specially, and just changes the
6042 casing of the domain that it ultimately uses when it transmits the message
6045 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6048 50. If a filter generated a file delivery with a non-absolute name (possible if
6049 no home directory exists for the router), the forbid_file option was not
6052 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6053 the existing equality matching.
6055 52. Exim was using ints instead of ino_t variables in some places where it was
6056 dealing with inode numbers.
6058 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6059 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6060 called TMPDIR, and if it finds it is different, it changes its value.
6062 54. The smtp_printf() function is now made available to local_scan() so
6063 additional output lines can be written before returning. There is also an
6064 smtp_fflush() function to enable the detection of a dropped connection.
6065 The variables smtp_input and smtp_batched_input are exported to
6068 55. Changed the default runtime configuration: the message "Unknown user"
6069 has been removed from the ACL, and instead placed on the localuser router,
6070 using the cannot_route_message feature. This means that any verification
6071 failures that generate their own messages won't get overridden. Similarly,
6072 the "Unrouteable address" message that was in the ACL for unverifiable
6073 relay addresses has also been removed.
6075 56. Added hosts_avoid_esmtp to the smtp transport.
6077 57. The exicyclog script was not checking for the esoteric option
6078 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6079 will work only if exicyclog is run under the appropriate euid.
6081 58. Following a discussion on the list, the rules by which Exim recognises line
6082 endings on incoming messages have been changed. The -dropcr and drop_cr
6083 options are now no-ops, retained only for backwards compatibility. The
6084 following line terminators are recognized: LF CRLF CR. However, special
6085 processing applies to CR:
6087 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6088 nor a local message in the state where . is a terminator.
6090 (ii) If a bare CR is encountered in a header line, an extra space is added
6091 after the line terminator so as not to end the header. The reasoning
6092 behind this is that bare CRs in header lines are most likely either
6093 to be mistakes, or people trying to play silly games.
6095 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6096 was being incorrectly given as 18 bytes larger than it should have been.
6097 This is a VOB (very old bug).
6099 60. This may never have affected anything current, but just in case it has:
6100 When the local host is found other than at the start of a list of hosts,
6101 the local host, those with the same MX, and any that follow, are discarded.
6102 When the list in question was part of a longer list of hosts, the following
6103 hosts (not currently being processed) were also being discarded. This no
6104 longer happens. I'm not sure if this situation could ever has previously
6107 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6110 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6111 header. This is something that is recommended in a new Internet Draft, and
6112 is something that is documented as being done by Sendmail. There are two
6113 possible values. For messages generated by the autoreply transport, Exim
6116 Auto-Submitted: auto-replied
6118 whereas for all other generated messages (e.g. bounces) it adds
6120 Auto-Submitted: auto-generated
6122 63. The "personal" condition in filters now includes a test for the
6123 Auto-Submitted: header. If it contains the string "auto-" the message it
6124 not considered personal.
6126 64. Added rcpt_include_affixes as a generic transport option.
6128 65. Added queue_only_override (default true).
6130 66. Added the syslog_duplication option.
6132 67. If what should have been the first header line of a message consisted of
6133 a space followed by a colon, Exim was mis-interpreting it as a header line.
6134 It isn't of course - it is syntactically invalid and should therefore be
6135 treated as the start of the message body. The misbehaviour could have
6136 caused a number of strange effects, including loss of data in subsequent
6137 header lines, and spool format errors.
6139 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6140 client host had authenticated. This control can now be exercised by an ACL
6141 for more flexibility.
6143 69. By default, callouts do not happen when testing with -bh. There is now a
6144 variant, -bhc, which does actually run the callout code, including
6145 consulting and updating the callout cache.
6147 70. Added support for saslauthd authentication, courtesy of Alexander
6150 71. If statvfs() failed on the spool or log directories while checking their
6151 size for availability, Exim confusingly gave the error "space shortage".
6152 Furthermore, in debugging mode it crashed with a floating point exception.
6153 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6154 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6155 serious problem, Exim now writes to the main and panic logs when this
6156 happens, with details of the failure. It then refuses to accept the
6157 incoming message, giving the message "spool directory problem" or "log
6158 directory problem" with a 421 code for SMTP messages.
6160 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6161 0, 1, and 2 exist, because some OS complain for execs without them (see
6162 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6163 descriptors. However, the code omitted to check that the open succeeded,
6164 causing mysterious errors if for some reason the permissions on /dev/null
6165 got screwed. Now Exim writes a message to the main and panic logs, and
6166 bombs out if it can't open /dev/null.
6168 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6169 interact so that it is all more flexible. It is supposed to remain
6170 backwards compatible. Also added extra_local_interfaces.
6172 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6173 to bomb out with an assertion failure - to the client this appears as a
6174 connection drop. This problem occurs in the part of the code that was taken
6175 from the Samba project. Fortunately, the assertion is in a very simple
6176 function, so I have fixed this by reproducing the function inline in the
6177 one place where it is called, and arranging for authentication to fail
6178 instead of killing the process with assert().
6180 75. The SPA client code was not working when the server requested OEM rather
6181 than Unicode encoding.
6183 76. Added code to make require_files with a specific uid setting more usable in
6184 the case where statting the file as root fails - usually a non-root-mounted
6185 NFS file system. When this happens and the failure is EACCES, Exim now
6186 forks a subprocess and does the per-uid checking as the relevant uid.
6188 77. Added process_log_path.
6190 78. If log_file_path was not explicitly set, a setting of check_log_space or
6191 check_log_inodes was ignored.
6193 79. If a space check for the spool or log partitions fails, the incident is now
6194 logged. Of course, in the latter case the data may get lost...
6196 80. Added the %p formatting code to string_format() so that it can be used to
6197 print addresses in debug_print(). Adjusted all the address printing in the
6198 debugging in store.c to use %p rather than %d.
6200 81. There was a concern that a line of code in smtp_in.c could overflow a
6201 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6202 initially expressed, the concern was not well-founded, because trailing
6203 spaces are removed early. However, if the trailing spaces were followed by
6204 a NULL, they did not get removed, so the overflow was possible. Two fixes
6207 (a) I re-wrote the offending code in a cleaner fashion.
6208 (b) If an incoming SMTP command contains a NULL character, it is rejected
6211 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6212 runs initgroups(), so that if the Exim user is in any additional groups,
6213 they will be used during message reception.
6219 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob