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18 EXPERIMENTAL_CERTNAMES is withdrawn.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
32 TL/01 When running the test suite, the README says that variables such as
33 no_msglog_check are global and can be placed anywhere in a specific
34 test's script, however it was observed that placement needed to be near
35 the beginning for it to behave that way. Changed the runtest perl
36 script to read through the entire script once to detect and set these
37 variables, reset to the beginning of the script, and then run through
38 the script parsing/test process like normal.
40 TL/02 The BSD's have an arc4random API. One of the functions to induce
41 adding randomness was arc4random_stir(), but it has been removed in
42 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
43 function when detected.
45 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
46 cause callback expansion.
48 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
49 syntax errors in an expansion can be treated as a string instead of
50 logging or causing an error, due to the internal use of bool_lax
51 instead of bool when processing it.
53 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
54 server certificates when making smtp deliveries.
56 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
58 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
60 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
62 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
63 Merged patch from Sebastian Wiedenroth.
65 JH/05 Fix results-pipe from transport process. Several recipients, combined
66 with certificate use, exposed issues where response data items split
67 over buffer boundaries were not parsed properly. This eventually
68 resulted in duplicates being sent. This issue only became common enough
69 to notice due to the introduction of conection certificate information,
70 the item size being so much larger. Found and fixed by Wolfgang Breyha.
72 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
73 size buffer was used, resulting in syntax errors when an expansion
76 JH/07 Add support for directories of certificates when compiled with a GnuTLS
77 version 3.3.6 or later.
79 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
80 is EXPERIMENTAL_EVENT, the main-configuration and transport options
81 both become "event_action", the variables become $event_name, $event_data
82 and $event_defer_errno. There is a new variable $verify_mode, usable in
83 routers, transports and related events. The tls:cert event is now also
84 raised for inbound connections, if the main configuration event_action
87 TL/06 In test suite, disable OCSP for old versions of openssl which contained
88 early OCSP support, but no stapling (appears to be less than 1.0.0).
90 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
91 server certificate names available under the smtp transport option
92 "tls_verify_cert_hostname" now do not permit multi-component wildcard
95 JH/10 Time-related extraction expansions from certificates now use the main
96 option "timezone" setting for output formatting, and are consistent
97 between OpenSSL and GnuTLS compilations. Bug 1541.
99 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
102 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
103 include certificate info, eximon was claiming there were spoolfile
106 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
108 JH/14 Log delivery-related information more consistently, using the sequence
109 "H=<name> [<ip>]" wherever possible.
114 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
115 checkers that were complaining about end of non-void function with no
118 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
119 This was a regression intruduced in 4.83 by another bugfix.
121 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
123 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
124 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
130 TF/01 Correctly close the server side of TLS when forking for delivery.
132 When a message was received over SMTP with TLS, Exim failed to clear up
133 the incoming connection properly after forking off the child process to
134 deliver the message. In some situations the subsequent outgoing
135 delivery connection happened to have the same fd number as the incoming
136 connection previously had. Exim would try to use TLS and fail, logging
137 a "Bad file descriptor" error.
139 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
140 utilities have not been installed.
142 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
143 temporary space as the ACL may create new global variables.
145 TL/01 LDAP support uses per connection or global context settings, depending
146 upon the detected version of the libraries at build time.
148 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
149 to extract and use the src ip:port in logging and expansions as if it
150 were a direct connection from the outside internet. PPv2 support was
151 updated based on HAProxy spec change in May 2014.
153 JH/02 Add ${listextract {number}{list}{success}{fail}}.
155 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
156 Properly escape header and check for NULL return.
158 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
161 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
163 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
164 characters in header names, implemented as a verify condition.
165 Contributed by Michael Fischer v. Mollard.
167 TL/05 Rename SPF condition results err_perm and err_temp to standardized
168 results permerror and temperror. Previous values are deprecated but
169 still accepted. In a future release, err_perm and err_temp will be
170 completely removed, which will be a backward incompatibility if the
171 ACL tests for either of these two old results. Patch contributed by
172 user bes-internal on the mailing list.
174 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
176 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
177 selectors, in both main and reject logs.
179 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
182 JH/07 Add malware type "sock" for talking to simple daemon.
184 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
186 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
187 routers/transports under cutthrough routing.
189 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
190 numbers. Touch up "bool" conditional to keep the same definition.
192 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
194 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
196 JH/12 Expand items in router/transport headers_add or headers_remove lists
197 individually rather than the list as a whole. Bug 1452.
199 Required for reasonable handling of multiple headers_ options when
200 they may be empty; requires that headers_remove items with embedded
201 colons must have them doubled (or the list-separator changed).
203 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
204 view the policy declared in the DMARC record. Currently, $dmarc_status
205 is a combined value of both the record presence and the result of the
208 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
210 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
211 dnslookup router and the smtp transport (applying to the forward
214 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
215 of ldap servers used for a specific lookup. Patch provided by Heiko
218 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
219 New variable $lookup_dnssec_authenticated for observability.
221 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
222 Patch submitted by Lars Timman.
224 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
226 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
227 Requires trusted mode and valid format message id, aborts otherwise.
228 Patch contributed by Heiko Schlichting.
230 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
231 certextract with support for various fields. Bug 1358.
233 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
234 is requested by default, modifiable by smtp transport option
237 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
238 operate on certificate variables to give certificate fingerprints
239 Also new ${sha256:cert_variable}.
241 JH/23 The PRDR feature is moved from being Experimental into the mainline.
243 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
244 Christian Aistleitner.
246 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
248 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
249 file. Patch from Wolfgang Breyha.
251 JH/25 Expand the coverage of the delivery $host and $host_address to
252 client authenticators run in verify callout. Bug 1476.
254 JH/26 Port service names are now accepted for tls_on_connect_ports, to
255 align with daemon_smtp_ports. Bug 72.
257 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
258 support and error reporting did not work properly.
260 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
261 and is readable. Patch from Andrew Colin Kissa.
263 TL/14 Enhance documentation of ${run expansion and how it parses the
264 commandline after expansion, particularly in the case when an
265 unquoted variable expansion results in an empty value.
267 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
269 PP/02 Fix internal collision of T_APL on systems which support RFC3123
270 by renaming away from it. Addresses GH issue 15, reported by
273 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
275 TL/15 SECURITY: prevent double expansion in math comparison functions
276 (can expand unsanitized data). Not remotely exploitable.
283 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
285 PP/02 Make -n do something, by making it not do something.
286 When combined with -bP, the name of an option is not output.
288 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
291 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
292 $sender_host_name and config options to manage this, and basic check
295 PP/05 DSCP support for outbound connections and control modifier for inbound.
297 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
298 (Only plugin which currently uses this is kerberos4, which nobody should
299 be using, but we should make it available and other future plugins might
300 conceivably use it, even though it would break NAT; stuff *should* be
301 using channel bindings instead).
303 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
304 name; added for Sendmail compatibility; requires admin caller.
305 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
306 just ignore it); requires trusted caller.
307 Also parse but ignore: -Ac -Am -X<logfile>
310 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
312 TL/02 Add +smtp_confirmation as a default logging option.
314 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
315 Patch by Magnus Holmgren from 2007-02-20.
317 TL/04 Bugzilla 1281 - Spec typo.
318 Bugzilla 1283 - Spec typo.
319 Bugzilla 1290 - Spec grammar fixes.
321 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
323 TL/06 Add Experimental DMARC support using libopendmarc libraries.
325 TL/07 Fix an out of order global option causing a segfault. Reported to dev
326 mailing list by by Dmitry Isaikin.
328 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
330 JH/02 Support "G" suffix to numbers in ${if comparisons.
332 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
334 NM/01 Bugzilla 1197 - Spec typo
335 Bugzilla 1196 - Spec examples corrections
337 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
339 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
340 gnutls_enable_pkcs11, but renamed to more accurately indicate its
343 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
344 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
346 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
347 "acl {{name}{arg}...}", and optional args on acl condition
350 JH/05 Permit multiple router/transport headers_add/remove lines.
352 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
354 JH/07 Avoid using a waiting database for a single-message-only transport.
355 Performance patch from Paul Fisher. Bugzilla 1262.
357 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
360 JH/09 Add $headers_added variable, with content from use of ACL modifier
361 add_header (but not yet added to the message). Bugzilla 199.
363 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
364 Pulled from Bugzilla 817 by Wolfgang Breyha.
366 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
368 (nb: this is the same fix as in Exim 4.80.1)
370 JH/11 Add A= logging on delivery lines, and a client_set_id option on
373 JH/12 Add optional authenticated_sender logging to A= and a log_selector
376 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
378 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
379 advertise SMTP AUTH mechanism to us, instead of a generic
380 protocol violation error. Also, make Exim more robust to bad
381 data from the Dovecot auth socket.
383 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
385 When a queue runner is handling a message, Exim first routes the
386 recipient addresses, during which it prunes them based on the retry
387 hints database. After that it attempts to deliver the message to
388 any remaining recipients. It then updates the hints database using
391 So if a recipient address works intermittently, it can get repeatedly
392 deferred at routing time. The retry hints record remains fresh so the
393 address never reaches the final cutoff time.
395 This is a fairly common occurrence when a user is bumping up against
396 their storage quota. Exim had some logic in its local delivery code
397 to deal with this. However it did not apply to per-recipient defers
398 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
400 This change adds a proper retry rule check during routing so that the
401 final cutoff time is checked against the message's age. We only do
402 this check if there is an address retry record and there is not a
403 domain retry record; this implies that previous attempts to handle
404 the address had the retry_use_local_parts option turned on. We use
405 this as an approximation for the destination being like a local
406 delivery, as in LMTP.
408 I suspect this new check makes the old local delivery cutoff check
409 redundant, but I have not verified this so I left the code in place.
411 TF/02 Correct gecos expansion when From: is a prefix of the username.
413 Test 0254 submits a message to Exim with the header
417 When I ran the test suite under the user fanf2, Exim expanded
418 the header to contain my full name, whereas it should have added
419 a Resent-Sender: header. It erroneously treats any prefix of the
420 username as equal to the username.
422 This change corrects that bug.
424 GF/01 DCC debug and logging tidyup
425 Error conditions log to paniclog rather than rejectlog.
426 Debug lines prefixed by "DCC: " to remove any ambiguity.
428 TF/03 Avoid unnecessary rebuilds of lookup-related code.
430 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
431 Bug spotted by Jeremy Harris; was flawed since initial commit.
432 Would have resulted in OCSP responses post-SNI triggering an Exim
433 NULL dereference and crash.
435 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
437 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
438 Bug detection, analysis and fix by Samuel Thibault.
439 Bugzilla 1331, Debian bug #698092.
441 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
443 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
444 Server implementation by Todd Lyons, client by JH.
445 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
446 config variable "prdr_enable" controls whether the server
447 advertises the facility. If the client requests PRDR a new
448 acl_data_smtp_prdr ACL is called once for each recipient, after
449 the body content is received and before the acl_smtp_data ACL.
450 The client is controlled by bolth of: a hosts_try_prdr option
451 on the smtp transport, and the server advertisement.
452 Default client logging of deliveries and rejections involving
453 PRDR are flagged with the string "PRDR".
455 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
456 fclose(). Diagnosis by Todd Lyons.
458 PP/17 Update configure.default to handle IPv6 localhost better.
459 Patch by Alain Williams (plus minor tweaks).
462 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
463 This is now consistent with GnuTLS, and is now documented: the
464 previous undocumented portable approach to treating the option as
465 unset was to force an expansion failure. That still works, and
466 an empty string is now equivalent.
468 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
469 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
470 not performing validation itself.
472 PP/20 Added force_command boolean option to pipe transport.
473 Patch from Nick Koston, of cPanel Inc.
475 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
478 TF/04 Added udpsend ACL modifer and hexquote expansion operator
480 PP/21 Fix eximon continuous updating with timestamped log-files.
481 Broken in a format-string cleanup in 4.80, missed when I repaired the
482 other false fix of the same issue.
483 Report and fix from Heiko Schlichting.
486 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
487 Report from Prashanth Katuri.
489 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
490 It's SecureTransport, so affects any MacOS clients which use the
491 system-integrated TLS libraries, including email clients.
493 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
494 using a MIME ACL for non-SMTP local injection.
495 Report and assistance in diagnosis by Warren Baker.
497 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
499 JH/16 Fix comparisons for 64b. Bugzilla 1385.
501 TL/09 Add expansion variable $authenticated_fail_id to keep track of
502 last id that failed so it may be referenced in subsequent ACL's.
504 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
507 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
508 ldap library initialization, allowing self-signed CA's to be
509 used. Also properly sets require_cert option later in code by
510 using NULL (global ldap config) instead of ldap handle (per
511 session). Bug diagnosis and testing by alxgomz.
513 TL/12 Enhanced documentation in the ratelimit.pl script provided in
514 the src/util/ subdirectory.
516 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
517 renamed to Transport Post Delivery Action by Jeremy Harris, as
520 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
521 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
522 redis_servers = needs to be configured which will be used by the redis
523 lookup. Patch from Warren Baker, of The Packet Hub.
525 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
527 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
528 hostname or reverse DNS when processing a host list. Used suggestions
529 from multiple comments on this bug.
531 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
533 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
534 Missed a few lines, added it to make the runtest require no keyboard
537 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
538 contains upper case chars. Make router use caseful_local_part.
540 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
541 support when GnuTLS has been built with p11-kit.
547 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
549 This, or similar/improved, will also be change PP/11 of 4.82.
555 PP/01 Handle short writes when writing local log-files.
556 In practice, only affects FreeBSD (8 onwards).
557 Bugzilla 1053, with thanks to Dmitry Isaikin.
559 NM/01 Bugzilla 949 - Documentation tweak
561 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
564 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
566 PP/02 Implemented gsasl authenticator.
568 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
570 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
571 `pkg-config foo` for cflags/libs.
573 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
574 with rest of GSASL and with heimdal_gssapi.
576 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
577 `pkg-config foo` for cflags/libs for the TLS implementation.
579 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
580 properties get this fed in as external SSF. A number of robustness
581 and debugging improvements to the cyrus_sasl authenticator.
583 PP/08 cyrus_sasl server now expands the server_realm option.
585 PP/09 Bugzilla 1214 - Log authentication information in reject log.
586 Patch by Jeremy Harris.
588 PP/10 Added dbmjz lookup type.
590 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
592 PP/12 MAIL args handles TAB as well as SP, for better interop with
593 non-compliant senders.
594 Analysis and variant patch by Todd Lyons.
596 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
597 Bug report from Lars Müller <lars@samba.org> (via SUSE),
598 Patch from Dirk Mueller <dmueller@suse.com>
600 PP/13 tls_peerdn now print-escaped for spool files.
601 Observed some $tls_peerdn in wild which contained \n, which resulted
602 in spool file corruption.
604 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
605 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
606 or write after TLS renegotiation, which otherwise led to messages
609 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
610 as a tracking header (ie: a signed header comes before the signature).
611 Patch from Wolfgang Breyha.
613 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
614 comma-sep list; embedded commas doubled.
616 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
618 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
620 Report and patch from Dmitry Banschikov.
622 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
623 Removed SSL_clear() after SSL_new() which led to protocol negotiation
624 failures. We appear to now support TLS1.1+ with Exim.
626 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
627 lets Exim select keys and certificates based upon TLS SNI from client.
628 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
629 before an outbound SMTP session. New log_selector, +tls_sni.
631 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
632 NULL dereference. Report and patch from Alun Jones.
634 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
635 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
636 Not seeing resolver debug output on NetBSD, but suspect this is a
637 resolver implementation change.
639 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
640 Left warnings. Added "eximon gdb" invocation mode.
642 PP/21 Defaulting "accept_8bitmime" to true, not false.
644 PP/22 Added -bw for inetd wait mode support.
646 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
647 locate the relevant includes and libraries. Made this the default.
649 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
650 Bugzilla 1246, report and most of solution from Tomasz Kusy.
652 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
653 This may cause build issues on older platforms.
655 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
656 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
657 gnutls_require_mac & gnutls_require_protocols (no longer supported).
658 Added SNI support via GnuTLS too.
659 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
661 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
663 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
664 Applied second patch from Janne, implementing suggestion to default
665 multiple-strings-in-record handling to match SPF spec.
667 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
669 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
670 read-only, out of scope).
671 Patch from Wolfgang Breyha, report from Stuart Northfield.
673 PP/29 Fix three issues highlighted by clang analyser static analysis.
674 Only crash-plausible issue would require the Cambridge-specific
675 iplookup router and a misconfiguration.
676 Report from Marcin Mirosław.
678 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
680 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
681 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
682 As part of this, removing so much warning spew let me fix some minor
683 real issues in debug logging.
685 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
686 assignment on my part. Fixed.
688 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
689 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
690 Janne Snabb (who went above and beyond: thank you).
692 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
693 string otherwise requires a connection and a bunch more work and it's
694 relatively easy to get wrong. Should also expose TLS library linkage
697 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
698 64-bit ${eval} (JH/03).
700 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
701 GNU libc to support some of the 64-bit stuff, should not lead to
702 conflicts. Defined before os.h is pulled in, so if a given platform
703 needs to override this, it can.
705 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
706 protection layer was required, which is not implemented.
707 Bugzilla 1254, patch from Wolfgang Breyha.
709 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
710 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
711 tls_dhparam take prime identifiers. Also unbreak combination of
712 OpenSSL+DH_params+TLSSNI.
714 PP/39 Disable SSLv2 by default in OpenSSL support.
720 PP/01 Solaris build fix for Oracle's LDAP libraries.
721 Bugzilla 1109, patch from Stephen Usher.
723 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
725 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
728 TF/02 Fix a couple more cases where we did not log the error message
729 when unlink() failed. See also change 4.74-TF/03.
731 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
732 lock up or crash if it happened to be inside a call to libc when it
733 got a SIGUSR1 from exiwhat.
735 The SIGUSR1 handler appends the current process status to the process
736 log which is later printed by exiwhat. It used to use the general
737 purpose logging code to do this, but several functions it calls are
738 not safe for signals.
740 The new output code in the SIGUSR1 handler is specific to the process
741 log, and simple enough that it's easy to inspect for signal safety.
742 Removing some special cases also simplifies the general logging code.
743 Removing the spurious timestamps from the process log simplifies
746 TF/04 Improved ratelimit ACL condition.
748 The /noupdate option has been deprecated in favour of /readonly which
749 has clearer semantics. The /leaky, /strict, and /readonly update modes
750 are mutually exclusive. The update mode is no longer included in the
751 database key; it just determines when the database is updated. (This
752 means that when you upgrde Exim will forget old rate measurements.)
754 Exim now checks that the per_* options are used with an update mode that
755 makes sense for the current ACL. For example, when Exim is processing a
756 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
757 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
758 must specify per_mail/readonly. If you omit the update mode it defaults to
759 /leaky where that makes sense (as before) or /readonly where required.
761 The /noupdate option is now undocumented but still supported for
762 backwards compatibility. It is equivalent to /readonly except that in
763 ACLs where /readonly is required you may specify /leaky/noupdate or
764 /strict/noupdate which are treated the same as /readonly.
766 A useful new feature is the /count= option. This is a generalization
767 of the per_byte option, so that you can measure the throughput of other
768 aggregate values. For example, the per_byte option is now equivalent
769 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
771 The per_rcpt option has been generalized using the /count= mechanism
772 (though it's more complicated than the per_byte equivalence). When it is
773 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
774 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
775 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
776 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
777 Note that using per_rcpt with a non-readonly update mode in more than
778 one ACL will cause the recipients to be double-counted. (The per_mail
779 and per_byte options don't have this problem.)
781 The handling of very low rates has changed slightly. If the computed rate
782 is less than the event's count (usually one) then this event is the first
783 after a long gap. In this case the rate is set to the same as this event's
784 count, so that the first message of a spam run is counted properly.
786 The major new feature is a mechanism for counting the rate of unique
787 events. The new per_addr option counts the number of different
788 recipients that someone has sent messages to in the last time period. It
789 behaves like per_rcpt if all the recipient addresses are different, but
790 duplicate recipient addresses do not increase the measured rate. Like
791 the /count= option this is a general mechanism, so the per_addr option
792 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
793 example, measure the rate that a client uses different sender addresses
794 with the options per_mail/unique=$sender_address. There are further
795 details in the main documentation.
797 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
799 TF/06 Removed a few PCRE remnants.
801 TF/07 Automatically extract Exim's version number from tags in the git
802 repository when doing development or release builds.
804 PP/02 Raise smtp_cmd_buffer_size to 16kB.
805 Bugzilla 879. Patch from Paul Fisher.
807 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
808 Heavily based on revision 40f9a89a from Simon Arlott's tree.
811 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
813 PP/05 Variable $av_failed, true if the AV scanner deferred.
814 Bugzilla 1078. Patch from John Horne.
816 PP/06 Stop make process more reliably on build failure.
817 Bugzilla 1087. Patch from Heiko Schlittermann.
819 PP/07 Make maildir_use_size_file an _expandable_ boolean.
820 Bugzilla 1089. Patch from Heiko Schlittermann.
822 PP/08 Handle ${run} returning more data than OS pipe buffer size.
823 Bugzilla 1131. Patch from Holger Weiß.
825 PP/09 Handle IPv6 addresses with SPF.
826 Bugzilla 860. Patch from Wolfgang Breyha.
828 PP/10 GnuTLS: support TLS 1.2 & 1.1.
830 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
833 PP/11 match_* no longer expand right-hand-side by default.
834 New compile-time build option, EXPAND_LISTMATCH_RHS.
835 New expansion conditions, "inlist", "inlisti".
837 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
839 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
841 PP/14 fix log_write() format string regression from TF/03.
842 Bugzilla 1152. Patch from Dmitry Isaikin.
848 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
850 PP/02 Harmonised TLS library version reporting; only show if debugging.
851 Layout now matches that introduced for other libraries in 4.74 PP/03.
853 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
855 PP/04 New "dns_use_edns0" global option.
857 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
860 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
861 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
863 TK/01 Updated PolarSSL code to 0.14.2.
864 Bugzilla 1097. Patch from Andreas Metzler.
866 PP/07 Catch divide-by-zero in ${eval:...}.
869 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
872 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
873 format-string attack -- SECURITY: remote arbitrary code execution.
875 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
876 time unintentionally subject to list matching rules, letting the header
877 cause arbitrary Exim lookups (of items which can occur in lists, *not*
878 arbitrary string expansion). This allowed for information disclosure.
880 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
881 INT_MIN/-1 -- value coerced to INT_MAX.
887 NM/01 Workround for PCRE version dependency in version reporting
890 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
891 This fixes portability to compilers other than gcc, notably
892 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
894 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
895 makefiles for portability to HP-UX and POSIX correctness.
897 PP/01 Permit LOOKUP_foo enabling on the make command-line.
898 Also via indented variable definition in the Makefile.
899 (Debugging by Oliver Heesakkers).
901 PP/02 Restore caching of spamd results with expanded spamd_address.
902 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
904 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
905 Improves build reliability. Fix from: Frank Elsner
907 NM/02 Fix wide character breakage in the rfc2047 coding
908 Fixes bug 1064. Patch from Andrey N. Oktyabrski
910 NM/03 Allow underscore in dnslist lookups
911 Fixes bug 1026. Patch from Graeme Fowler
913 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
914 Code patches from Adam Ciarcinski of NetBSD.
916 NM/04 Fixed exiqgrep to cope with mailq missing size issue
919 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
920 is logged, to avoid truncation. Patch from John Horne.
922 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
923 Patch from Jakob Hirsch.
925 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
926 SQL string expansion failure details.
927 Patch from Andrey Oktyabrski.
929 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
930 Patch from Simon Arlott.
932 PP/09 New lookups functionality failed to compile on old gcc which rejects
933 extern declarations in function scope.
934 Patch from Oliver Fleischmann
936 PP/10 Use sig_atomic_t for flags set from signal handlers.
937 Check getgroups() return and improve debugging.
938 Fixed developed for diagnosis in bug 927 (which turned out to be
941 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
942 Patch from Mark Zealey.
944 PP/12 Bugzilla 1056: Improved spamd server selection.
945 Patch from Mark Zealey.
947 PP/13 Bugzilla 1086: Deal with maildir quota file races.
948 Based on patch from Heiko Schlittermann.
950 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
951 Patch from Uwe Doering, sign-off by Michael Haardt.
953 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
954 variable declaration deep within a block. Bug and patch from
957 PP/15 lookups-Makefile IRIX compatibilty coercion.
959 PP/16 Make DISABLE_DKIM build knob functional.
961 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
962 Patch by Simon Arlott
964 TF/03 Fix valgrind.h portability to C89 compilers that do not support
965 variable argument macros. Our copy now differs from upstream.
971 TF/01 Failure to get a lock on a hints database can have serious
972 consequences so log it to the panic log.
974 TF/02 Log LMTP confirmation messages in the same way as SMTP,
975 controlled using the smtp_confirmation log selector.
977 TF/03 Include the error message when we fail to unlink a spool file.
979 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
980 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
981 for maintaining out-of-tree patches for some time.
983 PP/01 Bugzilla 139: Documentation and portability issues.
984 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
985 Handle per-OS dynamic-module compilation flags.
987 PP/02 Let /dev/null have normal permissions.
988 The 4.73 fixes were a little too stringent and complained about the
989 permissions on /dev/null. Exempt it from some checks.
990 Reported by Andreas M. Kirchwitz.
992 PP/03 Report version information for many libraries, including
993 Exim version information for dynamically loaded libraries. Created
994 version.h, now support a version extension string for distributors
995 who patch heavily. Dynamic module ABI change.
997 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
998 privilege escalation vulnerability whereby the Exim run-time user
999 can cause root to append content of the attacker's choosing to
1002 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1005 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1006 If dropping privileges for untrusted macros, we disabled normal logging
1007 on the basis that it would fail; for the Exim run-time user, this is not
1008 the case, and it resulted in successful deliveries going unlogged.
1009 Fixed. Reported by Andreas Metzler.
1015 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1016 only prepend for the Resent-* case. Fixes regression introduced in
1017 Exim 4.70 by NM/22 for Bugzilla 607.
1019 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1020 increasing numbers of administrators be bitten by this.
1022 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1024 PP/03 Bugzilla 994: added openssl_options main configuration option.
1026 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1028 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1030 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1032 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1033 without a peer certificate, leading to a segfault because of an
1034 assumption that peers always have certificates. Be a little more
1035 paranoid. Problem reported by Martin Tscholak.
1037 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1038 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1039 NB: ClamAV planning to remove STREAM in "middle of 2010".
1040 CL also introduces -bmalware, various -d+acl logging additions and
1041 more caution in buffer sizes.
1043 PP/09 Implemented reverse_ip expansion operator.
1045 PP/10 Bugzilla 937: provide a "debug" ACL control.
1047 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1049 PP/12 Bugzilla 973: Implement --version.
1051 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1053 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1055 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1057 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1058 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1059 ignore trailing whitespace.
1061 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1063 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1066 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1067 Notification from Dr Andrew Aitchison.
1069 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1070 ExtendedDetectionInfo response format.
1071 Notification from John Horne.
1073 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1076 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1077 XSL and documented dependency on system catalogs, with examples of how
1080 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1083 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1084 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1085 configuration file which is writeable by the Exim user or group.
1087 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1088 of configuration files to cover files specified with the -C option if
1089 they are going to be used with root privileges, not just the default
1092 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1093 option (effectively making it always true).
1095 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1096 files to be used while preserving root privileges.
1098 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1099 that rogue child processes cannot use them.
1101 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1102 run-time user, instead of root.
1104 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1105 Exim run-time user without dropping privileges.
1107 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1108 result string, instead of calling string_vformat() twice with the same
1111 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1112 for other users. Others should always drop root privileges if they use
1113 -C on the command line, even for a whitelisted configure file.
1115 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1117 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1123 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1124 $data_path, and $header_path variables; fixed documentation bugs and
1127 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1128 exipick to access non-standard spools, including the "frozen" queue
1131 NM/01 Bugzilla 965: Support mysql stored procedures.
1132 Patch from Alain Williams
1134 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1136 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1137 Patch from Andreas Metzler
1139 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1140 Patch from Kirill Miazine
1142 NM/05 Bugzilla 671: Added umask to procmail example.
1144 JJ/03 installed exipick 20100323.0, fixing doc bug
1146 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1147 directory. Notification and patch from Dan Rosenberg.
1149 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1151 TK/02 Improve log output when DKIM signing operation fails.
1153 MH/01 Treat the transport option dkim_domain as a colon separated
1154 list, not as a single string, and sign the message with each element,
1155 omitting multiple occurences of the same signer.
1157 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1158 Bugzilla 985, 986. Patch by Simon Arlott
1160 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1161 Patch by Simon Arlott
1163 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1164 MBX locking. Notification from Dan Rosenberg.
1170 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1172 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1174 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1176 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1178 NM/04 Bugzilla 915: Flag broken perl installation during build.
1184 TK/01 Added patch by Johannes Berg that expands the main option
1185 "spamd_address" if it starts with a dollar sign.
1187 TK/02 Write list of recipients to X-Envelope-Sender header when building
1188 the mbox-format spool file for content scanning (suggested by Jakob
1191 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1192 (http://www.dcc-servers.net/) support via dccifd. Activated by
1193 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1195 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1196 by Mark Daniel Reidel <mr@df.eu>.
1198 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1199 When building exim an external PCRE library is now needed -
1200 PCRE is a system library on the majority of modern systems.
1201 See entry on PCRE_LIBS in EDITME file.
1203 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1204 conversation. Added nologin parameter to request.
1205 Patch contributed by Kirill Miazine.
1207 TF/01 Do not log submission mode rewrites if they do not change the address.
1209 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1211 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1212 log files in place. Contributed by Roberto Lima.
1214 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1216 TF/03 Bugzilla 615: When checking the local_parts router precondition
1217 after a local_part_suffix or local_part_prefix option, Exim now
1218 does not use the address's named list lookup cache, since this
1219 contains cached lookups for the whole local part.
1221 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1222 Robert Millan. Documentation is in experimental-spec.txt.
1224 TF/04 Bugzilla 668: Fix parallel build (make -j).
1226 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1228 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1229 Patch provided by Jan Srzednicki.
1231 TF/05 Leading white space used to be stripped from $spam_report which
1232 wrecked the formatting. Now it is preserved.
1234 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1235 that they are available at delivery time.
1237 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1239 TF/08 TLS error reporting now respects the incoming_interface and
1240 incoming_port log selectors.
1242 TF/09 Produce a more useful error message if an SMTP transport's hosts
1243 setting expands to an empty string.
1245 NM/06 Bugzilla 744: EXPN did not work under TLS.
1246 Patch provided by Phil Pennock.
1248 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1249 Patch provided by Richard Godbee.
1251 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1252 acl_smtp_notquit, added index entry.
1254 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1255 Patch provided by Eugene Bujak.
1257 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1258 accept(). Patch provided by Maxim Dounin.
1260 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1261 Patch provided by Phil Pennock.
1263 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1265 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1266 Patch provided by Brad "anomie" Jorsch.
1268 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1269 Patch provided by Dean Brooks.
1271 TK/05 Add native DKIM support (does not depend on external libraries).
1273 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1274 Patch provided by Graeme Fowler.
1276 NM/16 Bugzilla 851: Documentation example syntax fix.
1278 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1280 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1283 NM/19 Bugzilla 745: TLS version reporting.
1284 Patch provided by Phil Pennock.
1286 NM/20 Bugzilla 167: bool: condition support.
1287 Patch provided by Phil Pennock.
1289 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1290 clients. Patch provided by Phil Pennock.
1292 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1293 Patch provided by Brad "anomie" Jorsch.
1295 NM/23 Bugzilla 687: Fix misparses in eximstats.
1296 Patch provided by Heiko Schlittermann.
1298 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1299 Patch provided by Heiko Schlittermann.
1301 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1302 plus update to original patch.
1304 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1306 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1307 Patch provided by David Brownlee.
1309 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1311 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1313 NM/30 Bugzilla 888: TLS documentation bugfixes.
1315 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1317 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1318 Unlike the original bugzilla I have changed all shell scripts in src tree.
1320 NM/33 Bugzilla 898: Transport filter timeout fix.
1321 Patch by Todd Rinaldo.
1323 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1324 Patch by Serge Demonchaux.
1326 NM/35 Bugzilla 39: Base64 decode bug fixes.
1327 Patch by Jakob Hirsch.
1329 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1331 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1333 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1335 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1341 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1342 ALT-N's libdkim that I have put here:
1343 http://duncanthrax.net/exim-experimental/
1345 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1346 were called 'true' and it seems that C99 defines that as a reserved
1347 keyword to be used with 'bool' variable types. That means you could
1348 not include C99-style headers which use bools without triggering
1349 build errors in sieve.c.
1351 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1352 as mailq or other aliases. Changed the --help handling significantly
1353 to do whats expected. exim_usage() emits usage/help information.
1355 SC/01 Added the -bylocaldomain option to eximstats.
1357 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1359 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1361 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1367 PH/01 Another patch from the Sieve maintainer.
1369 PH/02 When an IPv6 address is converted to a string for single-key lookup
1370 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1371 dots are used instead of colons so that keys in lsearch files need not
1372 contain colons. This was done some time before quoting was made available
1373 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1374 (notated using the quote facility) so as to distinguish them from IPv4
1375 keys. This meant that lookups for IP addresses in host lists did not work
1376 for iplsearch lookups.
1378 This has been fixed by arranging for IPv6 addresses to be expressed with
1379 colons if the lookup type is iplsearch. This is not incompatible, because
1380 previously such lookups could never work.
1382 The situation is now rather anomolous, since one *can* have colons in
1383 ordinary lsearch keys. However, making the change in all cases is
1384 incompatible and would probably break a number of configurations.
1386 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1389 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1390 conversion specification without a maximum field width, thereby enabling
1391 a rogue spamd server to cause a buffer overflow. While nobody in their
1392 right mind would setup Exim to query an untrusted spamd server, an
1393 attacker that gains access to a server running spamd could potentially
1394 exploit this vulnerability to run arbitrary code as the Exim user.
1396 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1397 $primary_hostname instead of what libspf2 thinks the hosts name is.
1399 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1400 a directory entry by the name of the lookup key. Previously, if a
1401 symlink pointed to a non-existing file or a file in a directory that
1402 Exim lacked permissions to read, a lookup for a key matching that
1403 symlink would fail. Now it is enough that a matching directory entry
1404 exists, symlink or not. (Bugzilla 503.)
1406 PH/03 The body_linecount and body_zerocount variables are now exported in the
1409 PH/04 Added the $dnslist_matched variable.
1411 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1412 This means they are set thereafter only if the connection becomes
1415 PH/06 Added the client_condition to authenticators so that some can be skipped
1416 by clients under certain conditions.
1418 PH/07 The error message for a badly-placed control=no_multiline_responses left
1419 "_responses" off the end of the name.
1421 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1423 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1424 (without spaces) instead of just copying the configuration text.
1426 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1428 PH/11 Added $max_received_linelength.
1430 PH/12 Added +ignore_defer and +include_defer to host lists.
1432 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1433 way in which PCRE > 7.0 is built.
1435 PH/14 Implemented queue_only_load_latch.
1437 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1438 MAIL command. The effect was to mangle the value on 64-bit systems.
1440 PH/16 Another patch from the Sieve maintainer.
1442 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1444 PH/18 If a system quota error occurred while trying to create the file for
1445 a maildir delivery, the message "Mailbox is full" was not appended to the
1446 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1447 only to a quota excession during the actual writing of the file.
1449 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1450 characters?) which causes problems in log lines. The DN values are now
1451 passed through string_printing() before being added to log lines.
1453 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1454 and InterBase are left for another time.)
1456 PH/21 Added message_body_newlines option.
1458 PH/22 Guard against possible overflow in moan_check_errorcopy().
1460 PH/23 POSIX allows open() to be a macro; guard against that.
1462 PH/24 If the recipient of an error message contained an @ in the local part
1463 (suitably quoted, of course), incorrect values were put in $domain and
1464 $local_part during the evaluation of errors_copy.
1470 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1471 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1474 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1475 issue a MAIL command.
1477 PH/02 In an ACL statement such as
1479 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1481 if a client was not listed at all, or was listed with a value other than
1482 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1483 the condition was not true (as it should be), so access was not denied.
1484 The bug was that the ! inversion was incorrectly passed on to the second
1485 item. This has been fixed.
1487 PH/03 Added additional dnslists conditions == and =& which are different from
1488 = and & when the dns lookup returns more than one IP address.
1490 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1491 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1493 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1494 FSYNC, which compiles an option called disable_fsync that allows for
1495 bypassing fsync(). The documentation is heavily laced with warnings.
1497 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1499 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1500 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1501 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1502 including adding "make clean"; (3) Added -fPIC when compiling the test
1503 dynamically loaded module, to get rid of a warning.
1505 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1506 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1507 = 0s. The bug is otherwise harmless.
1509 PH/07 There was a bug in the dovecot authenticator such that the value of
1510 $auth1 could be overwritten, and so not correctly preserved, after a
1511 successful authentication. This usually meant that the value preserved by
1512 the server_setid option was incorrect.
1514 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1516 PH/09 Installed PCRE release 7.0.
1518 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1519 run for batched SMTP input. It is now run at the start of every message
1520 in the batch. While fixing this I discovered that the process information
1521 (output by running exiwhat) was not always getting set for -bs and -bS
1522 input. This is fixed, and it now also says "batched" for BSMTP.
1524 PH/11 Added control=no_pipelining.
1526 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1527 patch, slightly modified), and move the expansion of helo_data till after
1528 the connection is made in the smtp transport (so it can use these
1531 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1533 PH/14 Added log_selector = +pid.
1535 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1537 PH/16 Add ${if forany and ${if forall.
1539 PH/17 Added dsn_from option to vary the From: line in DSNs.
1541 PH/18 Flush SMTP output before performing a callout, unless control =
1542 no_callout_flush is set.
1544 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1545 was true (the default) a successful delivery failed to delete the retry
1546 item, thus causing premature timeout of the address. The bug is now
1549 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1551 PH/21 Long custom messages for fakedefer and fakereject are now split up
1552 into multiline reponses in the same way that messages for "deny" and
1553 other ACL rejections are.
1555 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1556 with slight modification.
1558 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1559 draft, changing the syntax and factoring some duplicate code".
1561 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1562 for deliveries of the second and subsequent messages over the same SMTP
1565 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1566 ${reduce, with only minor "tidies".
1568 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1570 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1571 expansion side effects.
1573 PH/27 When a message times out after an over-quota error from an Exim-imposed
1574 quota, the bounce message says "mailbox is full". This message was not
1575 being given when it was a system quota that was exceeded. It now should
1578 MH/03 Made $recipients available in local_scan(). local_scan() already has
1579 better access to the recipient list through recipients_list[], but
1580 $recipients can be useful in postmaster-provided expansion strings.
1582 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1583 in the case of a MAIL command with additional options following the
1584 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1585 were accidentally chopped off.
1587 PH/29 SMTP synchronization checks are implemented when a command is read -
1588 there is a check that no more input is waiting when there shouldn't be
1589 any. However, for some commands, a delay in an ACL can mean that it is
1590 some time before the response is written. In this time, more input might
1591 arrive, invalidly. So now there are extra checks after an ACL has run for
1592 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1593 pipelining has not been advertised.
1595 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1597 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1598 returned in the SMTP response when smtp_return_error_details was false.
1599 This has been fixed.
1601 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1602 instead of the C I/O that was originally supplied, because problems were
1603 reported on Solaris.
1605 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1606 Exim which did not show up earlier: it was assuming that a call to
1607 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1608 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1609 was a macro that became an assignment, so it seemed to work. This has
1610 changed to a proper function call with a void return, hence the compile
1611 error. Exim's code has been fixed.
1613 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1616 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1618 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1620 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1621 because it turns out to be too big or there is a timeout) there is no
1622 "Completed" line in the log. When some messages of this type were
1623 selected by exigrep, they were listed as "not completed". Others were
1624 picked up by some special patterns. I have improved the selection
1625 criteria to be more general.
1627 PH/38 The host_find_failed option in the manualroute router can now be set
1628 to "ignore", to completely ignore a host whose IP address cannot be
1629 found. If all hosts are ignored, the behaviour is controlled by the new
1630 host_all_ignored option.
1632 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1633 homing or because of multiple MX records with /mx) generated more than
1634 one IP address, and the following item turned out to be the local host,
1635 all the secondary addresses of the first item were incorrectly removed
1636 from the list, along with the local host and any following hosts (which
1637 is what is supposed to happen).
1639 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1640 whoever called Exim into the -H file. In the case of the daemon it was
1641 behaving confusingly. When first started, it used values for whoever
1642 started the daemon, but after a SIGHUP it used the Exim user (because it
1643 calls itself on a restart). I have changed the code so that it now always
1646 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1647 message are rejected with the same error (e.g. no authentication or bad
1648 sender address), and a DATA command is nevertheless sent (as can happen
1649 with PIPELINING or a stupid MUA), the error message that was given to the
1650 RCPT commands is included in the rejection of the DATA command. This is
1651 intended to be helpful for MUAs that show only the final error to their
1654 PH/42 Another patch from the Sieve maintainer.
1656 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1657 Eximstats - Fixed some broken HTML links and added missing column headers
1659 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1660 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1662 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1664 PH/43 Yet another patch from the Sieve maintainer.
1666 PH/44 I found a way to check for a TCP/IP connection going away before sending
1667 the response to the final '.' that terminates a message, but only in the
1668 case where the client has not sent further data following the '.'
1669 (unfortunately, this is allowed). However, in many cases there won't be
1670 any further data because there won't be any more messages to send. A call
1671 to select() can be used: if it shows that the input is "ready", there is
1672 either input waiting, or the socket has been closed. An attempt to read
1673 the next input character can distinguish the two cases. Previously, Exim
1674 would have sent an OK response which the client would never have see.
1675 This could lead to message repetition. This fix should cure that, at
1676 least in a lot of common cases.
1678 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1679 advertised in response to EHLO.
1685 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1686 fixed by 4.65/MH/01 (is this a record?) are fixed:
1688 (i) An empty string was always treated as zero by the numeric comparison
1689 operators. This behaviour has been restored.
1691 (ii) It is documented that the numeric comparison operators always treat
1692 their arguments as decimal numbers. This was broken in that numbers
1693 starting with 0 were being interpreted as octal.
1695 While fixing these problems I realized that there was another issue that
1696 hadn't been noticed. Values of message_size_limit (both the global option
1697 and the transport option) were treated as octal if they started with 0.
1698 The documentation was vague. These values are now always treated as
1699 decimal, and I will make that clear in the documentation.
1705 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1706 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1709 MH/01 Don't check that the operands of numeric comparison operators are
1710 integers when their expansion is in "skipping" mode (fixes bug
1711 introduced by 4.64-PH/07).
1713 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1714 child addresses, Exim now panics and dies. Previously, because the count
1715 is held in a short int, deliveries were likely to be lost. As such a
1716 large number of recipients for a single message is ridiculous
1717 (performance will be very, very poor), I have chosen to impose a limit
1718 rather than extend the field.
1724 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1725 leftover -K file (the existence of which was triggered by #402).
1726 While we were at it, introduced process PID as part of the -K
1727 filename. This should rule out race conditions when creating
1730 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1731 processing considerably. Previous code took too long for large mails,
1732 triggering a timeout which in turn triggers #401.
1734 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1735 in the DK code in transports.c. sendfile() is not really portable,
1736 hence the _LINUX specificness.
1738 TF/01 In the add_headers option to the mail command in an Exim filter,
1739 there was a bug that Exim would claim a syntax error in any
1740 header after the first one which had an odd number of characters
1743 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1744 callout verification, Exim cached a "reject" for the entire domain. This
1745 is correct for most verifications, but it is not correct for a recipient
1746 verification with use_sender or use_postmaster set, because in that case
1747 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1748 case of MAIL FROM:<> rejection from other early rejections (e.g.
1749 rejection of HELO). When verifying a recipient using a non-null MAIL
1750 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1751 Whatever the result of the callout, the value of the domain cache is
1752 left unchanged (for any other kind of callout, getting as far as trying
1753 RCPT means that the domain itself is ok).
1755 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1758 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1759 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1760 session, the daemon ignores ECONNECT errors and logs others; it now
1761 ignores EPIPE as well.
1763 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1764 (quoted-printable decoding).
1766 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1767 later the small subsequent patch to fix an introduced bug.
1769 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1771 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1773 PH/08 An error is now given if message_size_limit is specified negative.
1775 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1776 to be given (somewhat) arbitrary names.
1778 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1781 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1782 miscellaneous code fixes
1784 PH/10 Added the log_reject_target ACL modifier to specify where to log
1787 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1788 hostname. This is wrong, because it relates to the incoming message (and
1789 probably the interface on which it is arriving) and not to the outgoing
1790 callout (which could be using a different interface). This has been
1791 changed to use the value of the helo_data option from the smtp transport
1792 instead - this is what is used when a message is actually being sent. If
1793 there is no remote transport (possible with a router that sets up host
1794 addresses), $smtp_active_hostname is used.
1796 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1797 tweaks were necessary in order to get it to work (see also 21 below):
1798 (a) The code assumed that strncpy() returns a negative number on buffer
1799 overflow, which isn't the case. Replaced with Exim's string_format()
1801 (b) There were several signed/unsigned issues. I just did the minimum
1802 hacking in of casts. There is scope for a larger refactoring.
1803 (c) The code used strcasecmp() which is not a standard C function.
1804 Replaced with Exim's strcmpic() function.
1805 (d) The code set only $1; it now sets $auth1 as well.
1806 (e) A simple test gave the error "authentication client didn't specify
1807 service in request". It would seem that Dovecot has changed its
1808 interface. Fortunately there's a specification; I followed it and
1809 changed what the client sends and it appears to be working now.
1811 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1814 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1815 address A is aliased to B and C, where B exists and C does not. Without
1816 -v the output is "A verified" because verification stops after a
1817 successful redirection if more than one address is generated. However,
1818 with -v the child addresses are also verified. Exim was outputting "A
1819 failed to verify" and then showing the successful verification for C,
1820 with its parentage. It now outputs "B failed to verify", showing B's
1821 parentage before showing the successful verification of C.
1823 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1824 look up a TXT record in a specific list after matching in a combined
1827 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1828 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1829 they consult the DNS. I had assumed they would set it the way they
1830 wanted; and indeed my experiments on Linux seem to show that in some
1831 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1832 To be on the safe side, however, I have now made the interface to
1833 host_find_byname() similar to host_find_bydns(), with an argument
1834 containing the DNS resolver options. The host_find_byname() function now
1835 sets these options at its start, just as host_find_bydns() does. The smtp
1836 transport options dns_qualify_single and dns_search_parents are passed to
1837 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1838 of host_find_byname() use the default settings of RES_DEFNAMES
1839 (qualify_single) but not RES_DNSRCH (search_parents).
1841 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1842 spool_read_header() do less string testing, by means of a preliminary
1843 switch on the second character of optional "-foo" lines. (This is
1844 overdue, caused by the large number of possibilities that now exist.
1845 Originally there were few.) While I was there, I also converted the
1846 str(n)cmp tests so they don't re-test the leading "-" and the first
1847 character, in the hope this might squeeze out yet more improvement.
1849 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1850 flag allowing group syntax was set by the header_syntax check but not
1851 turned off, possible causing trouble later; (2) The flag was not being
1852 set at all for the header_verify test, causing "group"-style headers to
1853 be rejected. I have now set it in this case, and also caused header_
1854 verify to ignore an empty address taken from a group. While doing this, I
1855 came across some other cases where the code for allowing group syntax
1856 while scanning a header line wasn't quite right (mostly, not resetting
1857 the flag correctly in the right place). These bugs could have caused
1858 trouble for malformed header lines. I hope it is now all correct.
1860 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1861 with the "reply" argument non-NULL. The code, however (which originally
1862 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1863 but it didn't always do it. This confused somebody who was copying the
1864 code for some other use. I have removed all the tests.
1866 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1867 feature that was used to support insecure browsers during the U.S. crypto
1868 embargo. It requires special client support, and Exim is probably the
1869 only MTA that supported it -- and would never use it because real RSA is
1870 always available. This code has been removed, because it had the bad
1871 effect of slowing Exim down by computing (never used) parameters for the
1872 RSA_EXPORT functionality.
1874 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1875 authenticator to fail if there's a tab character in the incoming data
1876 (there should never be unless someone is messing about, as it's supposed
1877 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1878 if the connection is using TLS or if the remote IP is the same as the
1879 local IP, and the "valid-client-cert option" if a client certificate has
1882 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1883 authenticators. This can be used for authorization after authentication
1884 succeeds. (In the case of plaintext, it servers for both authentication
1887 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1888 if any retry times were supplied.
1890 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1891 connection, where there is no client IP address to check. In this
1892 situation, the verify now always succeeds.
1894 PH/25 Applied John Jetmore's -Mset patch.
1896 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1898 PH/27 In a string expansion for a processed (not raw) header when multiple
1899 headers of the same name were present, leading whitespace was being
1900 removed from all of them, but trailing whitespace was being removed only
1901 from the last one. Now trailing whitespace is removed from each header
1902 before concatenation. Completely empty headers in a concatenation (as
1903 before) are ignored.
1905 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1906 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1908 PH/29 [Removed. This was a change that I later backed out, and forgot to
1909 correct the ChangeLog entry (that I had efficiently created) before
1910 committing the later change.]
1912 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1913 address errors (4xx response to RCPT) over the same connection as other
1914 messages routed to the same hosts. Such deliveries are always "forced",
1915 so retry times are not inspected. This resulted in far too many retries
1916 for the affected addresses. The effect occurred only when there were more
1917 hosts than the hosts_max_try setting in the smtp transport when it had
1918 the 4xx errors. Those hosts that it had tried were not added to the list
1919 of hosts for which the message was waiting, so if all were tried, there
1920 was no problem. Two fixes have been applied:
1922 (i) If there are any address or message errors in an SMTP delivery, none
1923 of the hosts (tried or untried) are now added to the list of hosts
1924 for which the message is waiting, so the message should not be a
1925 candidate for sending over the same connection that was used for a
1926 successful delivery of some other message. This seems entirely
1927 reasonable: after all the message is NOT "waiting for some host".
1928 This is so "obvious" that I'm not sure why it wasn't done
1929 previously. Hope I haven't missed anything, but it can't do any
1930 harm, as the worst effect is to miss an optimization.
1932 (ii) If, despite (i), such a delivery is accidentally attempted, the
1933 routing retry time is respected, so at least it doesn't keep
1934 hammering the server.
1936 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1937 in ${readsocket because some servers need this prod.
1939 PH/32 Added some extra debug output when updating a wait-xxx database.
1941 PH/33 The hint "could be header name not terminated by colon", which has been
1942 given for certain expansion errors for a long time, was not being given
1943 for the ${if def:h_colon_omitted{... case.
1945 PH/34 The spec says: "With one important exception, whenever a domain list is
1946 being scanned, $domain contains the subject domain." There was at least
1947 one case where this was not true.
1949 PH/35 The error "getsockname() failed: connection reset by peer" was being
1950 written to the panic log as well as the main log, but it isn't really
1951 panic-worthy as it just means the connection died rather early on. I have
1952 removed the panic log writing for the ECONNRESET error when getsockname()
1955 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1956 runs only) independently of the message's sender address. This meant
1957 that, if the 4xx error was in fact related to the sender, a different
1958 message to the same recipient with a different sender could confuse
1959 things. In particualar, this can happen when sending to a greylisting
1960 server, but other circumstances could also provoke similar problems.
1961 I have changed the default so that the retry time for these errors is now
1962 based a combination of the sender and recipient addresses. This change
1963 can be overridden by setting address_retry_include_sender=false in the
1966 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1967 remote server are returned as part of bounce messages. This was not
1968 happening for LMTP over a pipe (the lmtp transport), but now it is the
1969 same for both kinds of LMTP.
1971 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1972 in header lines that were in fact CNAMEs. This is no longer the case.
1974 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1975 and queue runs started by the daemon processed all messages. This has
1976 been fixed so that -R and -S can now usefully be given with -q<time>.
1978 PH/40 Import PCRE release 6.7 (fixes some bugs).
1980 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1982 PH/42 Give an error if -q is specified more than once.
1984 PH/43 Renamed the variables $interface_address and $interface_port as
1985 $received_ip_address and $received_port, to make it clear that these
1986 values apply to message reception, and not to the outgoing interface when
1987 a message is delivered. (The old names remain recognized, of course.)
1989 PH/44 There was no timeout on the connect() call when using a Unix domain
1990 socket in the ${readsocket expansion. There now is.
1992 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1993 be meaningful with "accept".
1995 SC/01 Eximstats V1.43
1996 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1998 SC/02 Eximstats V1.44
1999 Use a glob alias rather than an array ref in the generated
2000 parser. This improves both readability and performance.
2002 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2003 Collect SpamAssassin and rejection statistics.
2004 Don't display local sender or destination tables unless
2005 there is data to show.
2006 Added average volumes into the top table text output.
2008 SC/04 Eximstats V1.46
2009 Collect data on the number of addresses (recipients)
2010 as well as the number of messages.
2012 SC/05 Eximstats V1.47
2013 Added 'Message too big' to the list of mail rejection
2014 reasons (thanks to Marco Gaiarin).
2016 SC/06 Eximstats V1.48
2017 Mainlog lines which have GMT offsets and are too short to
2018 have a flag are now skipped.
2020 SC/07 Eximstats V1.49 (Alain Williams)
2021 Added the -emptyok flag.
2023 SC/08 Eximstats V1.50
2024 Fixes for obtaining the IP address from reject messages.
2026 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2027 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2028 whitesspace changes from 4.64-PH/27
2030 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2033 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2034 are found, allow negative numbers in numeric criteria)
2036 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2038 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2039 to match changes made in 4.64-PH/43
2041 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2043 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2044 "domain" is actually an IP address, and force a failure. This locks out
2045 those revolvers/nameservers that support "A-for-A" lookups, in
2046 contravention of the specifications.
2048 PH/48 When a host name was looked up from an IP address, and the subsequent
2049 forward lookup of the name timed out, the host name was left in
2050 $sender_host_name, contrary to the specification.
2052 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2053 restricted to single-key lookups, Exim was not diagnosing an error if
2054 * or *@ was used with a query-style lookup.
2056 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2058 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2059 long ago, but noone interested enough thought of it. Let's just say that
2060 the "1.1" means that there are some new functions that weren't there at
2061 some point in the past.
2063 PH/51 Error processing for expansion failure of helo_data from an smtp
2064 transport during callout processing was broken.
2066 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2067 tested/used via the -bh/-bhc/-bs options.
2069 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2070 bug, fixed in subsequent PCRE releases).
2072 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2073 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2075 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2081 SC/01 Use a glob alias rather than an array ref in eximstats generated
2082 parser. This improves both readability and performance.
2084 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2085 Don't display local sender or destination tables in eximstats unless
2086 there is data to show.
2087 Added average volumes into the eximstats top table text output.
2089 SC/03 Collect data on the number of addresses (recipients) as well
2090 as the number of messages in eximstats.
2092 TF/01 Correct an error in the documentation for the redirect router. Exim
2093 does (usually) call initgroups() when daemonizing.
2095 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2096 with consistent privilege compared to when running as a daemon.
2098 TF/03 Note in the spec that $authenticated_id is not set for local
2099 submissions from trusted users.
2101 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2102 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2104 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2105 by adding some example configuration directives to the default
2106 configuration file. A little bit of work is required to uncomment the
2107 directives and define how usernames and passwords are checked, but
2108 there is now a framework to start from.
2110 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2111 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2112 without this. I don't know how relevant this is to other LDAP libraries.
2114 PH/02 Add the verb name to the "unknown ACL verb" error.
2116 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2118 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2120 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2121 directory not expanded when it should be if an expanded home directory
2122 was set for the address (which is overridden by the transport).
2124 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2127 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2128 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2129 because it is too late at that time, and has no effect.
2131 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2132 security issue with \' (bugzilla #107). I could not use the
2133 PQescapeStringConn() function, because it needs a PGconn value as one of
2136 PH/08 When testing addresses using -bt, indicate those final addresses that
2137 are duplicates that would not cause an additional delivery. At least one
2138 person was confused, thinking that -bt output corresponded to deliveries.
2139 (Suppressing duplicates isn't a good idea as you lose the information
2140 about possibly different redirections that led to the duplicates.)
2142 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2143 systems where poll() doesn't work, in particular OS X.
2145 PH/10 Added more information to debugging output for retry time not reached.
2147 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2148 operations in malware.c.
2150 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2153 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2154 syslog_duplication set false, log lines that would normally be written
2155 both the the main log and to the reject log were not written to syslog at
2158 PH/14 In the default configuration, change the use of "message" in ACL warn
2159 statements to "add_header".
2161 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2162 not followed by a command (e.g. "seen endif").
2164 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2165 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2168 PH/17 Added extra conditions to the default value of delay_warning_condition
2172 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2173 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2174 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2177 The Auto-Submitted: and various List- headers are standardised, whereas I
2178 don't think Precedence: ever was.
2180 PH/18 Refactored debugging code in route_finduser() to show more information,
2181 in particular, the error code if getpwnam() issues one.
2183 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2184 This is apparently needed in addition to the PH/07 change above to avoid
2185 any possible encoding problems.
2187 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2188 but not after initializing Perl.
2190 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2191 output them only if debugging. By default they are written stderr,
2192 apparently, which is not desirable.
2194 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2197 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2200 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2202 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2203 authenticated or an ident call has been made. Suppress the default
2204 values for $authenticated_id and $authenticated_sender (but permit -oMai
2205 and -oMas) when testing with -bh.
2207 PH/24 Re-jigged the order of the tests in the default configuration so that the
2208 tests for valid domains and recipients precede the DNS black list and CSA
2209 tests, on the grounds that those ones are more expensive.
2211 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2212 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2213 This bug exists in every version of Exim that I still have, right back to
2216 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2217 However, an attempt to turn on case-sensitivity in a regex key by
2218 including (?-i) didn't work because the subject string was already
2219 lowercased, and the effects were non-intuitive. It turns out that a
2220 one-line patch can be used to allow (?-i) to work as expected.
2226 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2227 other effects) broke the use of negated acl sub-conditions.
2229 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2232 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2233 "Deny" causes Exim to reject the incoming connection with a 554 error.
2234 Unfortunately, if there is a major crisis, such as a disk failure,
2235 tcp-wrappers gives "deny", whereas what one would like would be some
2236 kind of temporary error. A kludge has been added to help with this.
2237 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2238 554 error is used if errno is still zero or contains ENOENT (which occurs
2239 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2242 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2244 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2245 errors. Otherwise a message that provokes a temporary error (when other
2246 messages do not) can cause a whole host to time out.
2248 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2249 addresses were routed directly to files or pipes from a redirect router.
2250 File deliveries just didn't batch; pipe deliveries might have suffered
2253 PH/06 A failure to get a lock for a hints database would erroneously always say
2254 "Failed to get write lock", even when it was really a read lock.
2256 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2257 of 0600. This has been changed to use the value of the lockfile_mode
2258 option (which defaults to 0600).
2260 PH/08 Applied small patch from the Sieve maintainer.
2262 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2263 folder from quota calculations, a direct delivery into this folder messed
2264 up the contents of the maildirsize file. This was because the regex was
2265 used only to exclude .Trash (or whatever) when the size of the mailbox
2266 was calculated. There was no check that a delivery was happening into an
2267 excluded directory. This bug has been fixed by ignoring all quota
2268 processing for deliveries into excluded directories.
2270 PH/10 Added the maildirfolder_create_regex option to appendfile.
2276 PH/01 The code for finding all the local interface addresses on a FreeBSD
2277 system running IPv6 was broken. This may well have applied to all BSD
2278 systems, as well as to others that have similar system calls. The broken
2279 code found IPv4 interfaces correctly, but gave incorrect values for the
2280 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2281 that it would not match correctly against @[] and not recognize the IPv6
2284 PH/02 The ipliteral router was not recognizing addresses of the form user@
2285 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2287 PH/03 Added disable_ipv6.
2289 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2290 decoded form, because it is most often used to construct To: headers
2291 lines in autoreplies, and the decoded form may well be syntactically
2292 invalid. However, $reply_address has leading white space removed, and all
2293 newlines turned into spaces so that the autoreply transport does not
2296 PH/05 If group was specified without a user on a router, and no group or user
2297 was specified on a transport, the group from the router was ignored.
2299 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2300 for visible compile-time settings that can be used to change these
2301 numbers, for those that want even more. Backwards compatibility with old
2302 spool files has been maintained. However, going back to a previous Exim
2303 release will lost any variables that are in spool files.
2305 PH/07 Two small changes when running in the test harness: increase delay when
2306 passing a TCP/IP connection to a new process, in case the original
2307 process has to generate a bounce, and remove special handling of
2308 127.0.0.2 (sic), which is no longer necessary.
2310 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2311 be the same on different OS.
2313 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2316 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2317 whether --show-vars was specified or not
2319 JJ/02 exipick: Added support for new ACL variable spool format introduced
2322 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2323 syntactically invalid From: or Reply-to: line, and a filter used this to
2324 generate an autoreply, and therefore failed to obtain an address for the
2325 autoreply, Exim could try to deliver to a non-existent relative file
2326 name, causing unrelated and misleading errors. What now happens is that
2327 it logs this as a hard delivery error, but does not attempt to create a
2330 PH/11 The exinext utility has a -C option for testing purposes, but although
2331 the given file was scanned by exinext itself; it wasn't being passed on
2332 when Exim was called.
2334 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2335 an end-of-file indication when reading a command response.
2337 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2338 compiled. In many other places in Exim, IPv6 addresses are always
2339 recognized, so I have changed this. It also means that IPv4 domain
2340 literals of the form [IPV4:n.n.n.n] are now always recognized.
2342 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2343 used if the router is not running as root, for example, when verifying at
2344 ACL time, or when using -bh. The debugging output from this situation was
2345 non-existent - all you got was a failure to exec. I have made two
2348 (a) Failures to set uid/gid, the current directory, or a process leader
2349 in a subprocess such as that created by queryprogram now generate
2350 suitable debugging ouput when -d is set.
2352 (b) The queryprogram router detects when it is not running as root,
2353 outputs suitable debugging information if -d is set, and then runs
2354 the subprocess without attempting to change uid/gid.
2356 PH/15 Minor change to Makefile for building test_host (undocumented testing
2359 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2360 additional section of a DNS packet that returns MX or SRV records.
2361 Instead, it always explicitly searches for A/AAAA records. This avoids
2362 major problems that occur when a DNS server includes only records of one
2363 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2364 fixed another bug: if SRV records were looked up and the corresponding
2365 address records were *not* found in the additional section, the port
2366 values from the SRV records were lost.
2368 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2369 using the correct key (the original address) when searching the retry
2370 rules in order to find which one to use for generating the retry hint.
2372 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2373 adding the default one. Similarly, if it contains a Reply-To: header, the
2374 errors_reply_to option, if set, is not used.
2376 PH/19 When calculating a retry time, Exim used to measure the "time since
2377 failure" by looking at the "first failed" field in the retry record. Now
2378 it does not use this if it is later than than the arrival time of the
2379 message. Instead it uses the arrival time. This makes for better
2380 behaviour in cases where some deliveries succeed, thus re-setting the
2381 "first failed" field. An example is a quota failure for a huge message
2382 when small messages continue to be delivered. Without this change, the
2383 "time since failure" will always be short, possible causing more frequent
2384 delivery attempts for the huge message than are intended.
2385 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2387 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2388 $1, $2, $3) because the numerical variables can be reset during some
2389 expansion items (e.g. "match"), thereby losing the authentication data.
2391 PH/21 Make -bV show the size of off_t variables so that the test suite can
2392 decide whether to run tests for quotas > 2G.
2394 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2395 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2396 filecount value is greater than 2G or if a quota value is greater than 2G
2397 on a system where the size of off_t is not greater than 4, a panic error
2400 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2401 never match. The debug and -bh output now contains an explicit error
2402 message indicating a malformed IPv4 address or mask.
2404 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2405 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2406 PH/23 above applies.
2408 PH/25 Do not write to syslog when running in the test harness. The only
2409 occasion when this arises is a failure to open the main or panic logs
2410 (for which there is an explicit test).
2412 PH/26 Added the /no_tell option to "control=freeze".
2414 PH/27 If a host name lookup failed very early in a connection, for example, if
2415 the IP address matched host_lookup and the reverse lookup yielded a name
2416 that did not have a forward lookup, an error message of the form "no IP
2417 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2418 could be logged. Now it outputs the IP address instead of "NULL".
2420 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2421 allows the sender and the authenticated sender to be set when
2422 submitting a message from within Exim. Since child_open_exim() is
2423 documented for local_scan(), the new function should be too.
2425 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2426 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2427 results in an empty string is now treated as unset.
2429 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2431 PH/31 Added sender_verify_fail logging option.
2433 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2434 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2435 tidied the source and removed it altogether.
2437 PH/33 When a queue run was abandoned because the load average was too high, a
2438 log line was always written; now it is written only if the queue_run log
2439 selector is set. In addition, the log line for abandonment now contains
2440 information about the queue run such as the pid. This is always present
2441 in "start" and "stop" lines but was omitted from the "abandon" line.
2443 PH/34 Omit spaces between a header name and the colon in the error message that
2444 is given when verify = headers_syntax fails (if there are lots of them,
2445 the message gets confusing).
2447 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2448 names, as there are now some PTR records that contain slashes. This check
2449 is only to protect against broken name servers that fall over on strange
2450 characters, so the fact that it applies to all lookups doesn't matter.
2452 PH/36 Now that the new test suite is complete, we can remove some of the
2453 special code in Exim that was needed for the old test suite. For example,
2454 sorting DNS records because real resolvers return them in an arbitrary
2455 order. The new test suite's fake resolver always returns records in the
2458 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2459 bounces) except when queue_only is set, to avoid logging races between
2460 the different processes.
2462 PH/38 Panic-die if .include specifies a non-absolute path.
2464 PH/39 A tweak to the "H" retry rule from its user.
2466 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2467 a label. They prevented compilation on older perls.
2469 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2470 a warning to be raised on newish perls.
2472 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2473 on queue. Changes to match documented behaviour of showing count of
2474 messages matching specified criteria.
2476 PH/40 Changed the default ident timeout from 30s to 5s.
2478 PH/41 Added support for the use of login_cap features, on those BSD systems
2479 that have them, for controlling the resources used by pipe deliveries.
2481 PH/42 The content-scanning code uses fopen() to create files in which to put
2482 message data. Previously it was not paying any attention to the mode of
2483 the files. Exim runs with umask(0) because the rest of the code creates
2484 files with open(), and sets the required mode explicitly. Thus, these
2485 files were ending up world-writeable. This was not a big issue, because,
2486 being within the spool directory, they were not world-accessible. I have
2487 created a function called modefopen, which takes an additional mode
2488 argument. It sets umask(777), creates the file, chmods it to the required
2489 mode, then resets the umask. All the relevant calls to fopen() in the
2490 content scanning code have been changed to use this function.
2492 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2493 to 24 hours. This avoids potential overflow problems when processing G
2494 and H retry rules. I suspect nobody ever tinkers with this value.
2496 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2498 PH/45 When the plaintext authenticator is running as a client, the server's
2499 challenges are checked to ensure they are valid base64 strings. By
2500 default, the authentication attempt is cancelled if an invalid string is
2501 received. Setting client_ignore_invalid_base64 true ignores these errors.
2502 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2503 they are received. Thus, the responses can be made to depend on the
2504 challenges. If an invalid string is ignored, an empty string is placed in
2507 PH/46 Messages that are created by the autoreply transport now contains a
2508 References: header, in accordance with RFCs 2822 and 3834.
2510 PH/47 Added authenticated_sender_force to the smtp transport.
2512 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2514 PH/49 Installed latest patch from the Sieve maintainer.
2516 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2517 was also set, the appendfile transport was unnecessarily scanning a
2518 directory of message files (e.g. for maildir delivery) to find the count
2519 of files (along with the size), even though it did not need this
2520 information. It now does the scan only if it needs to find either the
2521 size of the count of files.
2523 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2525 PH/52 Two bugs concerned with error handling when the smtp transport is
2528 (i) Exim was not creating retry information for temporary errors given
2529 for individual recipients after the DATA command when the smtp transport
2530 was used in LMTP mode. This meant that they could be retried too
2531 frequently, and not timed out correctly.
2533 (ii) Exim was setting the flag that allows error details to be returned
2534 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2535 recipients that were returned after the DATA command.
2537 PH/53 This is related to PH/52, but is more general: for any failing address,
2538 when detailed error information was permitted to be returned to the
2539 sender, but the error was temporary, then after the final timeout, only
2540 "retry timeout exceeded" was returned. Now it returns the full error as
2541 well as "retry timeout exceeded".
2543 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2544 do this, and (what is worse) MTAs that accept it.
2546 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2547 will now be deprecated.
2549 PH/56 New os.c-cygwin from the Cygwin maintainer.
2551 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2552 formats (previously only available in exim formats via -bpr, -bpru,
2553 and -bpra. Now also available in native and exiqgrep formats)
2555 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2556 with very large, slow to parse queues
2558 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2560 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2562 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2563 responses to authentication challenges, though it was showing the
2564 challenges; (ii) I've removed the CR characters from the debug output for
2567 PH/58 Allow for the insertion of a newline as well as a space when a string
2568 is turned into more than one encoded-word during RFC 2047 encoding. The
2569 Sieve code now uses this.
2571 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2572 data_4xx, lost_connection, tls_required.
2574 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2575 message was being sent as an SMTP response.
2577 PH/61 Add -l and -k options to exicyclog.
2579 PH/62 When verifying, if an address was redirected to one new address, so that
2580 verification continued, and the new address failed or deferred after
2581 having set something in $address_data, the value of $address_data was not
2582 passed back to the ACL. This was different to the case when no
2583 redirection occurred. The value is now passed back in both cases.
2585 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2586 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2587 use login_cap.h, so on its own it isn't the distinguishing feature. The
2588 new name refers directly to the setclassresources() function.
2590 PH/65 Added configuration files for NetBSD3.
2592 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2594 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2595 is preferred over IPv4.
2597 PH/68 The bounce_return_message and bounce_return_body options were not being
2598 honoured for bounces generated during the reception of non-SMTP messages.
2599 In particular, this applied to messages rejected by the ACL. This bug has
2600 been fixed. However, if bounce_return_message is true and bounce_return_
2601 body is false, the headers that are returned for a non-SMTP message
2602 include only those that have been read before the error was detected.
2603 (In the case of an ACL rejection, they have all been read.)
2605 PH/69 The HTML version of the specification is now built in a directory called
2606 spec_html instead of spec.html, because the latter looks like a path with
2607 a MIME-type, and this confuses some software.
2609 PH/70 Catch two compiler warnings in sieve.c.
2611 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2612 function verify_get_ident() calls ip_connect() to connect a socket, but
2613 if the "connect()" function timed out, ip_connect() used to close the
2614 socket. However, verify_get_ident() also closes the socket later, and in
2615 between Exim writes to the log, which may get opened at this point. When
2616 the socket was closed in ip_connect(), the log could get the same file
2617 descriptor number as the socket. This naturally causes chaos. The fix is
2618 not to close the socket in ip_connect(); the socket should be closed by
2619 the function that creates it. There was only one place in the code where
2620 this was missing, in the iplookup router, which I don't think anybody now
2621 uses, but I've fixed it anyway.
2623 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2624 well as to direct DNS lookups. Otherwise the handling of names in host
2625 lists is inconsistent and therefore confusing.
2631 PH/01 Two changes to the default runtime configuration:
2633 (1) Move the checks for relay_from_hosts and authenticated clients from
2634 after to before the (commented out) DNS black list checks.
2636 (2) Add control=submission to the relay_from_hosts and authenticated
2637 clients checks, on the grounds that messages accepted by these
2638 statements are most likely to be submissions.
2640 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2642 (1) Generate an error if the third argument for the ${prvs expansion is
2645 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2648 (3) Reset the variables that are obtained from the first argument of
2649 ${prvscheck and used in the second argument before leaving the code,
2650 because their memory is reclaimed, so using them afterwards may do
2653 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2654 one (it's much easier than Tom thought :-).
2656 (5) Because of (4), we can now allow for the use of $prvscheck_result
2657 inside the third argument.
2659 PH/03 For some reason, the default setting of PATH when running a command from
2660 a pipe transport was just "/usr/bin". I have changed it to
2663 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2664 anything to be listed in the output from -bV.
2666 PH/05 When a filter generated an autoreply, the entire To: header line was
2667 quoted in the delivery log line, like this:
2669 => >A.N.Other <ano@some.domain> <original@ddress> ...
2671 This has been changed so that it extracts the operative address. There
2672 may be more than one such address. If so, they are comma-separated, like
2675 => >ano@some.domain,ona@other.domain <original@ddress> ...
2677 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2678 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2679 not being looked up in the rDNS to get a host name, Exim was showing the
2680 IP address twice in Received: lines, even though the IP addresses were
2681 identical. For example:
2683 Received: from [1.2.3.4] (helo=[1.2.3.4])
2685 However, if the real host name was known, it was omitting the HELO data
2686 if it matched the actual IP address. This has been tidied up so that it
2687 doesn't show the same IP address twice.
2689 PH/07 When both +timestamp and +memory debugging was on, the value given by
2690 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2691 called by the debug printing, thereby overwriting the timestamp buffer.
2692 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2694 PH/08 When the original message was included in an autoreply transport, it
2695 always said "this is a copy of the message, including all the headers",
2696 even if body_only or headers_only was set. It now gives an appropriate
2699 PH/09 Applied a patch from the Sieve maintainer which:
2701 o fixes some comments
2702 o adds the (disabled) notify extension core
2703 o adds some debug output for the result of if/elsif tests
2704 o points to the current vacation draft in the documentation
2705 and documents the missing references header update
2709 o fixes a bug in processing the envelope test (when testing
2710 multiple envelope elements, the last element determinted the
2713 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2714 Electronic Mail") by including:
2716 Auto-submitted: auto-generated
2718 in the messages that it generates (bounce messages and others, such as
2719 warnings). In the case of bounce messages for non-SMTP mesages, there was
2720 also a typo: it was using "Auto_submitted" (underscore instead of
2721 hyphen). Since every message generated by Exim is necessarily in response
2722 to another message, thes have all been changed to:
2724 Auto-Submitted: auto-replied
2726 in accordance with these statements in the RFC:
2728 The auto-replied keyword:
2730 - SHOULD be used on messages sent in direct response to another
2731 message by an automatic process,
2733 - MUST NOT be used on manually-generated messages,
2735 - MAY be used on Delivery Status Notifications (DSNs) and Message
2736 Disposition Notifications (MDNs),
2738 - MUST NOT be used on messages generated by automatic or periodic
2739 processes, except for messages which are automatic responses to
2742 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2743 to the default Received: header definition.
2745 PH/12 Added log selector acl_warn_skipped (default on).
2747 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2748 variables because (a) they are in the wrong storage pool and (b) even if
2749 they were copied, it wouldn't work properly because of the caching.
2751 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2752 checking when decoding. Apparently there are clients that generate
2753 overlong encoded strings. Why am I not surprised?
2755 PH/15 If the first argument of "${if match_address" was not empty, but did not
2756 contain an "@" character, Exim crashed. Now it writes a panic log message
2757 and treats the condition as false.
2759 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2761 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2762 extension "envelope-auth". The code is finished and in agreement with
2763 other implementations, but there is no documentation so far and in fact,
2764 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2765 not changing the active code.
2767 Print executed "if" and "elsif" statements when debugging is used. This
2768 helps a great deal to understand what a filter does.
2770 Document more things not specified clearly in RFC3028. I had all this
2771 sorted out, when out of a sudden new issues came to my mind. Oops."
2773 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2776 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2777 canonical form (as documented). However, after a host name lookup from
2778 the IP address, check_host() was doing a simple string comparison with
2779 addresses acquired from the DNS when checking that the found name did
2780 have the original IP as one of its addresses. Since any found IPv6
2781 addresses are likely to be in abbreviated form, the comparison could
2782 fail. Luckily, there already exists a function for doing the comparison
2783 by converting both addresses to binary, so now that is used instead of
2784 the text comparison.
2786 PH/20 There was another similar case to PH/19, when a complete host name was
2787 given in a host list; looking up its IP address could give an abbreviated
2788 form, whereas the current host's name might or might not be abbreviated.
2789 The same fix has been applied.
2795 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2796 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2799 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2800 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2802 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2804 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2805 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2806 Also, for "Auto-Submitted", treat anything other than "no" as
2807 non-personal, in accordance with RFC 3834. (Previously it treated
2808 anything starting "auto-" as non-personal.)
2810 TF/01 The control=submission/name=... option had a problem with syntax
2811 errors if the name included a slash character. The /name= option
2812 now slurps the rest of the string, so it can include any characters
2813 but it must come last in the list of options (after /sender_retain
2816 PH/05 Some modifications to the interface to the fake nameserver for the new
2824 TK/01 Added the "success_on_redirect" address verification option. See
2825 NewStuff for rationale and an example.
2827 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2829 PH/02 Patch to exigrep to allow it to work on syslog lines.
2831 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2832 fread() to skip over the body file's header line, because in Cygwin the
2833 header line is locked and is inaccessible.
2835 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2836 co-exist for some time) to make it clear that it is the Exim ID that is
2837 referenced, not the Message-ID: header line.
2839 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2840 string_format() function, because snprintf() does not exist on all
2843 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2844 ${stat: expansion item.
2846 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2847 protocol synchronization error", to keep the pedants happy.
2849 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2850 well as for IRIX systems, when gcc is being used. See the host.c source
2853 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2855 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2858 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2859 transport and to the smtp transport in LMTP mode.
2861 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2863 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2864 run a filter in a subprocess. This could lead to confusion in subsequent
2865 lookups in the parent process. There should also be a search_tidyup() at
2866 the end of the subprocess.
2868 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2869 only if the host matched helo_try_verify_hosts, which caused the
2870 verification to occur when the EHLO/HELO command was issued. The ACL just
2871 tested the remembered result. Now, if a previous verification attempt has
2872 not happened, "verify = helo" does it there and then.
2874 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2876 TK/03 Fix log output including CR from clamd.
2878 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2879 exist provoked a memory error which could cause a segfault.
2881 PH/15 Installed PCRE 6.2
2883 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2885 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2886 of the problem. Specifically, suggested +O2 rather than +O1 for the
2889 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2891 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2892 record, the port was not correctly logged when the outgoing_port log
2893 selector was set (it logged the transort's default port).
2895 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2896 fallback_hosts, and "hosts" in the smtp transport.
2898 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2899 host errors such as "Connection refused".
2901 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2902 authentication with radiusclient 0.4.9:
2904 - Error returned from rc_read_config was caught wrongly
2905 - Username/password not passed on to radius server due to wrong length.
2907 The presumption is that some radiusclient API changes for 4.51/PH/17
2908 were not taken care of correctly. The code is still untested by me (my
2909 Linux distribution still has 0.3.2 of radiusclient), but it was
2910 contributed by a Radius user.
2912 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2913 expanding the "port" option of the smtp transport.
2915 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2916 while reading a MIME header. Thanks to Tom Hughes for a patch.
2918 PH/24 Include config.h inside local_scan.h so that configuration settings are
2921 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2922 that in an ACL for RCPT (for example), you can examine exactly what was
2925 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2926 commands, but it was not correctly comparing the address with the actual
2927 client host address. Thus, it would show the EHLO address in Received:
2928 header lines when this was not necessary.
2930 PH/27 Added the % operator to ${eval:}.
2932 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2933 it should be ignoring failures (because with -C, for example, it has lost
2934 privilege). It wasn't ignoring creation failures other than "already
2937 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2940 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2941 because an input line was too long, either on its own, or by virtue of
2942 too many continuations, the temporary file was not being removed, and the
2943 return code was incorrect.
2945 PH/31 Missing "BOOL" in function definition in filtertest.c.
2947 PH/32 Applied Sieve patches from the maintainer.
2949 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2951 PH/33 Added "verify = not_blind".
2953 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2954 Local/Makefile (with some defaults set). These are used in built scripts
2955 such as exicyclog, but they have never been used in the exim_install
2956 script (though there are many overriding facilities there). I have
2957 arranged that the exim_install script now takes note of these two
2960 PH/35 Installed configuration files for Dragonfly.
2962 PH/36 When a locally submitted message by a trusted user did not contain a
2963 From: header, and the sender address was obtained from -f or from an SMTP
2964 MAIL command, and the trusted user did not use -F to supply a sender
2965 name, $originator_name was incorrectly used when constructing a From:
2966 header. Furthermore, $originator_name was used for submission mode
2967 messages from external hosts without From: headers in a similar way,
2968 which is clearly wrong.
2970 PH/37 Added control=suppress_local_fixups.
2972 PH/38 When log_selector = +received_sender was set, and the addition of the
2973 sender made the log line's construction buffer exactly full, or one byte
2974 less than full, an overflow happened when the terminating "\n" was
2977 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2978 when the result of a list match is failure because a DNS lookup failed.
2980 PH/40 RM_COMMAND is now used in the building process.
2982 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2983 the "build-* directories that it finds.
2985 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2986 address in a domain literal was a prefix of an interface address.
2988 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2989 when verifying a sender address, unless rewrite_headers is false.
2991 PH/44 Wrote a long comment about why errors_to addresses are verified as
2992 recipients, not senders.
2994 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2995 the ratelimit ACL was added.
2997 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2999 PH/46 Added extra information about PostgreSQL errors to the error string.
3001 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3002 avoiding the need to install special zones in a real server. This is
3003 backwards compatible; if it can't find the fake resolver, it drops back.
3004 Thus, both old and new test suites can be run.
3006 TF/02 Added util/ratelimit.pl
3008 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3009 clock is set back in time.
3011 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3012 Candler <B.Candler@pobox.com>.
3014 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3015 OK for addresses that are the result of redirections.
3017 PH/48 A number of further additions for the benefit of the new test suite,
3018 including a fake gethostbyname() that interfaces to the fake DNS resolver
3021 TF/06 The fix for widen_domains has also been applied to qualify_single and
3022 search_parents which are the other dnslookup options that can cause
3025 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3028 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3030 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3031 service" if it encountered an MX record with an empty target hostname.
3032 The message is now "an MX or SRV record indicated no SMTP service".
3034 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3035 if the DNS of the sending site is misconfigured. This is quite a
3036 common situation. This change restores the behaviour of treating a
3037 helo verification defer as a failure.
3039 PH/51 If self=fail was set on a router, the bounce message did not include the
3040 actual error message.
3046 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3048 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3049 command itself ended in error, the underlying message about the transport
3050 filter timeout was being overwritten with the pipe command error. Now the
3051 underlying error message should be appended to the second error message.
3053 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3055 PH/02 Reduce the amount of output that "make" produces by default. Full output
3056 can still be requested.
3058 PH/03 The warning log line about a condition test deferring for a "warn" verb
3059 was being output only once per connection, rather than after each
3060 occurrence (because it was using the same function as for successful
3061 "warn" verbs). This seems wrong, so I have changed it.
3063 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3064 it should not have, which might have caused a crash in the right
3065 circumstances, but probably never did.
3067 PH/04 Installed a modified version of Tony Finch's patch to make submission
3068 mode fix the return path as well as the Sender: header line, and to
3069 add a /name= option so that you can make the user's friendly name appear
3072 TF/03 Added the control = fakedefer ACL modifier.
3074 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3075 Mark Lowes for thorough testing.
3077 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3079 TK/03 Merged latest SRS patch from Miles Wilton.
3081 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3082 with the definition in sysexits.h (which is #included earlier).
3083 Fortunately, Exim does not actually use EX_OK. The code used to try to
3084 preserve the sysexits.h value, by assumimg that macro definitions were
3085 scanned for macro replacements. I have been disabused of this notion,
3086 so now the code just undefines EX_OK before #including unistd.h.
3088 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3089 in the smtp transport. When a block could not be written in a single
3090 write() function, the timeout was being re-applied to each part-write.
3091 This seems wrong - if the receiver was accepting one byte at a time it
3092 would take for ever. The timeout is now adjusted when this happens. It
3093 doesn't have to be particularly precise.
3095 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3096 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3098 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3099 without a domain if the check to <postmaster@domain> fails.
3101 SC/01 Eximstats: added -xls and the ability to specify output files
3102 (patch written by Frank Heydlauf).
3104 SC/02 Eximstats: use FileHandles for outputing results.
3106 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3108 SC/04 Eximstats: fixed display of large numbers with -nvr option
3110 SC/05 Eximstats: fixed merging of reports with empty tables.
3112 SC/06 Eximstats: added the -include_original_destination flag
3114 SC/07 Eximstats: removed tabs and trailing whitespace.
3116 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3118 TK/06 MBOX spool code: Add real "From " MBOX separator line
3119 so the .eml file is really in mbox format (even though
3120 most programs do not really care). Patch from Alex Miller.
3122 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3123 The latter is generated from $received_to and is only set if the
3124 message has one envelope recipient. SA can use these headers,
3125 obviously out-of-the-box. Patch from Alex Miller.
3127 PH/08 The ${def test on a variable was returning false if the variable's
3128 value was "0", contrary to what the specification has always said!
3129 The result should be true unless the variable is empty.
3131 PH/09 The syntax error of a character other than { following "${if
3132 def:variable_name" (after optional whitespace) was not being diagnosed.
3133 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3134 accidental colon was present, for example, could give incorrect results.
3136 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3137 result is used (not including appendfile, where other changes are about
3140 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3141 This involved changing a lot of size variables from int to off_t. It
3142 should work with maildirs and everything.
3144 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3145 spamd dying while we are connected to it.
3147 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3150 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3151 function for BDB 4.3.
3153 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3155 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3156 the sources for PCRE have been re-arranged and more files are now
3159 PH/15 The code I had for printing potentially long long variables in PH/11
3160 above was not the best (it lost precision). The length of off_t variables
3161 is now inspected at build time, and an appropriate printing format (%ld
3162 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3163 to be "long long int" or "long int". This is needed for the internal
3164 formatting function string_vformat().
3166 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3167 the configuration file to be ":syslog", then the script "guesses" where
3168 the logs files are, rather than using the compiled in default. In our
3169 case the guess is not the same as the compiled default, so the script
3170 suddenly stopped working when I started to use syslog. The patch checks
3171 to see if log_file_path is "". If so, it attempts to read it from exim
3172 with no configuration file to get the compiled in version, before it
3173 falls back to the previous guessing code."
3175 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3176 implementing BATV in an Exim configuration. See NewStuff for the gory
3179 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3180 Makefile that are specific to HP-UX.
3182 PH/18 If the "use_postmaster" option was set for a recipient callout together
3183 with the "random" option, the postmaster address was used as the MAIL
3184 FROM address for the random test, but not for the subsequent recipient
3185 test. It is now used for both.
3187 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3188 patch removes a few documentation additions to RFC 3028, because the
3189 latest draft now contains them. It adds the new en;ascii-case comparator
3190 and a new error check for 8bit text in MIME parts. Comparator and
3191 require names are now matched exactly. I enabled the subaddress
3192 extension, but it is not well tested yet (read: it works for me)."
3194 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3195 rework some of the code of TK/09 above to avoid the hardwired use of
3196 "%lld" and "long long". Replaced the call to snprintf() with a call to
3199 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3200 records point to non-existent hosts", "retry timeout exceeded", and
3201 "retry time not reached for any host after a long failure period".
3203 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3204 experimental DomainKeys support:
3206 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3207 (2) On an error such as an illegally used "control", the wrong name for
3208 the control was given.
3210 These problems did NOT occur unless DomainKeys support was compiled.
3212 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3214 PH/24 Added ${if match_ip condition.
3216 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3217 databases so that it will be absolutely obvious if a crash occurs in the
3218 DB library. This is a regular occurrence (often caused by mis-matched
3221 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3222 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3223 picked up on a user's system that detects such things. There doesn't seem
3224 to be a gcc warning option for this - only an attribute that has to be
3225 put on the function's prototype. It seems that in Fedora Core 4 they have
3226 set this on a number of new functions. No doubt there will be more in due
3229 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3230 specify a transport. However, if an address that was verified by such a
3231 router was the subject of a callout, Exim crashed because it tried to
3232 read the rcpt_include_affixes from the non-existent transport. Now it
3233 just assumes that the setting of that option is false. This bug was
3234 introduced by 4.51/PH/31.
3236 PH/28 Changed -d+all to exclude +memory, because that information is very
3237 rarely of interest, but it makes the output a lot bigger. People tend to
3238 do -d+all out of habit.
3240 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3241 code in os-type was giving problems when libc.so lives in lib64, like on
3244 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3245 aren't the modern standard, and it seems that some systems' include files
3246 don't always have them. Exim was already checking for some of the newer
3247 ones like T_AAAA, and defining it itself. I've added checks for all the
3248 record types that Exim uses.
3250 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3251 not automatically generating a new one, as it is supposed to. This
3252 prevented TLS from working. If the file did exist, but contained invalid
3253 data, a new version was generated, as expected. It was only the case of a
3254 non-existent file that was broken.
3256 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3257 with a change in libdomainkeys > 0.64.
3259 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3260 from DNS. If the selector record carries the flag, it now has
3261 precedence over the domain-wide flag.
3263 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3265 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3266 the use of a port name that isn't defined in /etc/services) occurred, the
3267 message was deferred as in a normal delivery, and thus remained on the
3268 spool, instead of being failed because of the mua_wrapper setting. This
3269 is now fixed, and I tidied up some of the mua_wrapper messages at the
3272 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3273 the messages in a hash of arrays rather than using individual hashes.
3274 This is a bit cleaner and results in dramatic memory savings, albeit
3275 at a slight CPU cost.
3277 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3278 as requested by Marc Sherman.
3280 SC/10 Eximstats: added histograms for user specified patterns as requested
3283 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3285 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3286 fopen() in the content-scanning modules that did not already have it.
3292 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3293 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3295 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3297 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3299 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3300 to test Sieve filters that use "vacation".
3302 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3303 that changes the way the GnuTLS parameters are stored in the cache file.
3304 The new format can be generated externally. For backward compatibility,
3305 if the data in the cache doesn't make sense, Exim assumes it has read an
3306 old-format file, and it generates new data and writes a new file. This
3307 means that you can't go back to an older release without removing the
3310 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3311 work if there are any delivery delays because "one_time" forces the
3312 parent to be marked "delivered", so its unseen clone is never tried
3313 again. For this reason, Exim now forbids the simultaneous setting of
3316 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3317 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3318 entry if you want to know the details. The fix, however, neglected to
3319 consider the case when local delivery batching is involved. The test for
3320 "previously delivered" was not happening when checking to see if an
3321 address could be batched with a previous (undelivered) one; under
3322 certain circumstances this could lead to multiple deliveries to the same
3325 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3326 in its include files, and this causes problems building Exim.
3328 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3329 header_syntax) but Exim was just ignoring anything given after a slash.
3330 In particular, this caused confusion with an attempt to use "verify =
3331 reverse_host_lookup/defer_ok". An error is now given when options are
3332 supplied for verify items that do not have them. (Maybe reverse_host_
3333 lookup should have a defer_ok option, but that's a different point.)
3335 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3336 defined by RFC 821) to 2048, because there were problems with some AUTH
3337 commands, and RFC 1869 says the size should be increased for extended
3338 SMTP commands that take arguments.
3340 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3343 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3344 "unknown" error; now it says that the functionality isn't in the binary.
3346 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3347 an address' error message when a string expansion fails (syntax or
3348 whatever). Otherwise the password may appear in the log. Following change
3349 PH/42 below, there is no longer a chance of it appearing in a bounce
3352 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3354 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3355 of its addresses was ever tried. (Bugzilla bug #2.)
3357 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3358 the result incorrectly in the debug output. (It correctly added a newline
3359 to what was transported.)
3361 TF/01 Added $received_time.
3363 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3364 commented out examples of how to interface to a virus scanner and to
3365 SpamAssassin. Also added commented examples of av_scanner and
3366 spamd_address settings.
3368 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3369 and controls are allowed in which ACLs. There were a couple of minor
3370 errors. Some of the entries in the conditions table (which is a table of
3371 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3372 negation of where the condition IS allowed.
3374 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3376 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3377 header file does not have a version number, so I've had to invent a new
3378 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3379 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3380 radiusclient), but it was contributed by a Radius user.
3382 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3383 files or directories, for OpenSSL.
3385 PH/19 When an Exim process that is running as root has to create an Exim log
3386 file, it does so in a subprocess that runs as exim:exim so as to get the
3387 ownership right at creation (otherwise, other Exim processes might see
3388 the file with the wrong ownership). There was no test for failure of this
3389 fork() call, which would lead to the process getting stuck as it waited
3390 for a non-existent subprocess. Forks do occasionally fail when resources
3391 run out. I reviewed all the other calls to fork(); they all seem to check
3394 PH/20 When checking for unexpected SMTP input at connect time (before writing
3395 the banner), Exim was not dealing correctly with a non-positive return
3396 from the read() function. If the client had disconnected by this time,
3397 the result was a log entry for a synchronization error with an empty
3398 string after "input=" when read() returned zero. If read() returned -1
3399 (an event I could not check), uninitialized data bytes were printed.
3400 There were reports of junk text (parts of files, etc) appearing after
3403 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3405 PH/22 Added support for macro redefinition, and (re)definition in between
3406 driver and ACL definitions.
3408 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3409 forgetting to use the resulting value; it was using the unexpanded value.
3411 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3412 hadn't been configured. The fix is from Juergen Kreileder, who
3413 understands it better than I do:
3415 "Here's what I see happening with three configured cyrus_sasl
3416 authenticators configured (plain, login, cram-md5):
3418 On startup auth_cyrus_sasl_init() gets called for each of these.
3419 This means three calls to sasl_listmech() without a specified mech_list.
3420 => SASL tests which mechs of all available mechs actually work
3421 => three warnings about OTP not working
3422 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3424 With the patch, sasl_listmech() also gets called three times. But now
3425 SASL's mech_list option is set to the server_mech specified in the the
3426 authenticator. Or in other words, the answer from sasl_listmech()
3427 gets limited to just the mech you're testing for (which is different
3429 => the return list contains just 'plain' or 'login', 'cram-md5' or
3430 nothing depending on the value of ob->server_mech.
3432 I've just tested the patch: Authentication still works fine,
3433 unavailable mechs specified in the exim configuration are still
3434 caught, and the auth.log warnings about OTP are gone."
3436 PH/25 When debugging is enabled, the contents of the command line are added
3437 to the debugging output, even when log_selector=+arguments is not
3440 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3441 answer is "GNU", and only if the return is "GNU/something" is the answer
3444 PH/27 $acl_verify_message is now set immediately after the failure of a
3445 verification in an ACL, and so is available in subsequent modifiers. In
3446 particular, the message can be preserved by coding like this:
3448 warn !verify = sender
3449 set acl_m0 = $acl_verify_message
3451 Previously, $acl_verify_message was set only while expanding "message"
3452 and "log_message" when a very denied access.
3454 PH/28 Modified OS/os.c-Linux with
3456 -#ifndef OS_LOAD_AVERAGE
3457 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3459 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3460 nomenclature these days.)
3462 PH/29 Installed patch from the Sieve maintainer that adds the options
3463 sieve_useraddress and sieve_subaddress to the redirect router.
3465 PH/30 In these circumstances:
3466 . Two addresses routed to the same list of hosts;
3467 . First host does not offer TLS;
3468 . First host accepts first address;
3469 . First host gives temporary error to second address;
3470 . Second host offers TLS and a TLS session is established;
3471 . Second host accepts second address.
3472 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3473 and peerdn, if requested) that were in fact used only for the second
3476 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3477 attention to any local part prefix or suffix that was matched by the
3478 router that accepted the address. It now behaves in the same way as it
3479 does for delivery: the affixes are removed from the local part unless
3480 rcpt_include_affixes is set on the transport.
3482 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3483 timeout during the DATA phase of an incoming message.
3485 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3486 applied a patch sanctioned by the Sieve maintainer.
3488 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3489 the uid or gid is negative. A case of a negative gid caused this to be
3490 noticed. The fix allows for either to be negative.
3492 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3493 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3496 PH/36 The change PH/12 above was broken. Fixed it.
3498 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3499 the grounds that routing the same address twice would always produce the
3500 same answer. This might have been true once, but it is certainly no
3501 longer true now. Routing a child address may depend on the previous
3502 routing that produced that child. Some complicated redirection strategies
3503 went wrong when messages had multiple recipients, and made Exim's
3504 behaviour dependent on the order in which the addresses were given.
3506 I have moved the duplicate checking until after the routing is complete.
3507 Exim scans the addresses that are assigned to local and remote
3508 transports, and removes any duplicates. This means that more work will be
3509 done, as duplicates will always all be routed, but duplicates are
3510 presumably rare, so I don't expect this is of any significance.
3512 For deliveries to pipes, files, and autoreplies, the duplicate checking
3513 still happens during the routing process, since they are not going to be
3516 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3517 It corrects a timeout issue with spamd. This is Ian's comment: "The
3518 background is that sometimes spamd either never reads data from a
3519 connection it has accepted, or it never writes response data. The exiscan
3520 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3521 blindly assumes that writes won't block so it may never time out."
3523 PH/39 Allow G after quota size as well as K and M.
3525 PH/40 The value set for $authenticated_id in an authenticator may not contain
3526 binary zeroes or newlines because the value is written to log lines and
3527 to spool files. There was no check on this. Now the value is run through
3528 the string_printing() function so that such characters are converted to
3529 printable escape sequences.
3531 PH/41 $message_linecount is a new variable that contains the total number of
3532 lines in the message. Compare $body_linecount, which is the count for the
3535 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3536 bounce and delay warning messages, except in certain special cases, which
3539 (a) An SMTP error message from a remote host;
3540 (b) A message specified in a :fail: redirection;
3541 (c) A message specified in a "fail" command in a system filter;
3542 (d) A message specified in a FAIL return from the queryprogram router;
3543 (e) A message specified by the cannot_route_message router option.
3545 In these cases only, Exim does include the error details in bounce and
3546 warning messages. There are also a few cases where bland messages such
3547 as "unrouteable address" or "local delivery error" are given.
3549 PH/43 $value is now also set for the "else" part of a ${run expansion.
3551 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3552 being worked on, but at least Exim now implements the latest version to
3555 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3556 process to complete was treated as a delivery failure, a timeout while
3557 writing the message to the pipe was logged, but erroneously treated as a
3558 successful delivery. Such timeouts include transport filter timeouts. For
3559 consistency with the overall process timeout, these timeouts are now
3560 treated as errors, giving rise to delivery failures by default. However,
3561 there is now a new Boolean option for the pipe transport called
3562 timeout_defer, which, if set TRUE, converts the failures into defers for
3563 both kinds of timeout. A transport filter timeout is now identified in
3566 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3567 systems where "make" and "gmake" are different, calling "gmake" at top
3568 level broke things. I've arranged for the value of $(MAKE) to be passed
3569 from the Makefile to this script so that it can call the same version of
3573 A note about Exim versions 4.44 and 4.50
3574 ----------------------------------------
3576 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3577 changes of various kinds. As a consequence, a big documentation update was
3578 needed. This delayed the release for rather longer than seemed good, especially
3579 in the light of a couple of (minor) security issues. Therefore, the changes
3580 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3581 release. So 4.44 and 4.50 are in effect two different branches that both start
3584 I have left the 4.50 change log unchanged; it contains all the changes since
3585 4.43. The change log for 4.44 is below; many of its items are identical to
3586 those for 4.50. This seems to be the most sensible way to preserve the
3587 historical information.
3593 1. Minor wording change to the doc/README.SIEVE file.
3595 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3596 computation of the current number of files was incorrect.
3598 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3599 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3602 4. Give more explanation in the error message when the command for a transport
3603 filter fails to execute.
3605 5. There are several places where Exim runs a non-Exim command in a
3606 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3607 was being done only for the command run by the queryprogram router. It is
3608 now done for all such subprocesses. The other cases are: ${run, transport
3609 filters, and the commands run by the lmtp and pipe transports.
3611 6. Added CONFIGURE_GROUP build-time option.
3613 7. Some older OS have a limit of 256 on the maximum number of file
3614 descriptors. Exim was using setrlimit() to set 1000 as a large value
3615 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3616 systems. I've change it so that if it can't get 1000, it tries for 256.
3618 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3619 was an oversight, and furthermore, ever since the addition of extra
3620 controls (e.g. 4.43/32), the checks on when to allow different forms of
3621 "control" were broken. There should now be diagnostics for all cases when a
3622 control that does not make sense is encountered.
3624 9. Added the /retain_sender option to "control=submission".
3626 10. $recipients is now available in the predata ACL (oversight).
3628 11. Tidy the search cache before the fork to do a delivery from a message
3629 received from the command line. Otherwise the child will trigger a lookup
3630 failure and thereby defer the delivery if it tries to use (for example) a
3631 cached ldap connection that the parent has called unbind on.
3633 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3634 of $address_data from the recipient verification was clobbered by the
3635 sender verification.
3637 13. The value of address_data from a sender verification is now available in
3638 $sender_address_data in subsequent conditions in the ACL statement.
3640 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3642 15. Added a new option "connect=<time>" to callout options, to set a different
3645 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3646 was its contents. (It was OK if the option was not defined at all.)
3648 17. A "Completed" log line is now written for messages that are removed from
3649 the spool by the -Mrm option.
3651 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3652 information about exactly what failed.
3654 19. Added -dd to debug only the daemon process.
3656 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3657 handles timeouts, both on the server side and network timeouts. Renamed the
3658 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3660 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3661 It is now set to "smtps".
3663 22. $host_address is now set to the target address during the checking of
3664 ignore_target_hosts.
3666 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3667 being passed; this would have caused $sender_host_name to have been used if
3668 matching the list had actually called for a host name (not very likely,
3669 since this list is usually IP addresses). A host name is now passed as
3672 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3673 code that specifies a non-restarting handler (typically sigaction() in
3674 modern systems) in an attempt to fix a rare and obscure crash bug.
3676 25. Narrowed the window for a race in the daemon that could cause it to ignore
3677 SIGCHLD signals. This is not a major problem, because they are used only to
3678 wake it up if nothing else does.
3680 26. A malformed maildirsize file could cause Exim to calculate negative values
3681 for the mailbox size or file count. Odd effects could occur as a result.
3682 The maildirsize information is now recalculated if the size or filecount
3685 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3686 support for a long time. Removed HAVE_SYS_VFS_H.
3688 28. Installed the latest version of exipick from John Jetmore.
3690 29. In an address list, if the pattern was not a regular expression, an empty
3691 subject address (from a bounce message) matched only if the pattern was an
3692 empty string. Non-empty patterns were not even tested. This was the wrong
3693 because it is perfectly reasonable to use an empty address as part of a
3694 database query. An empty address is now tested by patterns that are
3695 lookups. However, all the other forms of pattern expect the subject to
3696 contain a local part and a domain, and therefore, for them, an empty
3697 address still always fails if the pattern is not itself empty.
3699 30. Exim went into a mad DNS loop when attempting to do a callout where the
3700 host was specified on an smtp transport, and looking it up yielded more
3701 than one IP address.
3703 31. Re-factored the code for checking spool and log partition space into a
3704 function that finds that data and another that does the check. The former
3705 is then used to implement four new variables: $spool_space, $log_space,
3706 $spool_inodes, and $log_inodes.
3708 32. The RFC2047 encoding function was originally intended for short strings
3709 such as real names; it was not keeping to the 75-character limit for
3710 encoded words that the RFC imposes. It now respects the limit, and
3711 generates multiple encoded words if necessary. To be on the safe side, I
3712 have increased the buffer size for the ${rfc2047: expansion operator from
3715 33. It is now permitted to omit both strings after an "if" condition; if the
3716 condition is true, the result is "true". As before, when the second string
3717 is omitted, a false condition yields an empty string. This makes it less
3718 cumbersome to write custom ACL and router conditions.
3720 34. Failure to deliver a bounce message always caused it to be frozen, even if
3721 there was an errors_to setting on the router. The errors_to setting is now
3724 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3725 canonical form (fully expanded) before being placed in
3726 $sender_host_address.
3728 36. The table in the code that translates DNS record types into text (T_A to
3729 "A" for instance) was missing entries for NS and CNAME. It is just possible
3730 that this could have caused confusion if both these types were looked up
3731 for the same domain, because the text type is used as part of Exim's
3732 per-process caching. But the chance of anyone hitting this buglet seems
3735 37. The dnsdb lookup has been extended in a number of ways.
3737 (1) There is a new type, "zns", which walks up the domain tree until it
3738 finds some nameserver records. It should be used with care.
3740 (2) There is a new type, "mxh", which is like "mx" except that it returns
3741 just the host names, not the priorities.
3743 (3) It is now possible to give a list of domains (or IP addresses) to be
3744 looked up. The behaviour when one of the lookups defers can be
3745 controlled by a keyword.
3747 (4) It is now possible to specify the separator character for use when
3748 multiple records are returned.
3750 38. The dnslists ACL condition has been extended: it is now possible to supply
3751 a list of IP addresses and/or domains to be looked up in a particular DNS
3754 39. Added log_selector=+queue_time_overall.
3756 40. When running the queue in the test harness, wait just a tad after forking a
3757 delivery process, to get repeatability of debugging output.
3759 41. Include certificate and key file names in error message when GnuTLS fails
3760 to set them up, because the GnuTLS error message doesn't include the name
3761 of the failing file when there is a problem reading it.
3763 42. Allow both -bf and -bF in the same test run.
3765 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3767 44. The "Exiscan patch" is now merged into the mainline Exim source.
3769 45. Sometimes the final signoff response after QUIT could fail to get
3770 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3771 < 0 before doing a fflush(). This bug looks as though it goes back to the
3772 introduction of TLS in release 3.20, but "sometimes" must have been rare
3773 because the tests only now provoked it.
3775 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3776 (this can affect the format of dates).
3778 47. exim_tidydb, when checking for the continued existence of a message for
3779 which it has found a message-specific retry record, was not finding
3780 messages that were in split spool directories. Consequently, it was
3781 deleting retry records that should have stayed in existence.
3783 48. Steve fixed some bugs in eximstats.
3785 49. The SPA authentication driver was not abandoning authentication and moving
3786 on to the next authenticator when an expansion was forced to fail,
3787 contradicting the general specification for all authenticators. Instead it
3788 was generating a temporary error. It now behaves as specified.
3790 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3791 (the order specifies the preference for clients). The order is now AES256,
3792 AES128, 3DES, ARCFOUR128.
3794 51. Small patch to Sieve code - explicitly set From: when generating an
3797 52. Exim crashed if a remote delivery caused a very long error message to be
3798 recorded - for instance if somebody sent an entire SpamAssassin report back
3799 as a large number of 550 error lines. This bug was coincidentally fixed by
3800 increasing the size of one of Exim's internal buffers (big_buffer) that
3801 happened as part of the Exiscan merge. However, to be on the safe side, I
3802 have made the code more robust (and fixed the comments that describe what
3805 53. Now that there can be additional text after "Completed" in log lines (if
3806 the queue_time_overall log selector is set), a one-byte patch to exigrep
3807 was needed to allow it to recognize "Completed" as not the last thing in
3810 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3811 patch that reportedly fixes this has been added. I am not expert enough to
3812 create a test for it. This is what the patch creator wrote:
3814 "I found a little strange behaviour of ldap code when working with
3815 Windows 2003 AD Domain, where users was placed in more than one
3816 Organization Units. When I tried to give exim partial DN, the exit code
3817 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3818 But simultaneously result of request was absolutely normal ldap result,
3819 so I produce this patch..."
3821 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3822 REFERENCE, so I have modified the code to exclude the patch when that macro
3825 55. Some experimental protocols are using DNS PTR records for new purposes. The
3826 keys for these records are domain names, not reversed IP addresses. The
3827 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3828 leaves it alone. Component reversal etc. now happens only for IP addresses.
3831 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3833 57. Double the size of the debug message buffer (to 2048) so that more of very
3834 long debug lines gets shown.
3836 58. The exicyclog utility now does better if the number of log files to keep
3837 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3839 59. Two changes related to the smtp_active_hostname option:
3841 (1) $smtp_active_hostname is now available as a variable.
3842 (2) The default for smtp_banner uses $smtp_active_hostname instead
3843 of $primary_hostname.
3845 60. The host_aton() function is supposed to be passed a string that is known
3846 to be a valid IP address. However, in the case of IPv6 addresses, it was
3847 not checking this. This is a hostage to fortune. Exim now panics and dies
3848 if the condition is not met. A case was found where this could be provoked
3849 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3850 components; fortuitously, this particular loophole had already been fixed
3851 by change 4.50/55 above.
3853 If there are any other similar loopholes, the new check in host_aton()
3854 itself should stop them being exploited. The report I received stated that
3855 data on the command line could provoke the exploit when Exim was running as
3856 exim, but did not say which command line option was involved. All I could
3857 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3858 running as the user.
3861 61. There was a buffer overflow vulnerability in the SPA authentication code
3862 (which came originally from the Samba project). I have added a test to the
3863 spa_base64_to_bits() function which I hope fixes it.
3866 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3867 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3869 63. The daemon start-up calls getloadavg() while still root for those OS that
3870 need the first call to be done as root, but it missed one case: when
3871 deliver_queue_load_max is set with deliver_drop_privilege. This is
3872 necessary for the benefit of the queue runner, because there is no re-exec
3873 when deliver_drop_privilege is set.
3875 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3876 This has been fixed.
3878 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3879 was in use, was not putting the data itself into the right store pool;
3880 consequently, it could be overwritten for a subsequent message in the same
3881 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3884 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3886 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3887 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3888 the function were treating the return as a boolean value, which happened to
3889 work because 0=false and not-0=true, but is not correct code.
3891 68. The host_aton() function was not handling scoped IPv6 addresses (those
3892 with, for example, "%eth0" on the end) correctly.
3894 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3895 negated items (that is, ~something) in unsigned ints. Some compilers
3896 apparently mutter when there is no cast.
3898 70. If an address verification called from an ACL failed, and did not produce a
3899 user-specific message (i.e. there was only a "system" message), nothing was
3900 put in $acl_verify_message. In this situation, it now puts the system
3903 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3904 session; change 4.31/43 added the unwanted input to the log line - except
3905 that it did not do this in the start of session case. It now does.
3907 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3908 This is wrong and can cause the other end to generate a synchronization
3909 error if it is another Exim or anything else that does the synchronization
3910 check. A QUIT command is no longer sent after a timeout.
3912 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3913 during host lookups.
3915 74. The defer_ok option of callout verification was not working if it was used
3916 when verifying addresses in header lines, that is, for this case:
3918 verify = header_sender/callout=defer_ok
3920 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3921 those file descriptors could be used for SMTP connections. If anything
3922 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3923 could be sent to the SMTP client, causing chaos. The daemon now opens
3924 stdin, stdout, and stderr to /dev/null when it puts itself into the
3927 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3928 log by default. The user can override this with suitable Perl magic.
3930 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3931 the log message when discard triggers, was not working for the DATA ACL or
3932 for the non-SMTP ACL.
3934 78. Error message wording change in sieve.c.
3936 79. If smtp_accept_max_per_host was set, the number of connections could be
3937 restricted to fewer than expected, because the daemon was trying to set up
3938 a new connection before checking whether the processes handling previous
3939 connections had finished. The check for completed processes is now done
3940 earlier. On busy systems, this bug wouldn't be noticed because something
3941 else would have woken the daemon, and it would have reaped the completed
3944 80. If a message was submitted locally by a user whose login name contained one
3945 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3946 It caused a spool format error. I have fixed the spool reading code. A
3947 related problem was that the "from" clause in the Received: line became
3948 illegal because of the space(s). It is now covered by ${quote_local_part.
3950 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3953 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3955 83. Patch from Sieve maintainer to fix unterminated string problem in
3956 "vacation" handling.
3958 84. Some minor changes to the Linux configuration files to help with other
3959 OS variants using glibc.
3961 85. One more patch for Sieve to update vacation handling to latest spec.
3964 ----------------------------------------------------
3965 See the note above about the 4.44 and 4.50 releases.
3966 ----------------------------------------------------
3972 1. Change 4.43/35 introduced a bug that caused file counts to be
3973 incorrectly computed when quota_filecount was set in an appendfile
3976 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3977 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3980 3. Give more explanation in the error message when the command for a transport
3981 filter fails to execute.
3983 4. There are several places where Exim runs a non-Exim command in a
3984 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3985 was being done only for the command run by the queryprogram router. It is
3986 now done for all such subprocesses. The other cases are: ${run, transport
3987 filters, and the commands run by the lmtp and pipe transports.
3989 5. Some older OS have a limit of 256 on the maximum number of file
3990 descriptors. Exim was using setrlimit() to set 1000 as a large value
3991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3992 systems. I've change it so that if it can't get 1000, it tries for 256.
3994 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3995 was an oversight, and furthermore, ever since the addition of extra
3996 controls (e.g. 4.43/32), the checks on when to allow different forms of
3997 "control" were broken. There should now be diagnostics for all cases when a
3998 control that does not make sense is encountered.
4000 7. $recipients is now available in the predata ACL (oversight).
4002 8. Tidy the search cache before the fork to do a delivery from a message
4003 received from the command line. Otherwise the child will trigger a lookup
4004 failure and thereby defer the delivery if it tries to use (for example) a
4005 cached ldap connection that the parent has called unbind on.
4007 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4008 of $address_data from the recipient verification was clobbered by the
4009 sender verification.
4011 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4012 was its contents. (It was OK if the option was not defined at all.)
4014 11. A "Completed" log line is now written for messages that are removed from
4015 the spool by the -Mrm option.
4017 12. $host_address is now set to the target address during the checking of
4018 ignore_target_hosts.
4020 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4021 being passed; this would have caused $sender_host_name to have been used if
4022 matching the list had actually called for a host name (not very likely,
4023 since this list is usually IP addresses). A host name is now passed as
4026 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4027 code that specifies a non-restarting handler (typically sigaction() in
4028 modern systems) in an attempt to fix a rare and obscure crash bug.
4030 15. Narrowed the window for a race in the daemon that could cause it to ignore
4031 SIGCHLD signals. This is not a major problem, because they are used only to
4032 wake it up if nothing else does.
4034 16. A malformed maildirsize file could cause Exim to calculate negative values
4035 for the mailbox size or file count. Odd effects could occur as a result.
4036 The maildirsize information is now recalculated if the size or filecount
4039 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4040 support for a long time. Removed HAVE_SYS_VFS_H.
4042 18. Updated exipick to current release from John Jetmore.
4044 19. Allow an empty sender to be matched against a lookup in an address list.
4045 Previously the only cases considered were a regular expression, or an
4048 20. Exim went into a mad DNS lookup loop when doing a callout where the
4049 host was specified on the transport, if the DNS lookup yielded more than
4052 21. The RFC2047 encoding function was originally intended for short strings
4053 such as real names; it was not keeping to the 75-character limit for
4054 encoded words that the RFC imposes. It now respects the limit, and
4055 generates multiple encoded words if necessary. To be on the safe side, I
4056 have increased the buffer size for the ${rfc2047: expansion operator from
4059 22. Failure to deliver a bounce message always caused it to be frozen, even if
4060 there was an errors_to setting on the router. The errors_to setting is now
4063 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4064 canonical form (fully expanded) before being placed in
4065 $sender_host_address.
4067 24. Updated eximstats to version 1.33
4069 25. Include certificate and key file names in error message when GnuTLS fails
4070 to set them up, because the GnuTLS error message doesn't include the name
4071 of the failing file when there is a problem reading it.
4073 26. Expand error message when OpenSSL has problems setting up cert/key files.
4076 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4077 (this can affect the format of dates).
4079 28. exim_tidydb, when checking for the continued existence of a message for
4080 which it has found a message-specific retry record, was not finding
4081 messages that were in split spool directories. Consequently, it was
4082 deleting retry records that should have stayed in existence.
4084 29. eximstats updated to version 1.35
4085 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4086 1.35 - bugfix such that pie charts by volume are generated correctly
4088 30. The SPA authentication driver was not abandoning authentication and moving
4089 on to the next authenticator when an expansion was forced to fail,
4090 contradicting the general specification for all authenticators. Instead it
4091 was generating a temporary error. It now behaves as specified.
4093 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4094 (the order specifies the preference for clients). The order is now AES256,
4095 AES128, 3DES, ARCFOUR128.
4097 31. Small patch to Sieve code - explicitly set From: when generating an
4100 32. Exim crashed if a remote delivery caused a very long error message to be
4101 recorded - for instance if somebody sent an entire SpamAssassin report back
4102 as a large number of 550 error lines. This bug was coincidentally fixed by
4103 increasing the size of one of Exim's internal buffers (big_buffer) that
4104 happened as part of the Exiscan merge. However, to be on the safe side, I
4105 have made the code more robust (and fixed the comments that describe what
4108 33. Some experimental protocols are using DNS PTR records for new purposes. The
4109 keys for these records are domain names, not reversed IP addresses. The
4110 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4111 leaves it alone. Component reversal etc. now happens only for IP addresses.
4114 34. The host_aton() function is supposed to be passed a string that is known
4115 to be a valid IP address. However, in the case of IPv6 addresses, it was
4116 not checking this. This is a hostage to fortune. Exim now panics and dies
4117 if the condition is not met. A case was found where this could be provoked
4118 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4119 components; fortuitously, this particular loophole had already been fixed
4120 by change 4.50/55 or 4.44/33 above.
4122 If there are any other similar loopholes, the new check in host_aton()
4123 itself should stop them being exploited. The report I received stated that
4124 data on the command line could provoke the exploit when Exim was running as
4125 exim, but did not say which command line option was involved. All I could
4126 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4127 running as the user.
4130 35. There was a buffer overflow vulnerability in the SPA authentication code
4131 (which came originally from the Samba project). I have added a test to the
4132 spa_base64_to_bits() function which I hope fixes it.
4135 36. The daemon start-up calls getloadavg() while still root for those OS that
4136 need the first call to be done as root, but it missed one case: when
4137 deliver_queue_load_max is set with deliver_drop_privilege. This is
4138 necessary for the benefit of the queue runner, because there is no re-exec
4139 when deliver_drop_privilege is set.
4141 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4142 was in use, was not putting the data itself into the right store pool;
4143 consequently, it could be overwritten for a subsequent message in the same
4144 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4147 38. Sometimes the final signoff response after QUIT could fail to get
4148 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4149 < 0 before doing a fflush(). This bug looks as though it goes back to the
4150 introduction of TLS in release 3.20, but "sometimes" must have been rare
4151 because the tests only now provoked it.
4157 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4158 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4159 Now it returns an integer. A number of calls were still expecting a T/F
4160 return. Fortuitously, in all cases, the tests worked in OK situations,
4161 which is the norm. However, things would have gone wrong on any write
4162 failures on the smtp file descriptor. This function is used when sending
4163 messages over SMTP and also when doing verify callouts.
4165 2. When Exim is called to do synchronous delivery of a locally submitted
4166 message (the -odf or -odi options), it no longer closes stderr before doing
4169 3. Implemented the mua_wrapper option.
4171 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4173 5. Implemented the functions header_remove(), header_testname(),
4174 header_add_at_position(), and receive_remove_recipient(), and exported them
4177 6. If an ACL "warn" statement specified the addition of headers, Exim already
4178 inserted X-ACL-Warn: at the start if there was no header name. However, it
4179 was not making this test for the second and subsequent header lines if
4180 there were newlines in the string. This meant that an invalid header could
4181 be inserted if Exim was badly configured.
4183 7. Allow an ACL "warn" statement to add header lines at the start or after all
4184 the Received: headers, as well as at the end.
4186 8. Added the rcpt_4xx retry error code.
4188 9. Added postmaster_mailfrom=xxx to callout verification option.
4190 10. Added mailfrom=xxxx to the callout verification option, for verify=
4193 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4194 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4196 12. Inserted some casts to stop certain compilers warning when using pointer
4197 differences as field lengths or precisions in printf-type calls (mostly
4198 affecting debugging statements).
4200 13. Added optional readline() support for -be (dynamically loaded).
4202 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4203 same clock tick as a message's arrival, so that its received time was the
4204 same as the "first fail" time on the retry record, and that message
4205 remained on the queue past the ultimate address timeout, every queue runner
4206 would try a delivery (because it was past the ultimate address timeout) but
4207 after another failure, the ultimate address timeout, which should have then
4208 bounced the address, did not kick in. This was a "< instead of <=" error;
4209 in most cases the first failure would have been in the next clock tick
4210 after the received time, and all would be well.
4212 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4213 being recognized when the domain list was tested by the match_domain
4214 condition in an expansion string.
4216 16. Added the ${str2b64: operator.
4218 17. Exim was always calling setrlimit() to set a large limit for the number of
4219 processes, without checking whether the existing limit was already
4220 adequate. (It did check for the limit on file descriptors.) Furthermore,
4221 errors from getrlimit() and setrlimit() were being ignored. Now they are
4222 logged to the main and panic logs, but Exim does carry on, to try to do its
4223 job under whatever limits there are.
4225 18. Imported PCRE 5.0.
4227 19. Trivial typo in log message " temporarily refused connection" (the leading
4230 20. If the log selector return_path_on_delivery was set and an address was
4231 redirected to /dev/null, the delivery process crashed because it assumed
4232 that a return path would always be set for a "successful" delivery. In this
4233 case, the whole delivery is bypassed as an optimization, and therefore no
4236 21. Internal re-arrangement: the function for sending a challenge and reading
4237 a response while authentication was assuming a zero-terminated challenge
4238 string. It's now changed to take a pointer and a length, to allow for
4239 binary data in such strings.
4241 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4243 23. Exim was not respecting finduser_retries when seeking the login of the
4244 uid under which it was called; it was always trying 10 times. (The default
4245 setting of finduser_retries is zero.) Also, it was sleeping after the final
4246 failure, which is pointless.
4248 24. Implemented tls_on_connect_ports.
4250 25. Implemented acl_smtp_predata.
4252 26. If the domain in control=submission is set empty, Exim assumes that the
4253 authenticated id is a complete email address when it generates From: or
4254 Sender: header lines.
4256 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4257 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4258 chown and chgrp in /bin and hostname in /usr/bin.
4260 28. Exim was keeping the "process log" file open after each use, just as it
4261 does for the main log. This opens the possibility of it remaining open for
4262 long periods when the USR1 signal hits a daemon. Occasional processlog
4263 errors were reported, that could have been caused by this. Anyway, it seems
4264 much more sensible not to leave this file open at all, so that is what now
4267 29. The long-running daemon process does not normally write to the log once it
4268 has entered its main loop, and it closes the log before doing so. This is
4269 so that log files can straightforwardly be renamed and moved. However,
4270 there are a couple of unusual error situations where the daemon does write
4271 log entries, and I had neglected to close the log afterwards.
4273 30. The text of an SMTP error response that was received during a remote
4274 delivery was being truncated at 512 bytes. This is too short for some of
4275 the long messages that one sometimes sees. I've increased the limit to
4278 31. It is now possible to make retry rules that apply only when a message has a
4279 specific sender, in particular, an empty sender.
4281 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4282 it possible to be selective about when SMTP synchronization is enforced.
4284 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4286 32. Implemented hosts_connection_nolog.
4288 33. Added an ACL for QUIT.
4290 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4293 35. Added mailbox_size and mailbox_filecount to appendfile.
4295 36. Added control = no_multiline_responses to ACLs.
4297 37. There was a bug in the logic of the code that waits for the clock to tick
4298 in the case where the clock went backwards by a substantial amount such
4299 that the microsecond fraction of "now" was more than the microsecond
4300 fraction of "then" (but the whole seconds number was less).
4302 38. Added support for the libradius Radius client library this is found on
4303 FreeBSD (previously only the radiusclient library was supported).
4309 1. When certain lookups returned multiple values in the form name=value, the
4310 quoting of the values was not always being done properly. Specifically:
4311 (a) If the value started with a double quote, but contained no whitespace,
4313 (b) If the value contained whitespace other than a space character (i.e.
4314 tabs or newlines or carriage returns) it was not quoted.
4315 This fix has been applied to the mysql and pgsql lookups by writing a
4316 separate quoting function and calling it from the lookup code. The fix
4317 should probably also be applied to nisplus, ibase and oracle lookups, but
4318 since I cannot test any of those, I have not disturbed their existing code.
4320 2. A hit in the callout cache for a specific address caused a log line with no
4321 reason for rejecting RCPT. Now it says "Previous (cached) callout
4322 verification failure".
4324 3. There was an off-by-one bug in the queryprogram router. An over-long
4325 return line was truncated at 256 instead of 255 characters, thereby
4326 overflowing its buffer with the terminating zero. As well as fixing this, I
4327 have increased the buffer size to 1024 (and made a note to document this).
4329 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4330 when Exim is waiting for an SMTP response from a remote server, Exim
4331 restarts its select() call on the socket, thereby resetting its timeout.
4332 This is not a problem when such interrupts are rare. Somebody set up a cron
4333 job to run exiwhat every 2 minutes, which is less than the normal select()
4334 timeout (5 or 10 minutes). This meant that the select() timeout never
4335 kicked in because it was always reset. I have fixed this by comparing the
4336 time when an interrupt arrives with the time at the start of the first call
4337 to select(). If more time than the timeout has elapsed, the interrupt is
4338 treated as a timeout.
4340 5. Some internal re-factoring in preparation for the addition of Sieve
4341 extensions (by MH). In particular, the "personal" test is moved to a
4342 separate function, and given an option for scanning Cc: and Bcc: (which is
4343 not set for Exim filters).
4345 6. When Exim created an email address using the login of the caller as the
4346 local part (e.g. when creating a From: or Sender: header line), it was not
4347 quoting the local part when it contained special characters such as @.
4349 7. Installed new OpenBSD configuration files.
4351 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4352 try to make them clearer.
4354 9. Callout options, other than the timeout value, were being ignored when
4355 verifying sender addresses in header lines. For example, when using
4357 verify = header_sender/callout=no_cache
4359 the cache was (incorrectly) being used.
4361 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4362 only the Cygwin environment.
4364 11. When return_path_on_delivery was set as a log selector, if different remote
4365 addresses in the same message used different return paths and parallel
4366 remote delivery occurred, the wrong values would sometimes be logged.
4367 (Whenever a remote delivery process finished, the return path value from
4368 the most recently started remote delivery process was logged.)
4370 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4371 lines when AUTH and/or TLS are in use. This is the "received protocol"
4372 field. Exim used to use "asmtp" for authenticated SMTP, without any
4373 indication (in the protocol name) for TLS use. Now it follows the RFC and
4374 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4375 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4376 names appear in log lines as well as in Received: header lines.
4378 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4379 extensions, and comparison tests, and to fix some bugs.
4381 14. Changes to the "personal" filter test:
4383 (1) The test was buggy in that it was just doing the equivalent of
4384 "contains" tests on header lines. For example, if a user's address was
4385 anne@some.where, the "personal" test would incorrectly be true for
4387 To: susanne@some.where
4389 This test is now done by extracting each address from the header in turn,
4390 and checking the entire address. Other tests that are part of "personal"
4391 are now done using regular expressions (for example, to check local parts
4392 of addresses in From: header lines).
4394 (2) The list of non-personal local parts in From: addresses has been
4395 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4396 taken from the Sieve specification recommendations.
4398 (3) If the message contains any header line starting with "List-" it is
4399 treated as non-personal.
4401 (4) The test for "circular" in the Subject: header line has been removed
4402 because it now seems ill-conceived.
4404 15. Minor typos in src/EDITME comments corrected.
4406 16. Installed latest exipick from John Jetmore.
4408 17. If headers_add on a router specified a text string that was too long for
4409 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4410 of string_sprintf() is now avoided.
4412 18. $message_body_size was not set (it was always zero) when running the DATA
4413 ACL and the local_scan() function.
4415 19. For the "mail" command in an Exim filter, no default was being set for
4416 the once_repeat time, causing a random time value to be used if "once" was
4417 specified. (If the value happened to be <= 0, no repeat happened.) The
4418 default is now 0s, meaning "never repeat". The "vacation" command was OK
4419 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4420 (I found it when inspecting the code).
4422 20. There is now an overall timeout for performing a callout verification. It
4423 defaults to 4 times the callout timeout, which applies to individual SMTP
4424 commands during the callout. The overall timeout applies when there is more
4425 than one host that can be tried. The timeout is checked before trying the
4426 next host. This prevents very long delays if there are a large number of
4427 hosts and all are timing out (e.g. when the network connections are timing
4428 out). The value of the overall timeout can be changed by specifying an
4429 additional sub-option for "callout", called "maxwait". For example:
4431 verify = sender/callout=5s,maxwait=20s
4433 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4434 to the end before writing, but this should make it even safer).
4436 22. Exim was forgetting that it had advertised PIPELINING for the second and
4437 subsequent messages on an SMTP connection. It was also not resetting its
4438 memory on STARTTLS and an internal HELO.
4440 23. When Exim logs an SMTP synchronization error within a session, it now
4441 records whether PIPELINING has been advertised or not.
4443 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4444 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4445 rather than long int.
4447 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4449 26. Added the never_mail option to autoreply.
4455 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4456 crash if the getsockname() call failed; this can happen if a connection is
4457 closed very soon after it is established. The problem was simply in the
4458 order in which certain operations were done, causing Exim to try to write
4459 to the SMTP stream before it had set up the file descriptor. The bug has
4460 been fixed by making things happen in the correct order.
4466 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4467 before the connection was closed, thus losing the rejection response.
4469 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4470 some early Solaris releases, but causes trouble in current releases where
4471 socklen_t is defined.
4473 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4476 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4479 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4480 that releasing the top of it at the end releases what was used for sub-
4481 expansions (unless the block got too big). However, discard this block if
4482 the first thing is a variable or header, so that we can use its block when
4483 it is dynamic (useful for very large $message_headers, for example).
4485 6. Lookups now cache *every* query, not just the most recent. A new, separate
4486 store pool is used for this. It can be recovered when all lookup caches are
4487 flushed. Lookups now release memory at the end of their result strings.
4488 This has involved some general refactoring of the lookup sources.
4490 7. Some code has been added to the store_xxx() functions to reduce the amount
4491 of flapping under certain conditions.
4493 8. log_incoming_interface used to affect only the <= reception log lines. Now
4494 it causes the local interface and port to be added to several more SMTP log
4495 lines, for example "SMTP connection from", and rejection lines.
4497 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4499 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4501 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4502 could be overwritten at the end of the current message (or the start of a
4503 new message if it was set in a HELO ACL). The value is now preserved for
4504 the duration of the SMTP connection.
4506 12. If a transport had a headers_rewrite setting, and a matching header line
4507 contained an unqualified address, that address was qualified, even if it
4508 did not match any rewriting rules. The underlying bug was that the values
4509 of the flags that permit the existence of unqualified sender and recipient
4510 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4511 non-local messages, and by -bnq for local messages) were not being
4512 preserved with the message after it was received.
4514 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4515 "next input=" as part of the text comprising the host identity instead of
4516 the correct text. The code was using the same buffer for two different
4517 strings. However, depending on which order the printing function evaluated
4518 its arguments, the bug did not always show up. Under Linux, for example, my
4519 test suite worked just fine.
4521 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4522 doesn't work with some older versions of Perl. It has been changed to "my",
4523 which in any case is probably the better facility to use.
4525 15. A really picky compiler found some instances of statements for creating
4526 error messages that either had too many or two few arguments for the format
4529 16. The size of the buffer for calls to the DNS resolver has been increased
4530 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4531 for addresses that have a lot of PTR records. This alleviates a problem; it
4532 does not fully solve it.
4534 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4535 buffer now truncates the list and logs the incident, which is the same
4536 action as happens when Exim is looking up a host name and its aliases.
4537 Previously in this situation something unpredictable would happen;
4538 sometimes it was "internal error: store_reset failed".
4540 18. If a server dropped the connection unexpectedly when an Exim client was
4541 using GnuTLS and trying to read a response, the client delivery process
4542 crashed while trying to generate an error log message.
4544 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4545 string, for example:
4547 warn message = H1: something\nH2: something
4549 the text was added as a single header line from Exim's point of view
4550 though it ended up OK in the delivered message. However, searching for the
4551 second and subsequent header lines using $h_h2: did not work. This has been
4552 fixed. Similarly, if a system filter added multiple headers in this way,
4553 the routers could not see them.
4555 20. Expanded the error message when iplsearch is called with an invalid key to
4556 suggest using net-iplsearch in a host list.
4558 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4559 ACLs are no longer actually imposed (and a message to that effect is
4562 22. If a "gecos" field in a passwd entry contained escaped characters, in
4563 particular, if it contained a \" sequence, Exim got it wrong when building
4564 a From: or a Sender: header from that name. A second bug also caused
4565 incorrect handling when an unquoted " was present following a character
4566 that needed quoting.
4568 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4569 was not being matched caselessly.
4571 24. Arranged for all hyphens in the exim.8 source to be escaped with
4574 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4575 properly. Recipient callout cache records were still being keyed to include
4576 the sender, even when use_sender was set false. This led to far more
4577 callouts that were necessary. The sender is no longer included in the key
4578 when use_sender is false.
4580 26. Added "control = submission" modifier to ACLs.
4582 27. Added the ${base62d: operator to decode base 62 numbers.
4584 28. dnsdb lookups can now access SRV records.
4586 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4587 the configuration file.
4589 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4590 (-v) mode. This makes the output for a verbose queue run more intelligible.
4592 31. Added a use_postmaster feature to recipient callouts.
4594 32. Added the $body_zerocount variable, containing the number of binary zero
4595 bytes in the message body.
4597 33. The time of last modification of the "new" subdirectory is now used as the
4598 "mailbox time last read" when there is a quota error for a maildir
4601 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4603 35. Added +ignore_unknown as a special item in host lists.
4605 36. Code for decoding IPv6 addresses in host lists is now included, even if
4606 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4607 address was recognized as an IP address, but was then not correctly decoded
4608 into binary, causing unexpected and incorrect effects when compared with
4615 1. Very minor rewording of debugging text in manualroute to say "list of
4616 hosts" instead of "hostlist".
4618 2. If verify=header_syntax was set, and a header line with an unqualified
4619 address (no domain) and a large number of spaces between the end of the
4620 name and the colon was received, the reception process suffered a buffer
4621 overflow, and (when I tested it) crashed. This was caused by some obsolete
4622 code that should have been removed. The fix is to remove it!
4624 3. When running in the test harness, delay a bit after writing a bounce
4625 message to get a bit more predictability in the log output.
4627 4. Added a call to search_tidyup() just before forking a reception process. In
4628 theory, someone could use a lookup in the expansion of smtp_accept_max_
4629 per_host which, without the tidyup, could leave open a database connection.
4631 5. Added the variables $recipient_data and $sender_data which get set from a
4632 lookup success in an ACL "recipients" or "senders" condition, or a router
4633 "senders" option, similar to $domain_data and $local_part_data.
4635 6. Moved the writing of debug_print from before to after the "senders" test
4638 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4639 problems for message scanning, either using a data ACL, or using
4640 local_scan() because the Received: header was not generated till after they
4641 were called (in order to set the time as the time of reception completion).
4642 I have revised the way this works. The header is now generated after the
4643 body is received, but before the ACL or local_scan() are called. After they
4644 are run, the timestamp in the header is updated.
4650 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4651 before starting a queue runner without re-exec. This happened only when
4652 deliver_drop_privilege was set or when the Exim user was set to root. The
4653 effect of the bug was that timeouts during subsequent deliveries caused
4654 crashes instead of being properly handled. The handler is now left at its
4655 default (and expected) setting.
4657 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4658 message, again when deliver_drop_privilege is set or Exim is run as root.
4659 The bug described in (1) was not present in this case, but the tidying up
4660 of the other signals was missing. I have made the two cases consistent.
4662 3. The ignore_target_hosts setting on a manualroute router was being ignored
4663 for hosts that were looked up using the /MX notation.
4665 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4668 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4669 operated on the sender address. After changing the $sender_address to <>
4670 for the sender address verify, Exim was re-instated it as the original
4671 (before rewriting) address, but remembering that it had rewritten it, so it
4672 wasn't rewriting it again. This bug also had the effect of breaking the
4673 sender address verification caching when the sender address was rewritten.
4675 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4676 This has been changed so that if the ip literal address matches
4677 ignore_target_hosts, the router declines.
4679 7. Added expansion conditions match_domain, match_address, and match_local_
4680 part (NOT match_host).
4682 8. The placeholder for the Received: header didn't have a length field set.
4684 9. Added code to Exim itself and to exim_lock to test for a specific race
4685 condition that could lead to file corruption when using MBX delivery. The
4686 issue is with the lockfile that is created in /tmp. If this file is removed
4687 after a process has opened it but before that process has acquired a lock,
4688 there is the potential for a second process to recreate the file and also
4689 acquire a lock. This could lead to two Exim processes writing to the file
4690 at the same time. The added code performs the same test as UW imapd; it
4691 checks after acquiring the lock that its file descriptor still refers to
4692 the same named file.
4694 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4695 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4698 11. Added the smtp_active_hostname option. If used, this will typically be made
4699 to depend on the incoming interface address. Because $interface_address is
4700 not set up until the daemon has forked a reception process, error responses
4701 that can happen earlier (such as "too many connections") no longer contain
4704 12. If an expansion in a condition on a "warn" statement fails because a lookup
4705 defers, the "warn" statement is abandoned, and the next ACL statement is
4706 processed. Previously this caused the whole ACL to be aborted.
4708 13. Added the iplsearch lookup type.
4710 14. Added ident_timeout as a log selector.
4712 15. Added tls_certificate_verified as a log selector.
4714 16. Added a global option tls_require_ciphers (compare the smtp transport
4715 option of the same name). This controls incoming TLS connections.
4717 17. I finally figured out how to make tls_require_ciphers do a similar thing
4718 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4719 before starting the TLS session.
4721 18. Tabs are now shown as \t in -bP output.
4723 19. If the log selector return_path_on_delivery was set, Exim crashed when
4724 bouncing a message because it had too many Received: header lines.
4726 20. If two routers both had headers_remove settings, and the first one included
4727 a superfluous trailing colon, the final name in the first list and the
4728 first name in the second list were incorrectly joined into one item (with a
4729 colon in the middle).
4735 1. Added -C and -D options to the exinext utility, mainly to make it easier
4736 to include in the automated testing, but these could be helpful when
4737 multiple configurations are in use.
4739 2. The exinext utility was not formatting the output nicely when there was
4740 an alternate port involved in the retry record key, nor when there was a
4741 message id as well (for retries that were specific to a specific message
4742 and a specific host). It was also confused by IPv6 addresses, because of
4743 the additional colons they contain. I have fixed the IPv4 problem, and
4744 patched it up to do a reasonable job for IPv6.
4746 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4747 delivery, the log line now contains "pipelined" if PIPELINING was used.
4749 4. An SMTP transport process used to panic and die if the bind() call to set
4750 an explicit outgoing interface failed. This has been changed; it is now
4751 treated in the same way as a connect() failure.
4753 5. A reference to $sender_host_name in the part of a conditional expansion
4754 that was being skipped was still causing a DNS lookup. This no longer
4757 6. The def: expansion condition was not recognizing references to header lines
4758 that used bh_ and bheader_.
4760 7. Added the _cache feature to named lists.
4762 8. The code for checking quota_filecount in the appendfile transport was
4763 allowing one more file than it should have been.
4765 9. For compatibility with Sendmail, the command line option
4773 and sets the incoming protocol and host name (for trusted callers). The
4774 host name and its colon can be omitted when only the protocol is to be set.
4775 Note the Exim already has two private options, -pd and -ps, that refer to
4776 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4777 "s", but I don't think that's a major issue.
4779 10. A number of refactoring changes to the code, none of which should affect
4782 (a) The number of logging options was getting close to filling up the
4783 32-bit word that was used as a bit map. I have split them into two classes:
4784 those that are passed in the argument to log_write(), and those that are
4785 only ever tested independently outside of that function. These are now in
4786 separate 32-bit words, so there is plenty of room for expansion again.
4787 There is no change in the user interface or the logging behaviour.
4789 (b) When building, for example, log lines, the code previously used a
4790 macro that called string_cat() twice, in order to add two strings. This is
4791 not really sufficiently general. Furthermore, there was one instance where
4792 it was actually wrong because one of the argument was used twice, and in
4793 one call a function was used. (As it happened, calling the function twice
4794 did not affect the overall behaviour.) The macro has been replaced by a
4795 function that can join an arbitrary number of extra strings onto a growing
4798 (c) The code for expansion conditions now uses a table and a binary chop
4799 instead of a serial search (which was left over from when there were very
4800 few conditions). Also, it now recognizes conditions like "pam" even when
4801 the relevant support is not compiled in: a suitably worded error message is
4802 given if an attempt is made to use such a condition.
4804 11. Added ${time_interval:xxxxx}.
4806 12. A bug was causing one of the ddress fields not to be passed back correctly
4807 from remote delivery subprocesses. The field in question was not being
4808 subsequently used, so this caused to problems in practice.
4810 13. Added new log selectors queue_time and deliver_time.
4812 14. Might have fixed a bug in maildirsizefile handling that threw up
4813 "unexpected character" debug warnings, and recalculated the data
4814 unnecessarily. In any case, I expanded the warning message to give more
4817 15. Added the message "Restricted characters in address" to the statements in
4818 the default ACL that block characters like @ and % in local parts.
4820 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4821 Three changes have been made:
4823 (a) There was a serious bug; a negative response to MAIL caused the whole
4824 recipient domain to be cached as invalid, thereby blocking all messages
4825 to all local parts at the same domain, from all senders. This bug has
4826 been fixed. The domain is no longer cached after a negative response to
4827 MAIL if the sender used is not empty.
4829 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4832 (c) A new callout option, "use_sender" has been added for people who want
4833 the modified behaviour.
4839 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4842 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4843 indeed breaks things for older releases.
4845 3. Added additional logging to the case where there is a problem reading data
4846 from a filter that is running in a subprocess using a pipe, in order to
4847 try to track down a specific problem.
4849 4. Testing facility fudge: when running in the test harness and attempting
4850 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4851 getting "No route to host". Convert this to a timeout.
4853 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4856 6. Some OS don't have socklen_t but use size_t instead. This affects the
4857 fifth argument of getsockopt() amongst other things. This is now
4858 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4859 can be set for individual OS. I have set it for SunOS5, OSF1, and
4860 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4861 some earlier ones do not.
4863 7. Change 4.30/15 was not doing the test caselessly.
4865 8. The standard form for an IPv6 address literal was being rejected by address
4866 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4867 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4868 this, as well as the form without the "IPv6" on the front (but only when
4869 address literals are enabled, of course).
4871 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4873 10. Exim crashed if a message with an empty sender address specified by -f
4874 encountered a router with an errors_to setting. This could be provoked only
4875 by a command such as
4879 where an empty string was supplied; "<>" did not hit this bug.
4881 11. Installed PCRE release 4.5.
4883 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4884 remained set. It is now erased.
4886 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4887 times from message ids (which are base 36 rather than the normal 62).
4889 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4890 were being counted as actual protocol errors, and logged if the log
4891 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4892 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4893 and DATA following a set of rejected RCPTs do not count as protocol errors.
4894 In other words, Exim assumes they were pipelined, though this may not
4895 actually be the case. Of course, in all cases the client gets an
4896 appropriate error code.
4898 15. If a lookup fails in an ACL condition, a message about the failure may
4899 be available; it is used if testing the ACL cannot continue, because most
4900 such messages specify what the cause of the deferral is. However, some
4901 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4902 that caused an old message to be retained and used if a later statement
4903 caused a defer, replacing the real cause of the deferral.
4905 16. If an IP address had so many PTR records that the DNS lookup buffer
4906 was not large enough to hold them, Exim could crash while trying to process
4907 the truncated data. It now detects and logs this case.
4909 17. Further to 4.21/58, another change has been made: if (and only if) the
4910 first line of a message (the first header line) ends with CRLF, a bare LF
4911 in a subsequent header line has a space inserted after it, so as not to
4912 terminate the header.
4914 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4915 unnecessarily, used atoi() instead of strtol(), and didn't check the
4916 termination when getting file sizes from file names by regex.
4918 19. Completely re-implemented the support for maildirsize files, in the light
4919 of a number of problems with the previous contributed implementation
4920 (4.30/29). In particular:
4922 . If the quota is zero, the maildirsize file is maintained, but no quota is
4925 . If the maildir directory does not exist, it is created before any attempt
4926 to write a maildirsize file.
4928 . The quota value in the file is just a cache; if the quota is changed in
4929 the transport, the new value overrides.
4931 . A regular expression is available for excluding directories from the
4934 20. The autoreply transport checks the characters in options that define the
4935 message's headers; it allows continued headers, but it was checking with
4936 isspace() after an embedded newline instead of explicitly looking for a
4939 21. If all the "regular" hosts to which an address was routed had passed their
4940 expiry times, and had not reached their retry times, the address was
4941 bounced, even if fallback hosts were defined. Now Exim should go on to try
4944 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4945 equivalent code in the SMTP transport. Some hosts send humungous responses
4946 to HELO/EHLO, more than 1024 it seems.
4948 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4949 gives compiler warnings in some environments. I've now done it "properly",
4952 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4953 (because of problems with the built-in one) was declared to return uschar *
4954 instead of char *, causing compiler failure.
4956 25. Fixed a file descriptor leak when processing alias/forward files.
4958 26. Fixed a minor format string issue in dbfn.c.
4960 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4962 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4963 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4966 29. When a sender address is verified, it is cached, to save repeating the test
4967 when there is more than one recipient in a message. However, when the
4968 verification involves a callout, it is possible for different callout
4969 options to be set for different recipients. It is too complicated to keep
4970 track of this in the cache, so now Exim always runs a verification when a
4971 callout is required, relying on the callout cache for the optimization.
4972 The overhead is duplication of the address routing, but this should not be
4975 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4976 to be verified with callout=postmaster, and the main callout worked but the
4977 postmaster check failed, the verification correctly failed. However, if a
4978 subsequent RCPT command asked for sender verification *without* the
4979 postmaster check, incorrect caching caused this verification also to fail,
4982 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4983 it was not caching the DNS options (qualify_single, search_parents) that
4984 were used when the lookup failed. A subsequent lookup with different
4985 options therefore always gave the same answer, though there were cases
4986 where it should not have. (Example: a "domains = !$mx_any" option on a
4987 dnslookup router: the "domains" option is always processed without any
4988 widening, but the router might have qualify_single set.) Now Exim uses the
4989 cached value only when the same options are set.
4991 32. Added John Jetmore's "exipick" utility to the distribution.
4993 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4994 than a timeout (e.g. a certificate is required, and is not provided), an
4995 Exim server now closes the connection immediately. Previously it waited for
4996 the client to close - but if the client is SSL, it seems that they each
4997 wait for each other, leading to a delay before one of them times out.
4999 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5000 maintained 0.8.x compatibility because I don't think many are using it, and
5001 it is clearly obsolete.
5003 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5006 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5007 client certificate was expired. A simple patch fixes this, though I don't
5008 understand the full logic of why the verify callback is called multiple
5011 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5012 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5013 which causes problems with some clients (such as the Certicom SSL Plus
5014 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5015 disables the coutermeasure allowing Eudora to connect."
5017 38. Exim was not checking that a write() to a log file succeeded. This could
5018 lead to Bad Things if a log got too big, in particular if it hit a file
5019 size limit. Exim now panics and dies if it cannot write to a log file, just
5020 as it does if it cannot open a log file.
5022 39. Modified OS/Makefile-Linux so that it now contains
5024 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5026 The two -D definitions ensure that Exim is compiled with large file
5027 support, which makes it possible to handle log files that are bigger than
5030 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5031 instance) a domain was checked against a named list that involved a lookup,
5032 causing $domain_data to be set, then another domain was checked against the
5033 same list, then the first domain was re-checked, the value of $domain_data
5034 after the final check could be wrong. In particular, if the second check
5035 failed, it could be set empty. This bug probably also applied to
5038 41. The strip_trailing_dot option was not being applied to the address given
5039 with the -f command-line option.
5041 42. The code for reading a message's header from the spool was incrementing
5042 $received_count, but never initializing it. This meant that the value was
5043 incorrect (doubled) while delivering a message in the same process in which
5044 it was received. In the most common configuration of Exim, this never
5045 happens - a fresh exec is done - but it can happen when
5046 deliver_drop_privilege is set.
5048 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5049 it now includes up to 150 characters of the unexpected data in the log
5052 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5053 and building data strings. The size of both of these buffers was 10 000
5054 bytes - far larger than anybody would *ever* want, thought I. Needless to
5055 say, somebody hit the limit. I have increased the maximum line length to
5056 20 000 and the maximum data length of concatenated lines to 100 000. I have
5057 also fixed two bugs, because there was no checking on these buffers. Tsk,
5058 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5059 buffer is too small.
5061 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5062 lsearch lookups. Now it does.
5064 46. When parsing a route_list item in a manualroute router, a fixed-length
5065 buffer was used for the list of hosts. I made this 1024 bytes long,
5066 thinking that nobody would ever have a list of hosts that long. Wrong.
5067 Somebody had a whole pile of complicated expansion conditions, and the
5068 string was silently truncated, leading to an expansion error. It turns out
5069 that it is easier to change to an unlimited length (owing to other changes
5070 that have happened since this code was originally written) than to build
5071 structure for giving a limitation error. The length of the item that
5072 expands into the list of hosts is now unlimited.
5074 47. The lsearch lookup could not handle data where the length of text line was
5075 more than 4095 characters. Such lines were truncated, leading to shortened
5076 data being returned. It should now handle lines of any length.
5078 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5079 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5082 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5083 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5084 no longer happens in comment lines. A list of replacements is now placed
5085 at the head of all of the source files, except those whose only change is
5086 to replace PERL_COMMAND in the very first #! line.
5088 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5089 messages on the queue, with a bottom-up merge sort, using code contributed
5090 by Michael Haardt. This should make operations like -bp somewhat faster on
5091 large queues. It won't affect queue runners, except when queue_run_in_order
5094 51. Installed eximstats 1.31 in the distribution.
5096 52. Added support for SRV lookups to the dnslookup router.
5098 53. If an ACL referred to $message_body or $message_body_end, the value was not
5099 reset for any messages that followed in the same SMTP session.
5101 54. The store-handling optimization for building very long strings was not
5102 differentiating between the different store pools. I don't think this
5103 actually made any difference in practice, but I've tidied it.
5105 55. While running the routers to verify a sender address, $sender_address
5106 was still set to the sender address. This is wrong, because when routing to
5107 send a bounce to the sender, it would be empty. Therefore, I have changed
5108 it so that, while verifying a sender address, $sender_address is set to <>.
5109 (There is no change to what happens when verifying a recipient address.)
5111 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5112 target A or AAAA records (if not already returned) without resetting the
5113 qualify_single or search_parents options of the DNS resolver. These are
5114 inappropriate in this case because the targets of MX and SRV records must
5115 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5116 target that, when qualified, matched something in the local domain. These
5117 two options are now turned off when doing these lookups.
5119 57. It seems that at least some releases of Reiserfs (which does not have the
5120 concept of a fixed number of inodes) returns zero and not -1 for the
5121 number of available inodes. This interacted badly with check_spool_inodes,
5122 which assumed that -1 was the "no such thing" setting. What I have done is
5123 to check that the total number of inodes is greater than zero before doing
5124 the test of how many are available.
5126 58. When a "warn" ACL statement has a log_message modifier, the message is
5127 remembered, and not repeated. This is to avoid a lot of repetition when a
5128 message has many recipients that cause the same warning to be written.
5129 Howewer, Exim was preserving the list of already written lines for an
5130 entire SMTP session, which doesn't seem right. The memory is now reset if a
5131 new message is started.
5133 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5134 debugging output unless log_rewrite was also set.
5136 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5137 of (int)(handle) when we know that handle contains (void *)(-1).
5139 61. The Exim daemon panic-logs an error return when it closes the incoming
5140 connection. However "connection reset by peer" seems to be common, and
5141 isn't really an error worthy of noting specially, so that particular error
5144 62. When Exim is trying to find all the local interfaces, it used to panic and
5145 die if the ioctl to get the interface flags failed. However, it seems that
5146 on at least one OS (Solaris 9) it is possible to have an interface that is
5147 included in the list of interfaces, but for which you get a failure error
5148 for this call. This happens when the interface is not "plumbed" into a
5149 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5150 failure of the "get flags" call assumes that the interface is down.
5152 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5153 makes life easier for people who are doing arithmetic on fields extracted
5154 from dates, where you often get leading zeros that should not be
5155 interpreted as octal.
5157 64. Added qualify_domain to the redirect router, to override the global
5160 65. If a pathologically long header line contained very many addresses (the
5161 report of this problem mentioned 10 000) and each of them was rewritten,
5162 Exim could use up a very large amount of memory. (It kept on making new
5163 copies of the header line as it rewrote, and never released the old ones.)
5164 At the expense of a bit more processing, the header rewriting function has
5165 been changed so that it no longer eats memory in this way.
5167 66. The generation of the Received: header has been moved from the time that a
5168 message starts to be received, to the time that it finishes. The timestamp
5169 in the Received: header should now be very close to that of the <= log
5170 line. There are two side-effects of this change:
5172 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5173 logged header lines no longer include the local Received: line, because
5174 it has not yet been created. The same applies to a copy of the message
5175 that is returned to a non-SMTP sender when a message is rejected.
5177 (b) When a filter file is tested using -bf, no additional Received: header
5178 is added to the test message. After some thought, I decided that this
5181 This change does not affect the value of $received_for. It is still set
5182 after address rewriting, but before local_scan() is called.
5184 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5186 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5187 gave an unhelpful panic error message, and a defer error. I have managed to
5188 change this behaviour so that it now rejects any supplied certificate,
5189 which seems right, as the list of acceptable certificates is empty.
5191 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5192 gave an unhelpful defer error. I have not managed to make this reject any
5193 supplied certificates, but the error message it gives is "no certificate
5194 supplied", which is not helpful.
5196 70. exigrep's output now also includes lines that are not associated with any
5197 message, but which match the given pattern. Implemented by a patch from
5198 Martin Sluka, which also tidied up the Perl a bit.
5200 71. Recipient callout verification, like sender verification, was using <> in
5201 the MAIL FROM command. This isn't really the right thing, since the actual
5202 sender may affect whether the remote host accepts the recipient or not. I
5203 have changed it to use the actual sender in the callout; this means that
5204 the cache record is now keyed on a recipient/sender pair, not just the
5205 recipient address. There doesn't seem to be a real danger of callout loops,
5206 since a callout by the remote host to check the sender would use <>.
5207 [SEE ABOVE: changed after hitting problems.]
5209 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5210 temporary errors. However, in the case of such a code being given after
5211 the end of a data transmission (i.e. after ".") Exim was failing to write
5212 a retry record for the message. (Yes, there was some broken host that was
5213 actually sending 8xx at this point.)
5215 73. An unknown lookup type in a host list could cause Exim to panic-die when
5216 the list was checked. (An example that provoked this was putting <; in the
5217 middle of a list instead of at the start.) If this happened during a DATA
5218 ACL check, a -D file could be left lying around. This kind of configuration
5219 error no longer causes Exim to die; instead it causes a defer errror. The
5220 incident is still logged to the main and panic logs.
5222 74. Buglet left over from Exim 3 conversion. The message "too many messages
5223 in one connection" was written to the rejectlog but not the mainlog, except
5224 when address rewriting (yes!) was being logged.
5226 75. Added write_rejectlog option.
5228 76. When a system filter was run not as root (that is, when system_filter_user
5229 was set), the values of the $n variables were not being returned to the
5230 main process; thus, they were not subsequently available in the $sn
5233 77. Added +return_path_on_delivery log selector.
5235 78. A connection timeout was being treated differently from recipients deferred
5236 when testing hosts_max_try with a message that was older than the host's
5237 retry timeout. (The host should not be counted, thus allowing all hosts to
5238 be tried at least once before bouncing.) This may have been the cause of an
5239 occasionally reported bug whereby a message would remain on the queue
5240 longer than the retry timeout, but would be bounced if a delivery was
5241 forced. I say "may" because I never totally pinned down the problem;
5242 setting up timeout/retry tests is difficult. See also the next item.
5244 79. The ultimate address timeout was not being applied to errors that involved
5245 a combination of host plus message (for example, a timeout on a MAIL
5246 command). When an address resolved to a number of possible hosts, and they
5247 were not all tried for each delivery (e.g. because of hosts_max_try), a
5248 message could remain on the queue longer than the retry timeout.
5250 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5253 81. Fixed an obscure SMTP outgoing bug which required at least the following
5254 conditions: (a) there was another message waiting for the same server;
5255 (b) the server returned 5xx to all RCPT commands in the first message so
5256 that the message was not completed; (c) the server dropped the connection
5257 or gave a negative response to the RSET that Exim sends to abort the
5258 transaction. The observed case was a dropped connection after DATA that had
5259 been sent in pipelining mode. That is, the server had advertised PIPELINING
5260 but was not implementing it correctly. The effect of the bug was incorrect
5261 behaviour, such as trying another host, and this could lead to a crash.
5267 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5268 and daemon.c were passed as pointers to ints; they should have been
5269 pointers to socklen_t variables (which are typically unsigned ints).
5271 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5274 3. Fixed a really odd bug that affected only the testing scheme; patching a
5275 certain fixed string in the binary changed the value of another string that
5276 happened to be identical to the end of the original first string.
5278 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5279 name", it returns that address as the IP address. On some operating
5280 systems (e.g. Solaris), it also passes back the IP address string as the
5281 "host name". However, on others (e.g. Linux), it passes back an empty
5282 string. Exim wasn't checking for this, and was changing the host name to an
5283 empty string, assuming it had been canonicized.
5285 5. Although rare, it is permitted to have more than one PTR record for a given
5286 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5287 all the names associated with an address, because they do in Solaris.
5288 However, it seems that they do not in Linux for data that comes from the
5289 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5290 I found this out when I moved to a new Linux workstation and tried to run
5291 the Exim test suite.
5293 To get round this problem I have changed the code so that it now does its
5294 own call to the DNS to look up PTR records when searching for a host name.
5295 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5296 addresses that are only in /etc/hosts are still found.
5298 This behaviour is, however, controlled by an option called host_lookup_
5299 order, which defaults to "bydns:byaddr". If people want to use the other
5300 order, or indeed, just use one or the other means of lookup, they can
5301 specify it in this variable.
5303 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5304 some operating systems, this comes back from gethostbyaddr() as an empty
5305 string, and this is what Exim used to test for. However, it seems that in
5306 other systems, "." is yielded. Exim now tests for this case too.
5308 7. The values of check_spool_space and check_log_space are now held internally
5309 as a number of kilobytes instead of an absolute number of bytes. If a
5310 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5311 kilobyte. This means that much larger values can be stored.
5313 8. Exim monitor: an attempt to get the action menu when not actually pointing
5314 at a message produces an empty menu entitled "No message selected". This
5315 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5316 no entries in it ("Shell widget menu has zero width and/or height"). So I
5317 have added a single, blank menu entry in this case.
5319 9. Added ${quote_local_part.
5321 10. MIME decoding is now applied to the contents of Subject: header lines when
5324 11. Now that a reference to $sender_host_address automatically causes a reverse
5325 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5326 host lookup before query-style lookups in lists that might use this
5327 variable. This has therefore been abolished, and the "net-" prefix is no
5328 longer necessary for query-style lookups.
5330 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5331 have been a typo for LFLAGS, so it has been changed.
5333 13. The install script calls Exim with "-C /dev/null" in order to find the
5334 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5335 to be output. Howeve, since Exim outputs its version number before the
5336 error, it didn't break the script. It just looked ugly. I fixed this by
5337 always allowing "-C /dev/null" if the caller is root.
5339 14. Ignore overlarge ACL variable number when reading spool file - insurance
5340 against a later release with more variables having written the file.
5342 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5343 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5344 this, as well as the form without the "IPv6" on the front.
5346 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5347 OS/Makefile-Darwin file.
5349 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5350 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5352 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5353 variable. However, some OS use ints for inodes. I've added cast to long int
5354 to get rid of the compiler warning.
5356 19. I had forgotten to lock out "/../" in configuration file names when
5357 ALT_CONFIG_PREFIX was set.
5359 20. Routers used for verification do not need to specify transports. However,
5360 if such a router generated a host list, and callout was configured, Exim
5361 crashed, because it could not find a port number from the (non-existent)
5362 transport. It now assumes port 25 in this circumstance.
5364 21. Added the -t option to exigrep.
5366 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5367 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5368 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5369 them is set, LOOKUP_LSEARCH is forced.
5371 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5373 24. Added sender and host information to the "rejected by local_scan()" log
5374 line; previously there was no indication of these.
5376 25. Added .include_if_exists.
5378 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5379 Linux. It turns out that not all file systems support this. Apparently some
5380 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5381 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5382 not supported on the file descriptor, is now ignored when Exim is trying to
5383 sync a directory. This applies only to Linux.
5385 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5387 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5388 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5391 29. Added support for maildirsize files from supplied patch (modified a bit).
5393 30. The use of :fail: followed by an empty string could lead Exim to respond to
5394 sender verification failures with (e.g.):
5396 550 Verification failed for <xxx>
5397 550 Sender verify failed
5399 where the first response line was missing the '-' that indicates it is not
5400 the final line of the response.
5402 31. The loop for finding the name of the user that called Exim had a hardwired
5403 limit of 10; it now uses the value of finduser_retries, which is used for
5404 all other user lookups.
5406 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5409 33. Exim was neglecting to zero errno before one call of strtol() when
5410 expanding a string and expecting an integer value. On some systems this
5411 resulted in spurious "integer overflow" errors. Also, it was casting the
5412 result into an int without checking.
5414 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5415 did not work. The code looks as if it has *never* worked, though it appears
5416 to have been documented since at least releast 1.62. I have made it work.
5418 35. The "timeout_DNS" error in retry rules, also documented since at least
5419 1.62, also never worked. As it isn't clear exactly what this means, and
5420 clearly it isn't a major issue, I have abolished the feature by treating it
5421 as "timeout", and writing a warning to the main and panic logs.
5423 36. The display of retry rules for -brt wasn't always showing the error code
5426 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5427 timeout_connect_A, timeout_connect_MX.
5429 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5430 to the empty sender.
5432 39. The daemon was not analysing the content of -oX till after it had closed
5433 stderr and disconnected from the controlling terminal. This meant that any
5434 syntax errors were only noted on the panic log, and the return code from
5435 the command was 0. By re-arranging the code a little, I've made the
5436 decoding happen first, so such errors now appear on stderr, and the return
5437 code is 1. However, the actual setting up of the sockets still happens in
5438 the disconnected process, so errors there are still only recorded on the
5441 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5442 connections (as happens on some IP stacks) was logged at start up time as
5443 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5444 it from "IPv6 and IPv4", which means that two separate sockets are being
5447 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5448 says whether AF_INET or AF_INET6 was passed as an argument.
5450 42. Exiwhat output was messed up when time zones were included in log
5453 43. Exiwhat now gives more information about the daemon's listening ports,
5454 and whether -tls-on-connect was used.
5456 44. The "port" option of the smtp transport is now expanded.
5458 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5459 silently ignored. Now an error message is written to the main and panic
5462 46. There's a new ACL modifier called "logwrite" which writes to a log file
5463 as soon as it is encountered.
5465 47. Added $local_user_uid and $local_user_gid at routing time.
5467 48. Exim crashed when trying to verify a sender address that was being
5470 49. Exim was recognizing only a space character after ".include". It now also
5471 recognizes a tab character.
5473 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5474 extracting the relevant information from the specification. The man page no
5475 longer contains scrambled data for the -d option, and I've added a section
5476 at the front about calling Exim under different names.
5478 51. Added "extra_headers" argument to the "mail" command in filter files.
5480 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5483 53. Installed eximstats 1.29.
5485 54. Added transport_filter_timeout as a generic transport option.
5487 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5488 Cc: header lines. This was required by RFC 822, but it not required by RFC
5491 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5492 incoming messages that did not have them. Now it does so only if the
5493 message originates locally, that is, if there is no associated remote host
5494 address. When Resent- header lines are present, this applies to the Resent-
5495 lines rather than the non-Resent- lines.
5497 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5498 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5500 58. Messages for configuration errors now include the name of the main
5501 configuration file - useful now that there may be more than one file in a
5502 list (.included file names were always shown).
5504 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5505 for those rare installations that do not start the daemon as root or run it
5506 setuid root. I've cut out the call to initgroups() if the daemon is not
5509 60. The Exim user and group can now be bound into the binary as text strings
5510 that are looked up at the start of Exim's processing.
5512 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5514 62. Added $mailstore_basename variable.
5516 63. Installed patch to sieve.c from Michael Haardt.
5518 64. When Exim failed to open the panic log after failing to open the main log,
5519 the original message it was trying to log was written to stderr and debug
5520 output, but if they were not available (the usual case in production), it
5521 was lost. Now it is written to syslog before the two lines that record the
5522 failures to open the logs.
5524 65. Users' Exim filters run in subprocesses under the user's uid. It is
5525 possible for a "deliver" command or an alias in a "personal" command to
5526 provoke an address rewrite. If logging of address rewriting is configured,
5527 this fails because the process is not running as root or exim. There may be
5528 a better way of dealing with this, but for the moment (because 4.30 needs
5529 to be released), I have disabled address rewrite logging when running a
5530 filter in a non-root, non-exim process.
5536 1. The buildconfig auxiliary program wasn't quoting the value set for
5537 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5538 not defined. This bug was masked in 4.22 by the effect that was fixed in
5541 2. Some messages that were rejected after a message id was allocated were
5542 shown as "incomplete" by exigrep. It no longer does this for messages that
5543 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5545 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5546 have allow_domain_literals set, the ID did not get logged in the <= line.
5547 Domain literals are now always recognized in Message-ID: header lines.
5549 4. The first argument for a ${extract expansion item is the key name or field
5550 number. Leading and trailing spaces in this item were not being ignored,
5551 causing some misleading effects.
5553 5. When deliver_drop_privilege was set, single queue runner processes started
5554 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5555 same command in the process it spins off) were not dropping privilege.
5557 6. When the daemon running as "exim" started a queue runner, it always
5558 re-executed Exim in the spun-off process. This is a waste of effort when
5559 deliver_drop_privilege is set. The new process now just calls the
5560 queue-runner function directly.
5566 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5569 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5570 ignored. Though the use of -oP was forcing the writing of a pid file, it
5571 was always written to the default place.
5573 3. If the message "no IP address found for host xxxx" is generated during
5574 incoming verification, it is now followed by identification of the incoming
5575 connection (so you can more easily find what provoked it).
5577 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5579 5. Added some features to "harden" Exim a bit more against certain attacks:
5581 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5582 be put in Local/Makefile. This is like the never_users runtime option,
5583 but it cannot be overridden. The default setting is "root".
5585 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5586 prefix string with which any file named in a -C command line option
5589 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5590 is retained for -C and -D only if the caller of Exim is root. Without
5591 it, the exim user may also use -C and -D and retain privilege.
5593 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5594 command line option is disabled.
5596 6. Macro names set by the -D option must start with an upper case letter, just
5597 like macro names defined in the configuration file.
5599 7. Added "dereference=" facility to LDAP.
5601 8. Two instances of the typo "uknown" in the source files are fixed.
5603 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5604 the Configure-Makefile script screwed up while processing it.
5606 10. Incorporated PCRE 4.4.
5608 11. The SMTP synchronization check was not operating right at the start of an
5609 SMTP session. For example, it could not catch a HELO sent before the client
5610 waited for the greeting. There is now a check for outstanding input at the
5611 point when the greeting is written. Because of the duplex, asynchronous
5612 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5613 way, but not yet received, when the check is performed.
5615 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5616 on TCP/IP sockets, because this apparently causes some broken clients to
5619 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5620 unchanged) from the Cygwin maintainer.
5622 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5623 was supported instead of testing for mbx. Effectively a typo.
5625 15. The spa authenticator server code was not checking that the input it
5626 received was valid base64.
5628 16. The debug output line for the "set" modifier in ACLs was not showing the
5629 name of the variable that was being set.
5631 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5633 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5634 lookup on its own. Something else was needed to trigger the lookup. For
5635 example, a match in host_lookup or the need for a host name in a host list.
5636 Now, if $sender_host_name is referenced and the host name has not yet been
5637 looked up, a lookup is performed. If the lookup fails, the variable remains
5638 empty, and $host_lookup_failed is set to "1".
5640 19. Added "eqi" as a case-independent comparison operator.
5642 20. The saslauthd authentication condition could segfault if neither service
5643 nor realm was specified.
5645 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5646 error message that was logged was misleading, and incoming SMTP
5647 connections were dropped. The message is now more accurate, and temporary
5648 errors are given to SMTP connections.
5650 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5651 (except RSET) with a 421 error, until QUIT is received. However, it was
5652 failing to send a response to QUIT.
5654 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5655 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5656 ACL is now run after the helo_try_verify_hosts code.
5658 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5659 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5660 case-independent, so other case variants are also recognized). Apparently
5661 some systems use these upper case variants.
5663 25. If more than two messages were waiting for the same host, and a transport
5664 filter was specified for the transport, Exim sent two messages over the
5665 same TCP/IP connection, and then failed with "socket operation on non-
5666 socket" when it tried to send the third.
5668 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5670 27. The extern definition of crypt16() in expand.c was not being excluded when
5671 the OS had its own crypt16() function.
5673 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5674 as a preferred synonym for return_size_limit, both as an option and as an
5677 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5679 30. Changed the default configuration ACL to relax the local part checking rule
5680 for addresses that are not in any local domains. For these addresses,
5681 slashes and pipe symbols are allowed within local parts, but the sequence
5682 /../ is explicitly forbidden.
5684 31. SPA server authentication was not clearing the challenge buffer before
5687 32. log_message in a "warn" ACL statement was writing to the reject log as
5688 well as to the main log, which contradicts the documentation and doesn't
5689 seem right (because no rejection is happening). So I have stopped it.
5691 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5692 However, I am unable to do any testing of this.
5694 34. Fixed an infelicity in the appendfile transport. When checking directories
5695 for a mailbox, to see if any needed to be created, it was accidentally
5696 using path names with one or more superfluous leading slashes; tracing
5697 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5699 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5700 contents are added to the log line that is written for every discarded
5701 recipient. (Previously a log_message setting was ignored.)
5703 36. The ${quote: operator now quotes the string if it is empty.
5705 37. The install script runs exim in order to find its version number. If for
5706 some reason other than non-existence or emptiness, which it checks, it
5707 could not run './exim', it was installing it with an empty version number,
5708 i.e. as "exim-". This error state is now caught, and the installation is
5711 38. An argument was missing from the function that creates an error message
5712 when Exim fails to connect to the socket for saslauthd authentication.
5713 This could cause Exim to crash, or give a corrupted message.
5715 39. Added isip, isip4, and isip6 to ${if conditions.
5717 40. The ACL variables $acl_xx are now saved with the message, and can be
5718 accessed later in routers, transports, and filters.
5720 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5721 strings in the file are not string-expanded.
5723 42. If a MAIL command specified a SIZE value that was too large to fit into an
5724 int variable, the check against message_size_limit failed. Such values are
5725 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5726 day this will have to be increased, but I don't think I want to be around
5727 when emails are that large.
5734 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5735 iconv() is not standard in FreeBSD.
5737 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5738 IPv6 enabled. The observed symptom was a segmentation fault on return from
5739 the function os_common_find_running_interfaces() in src/os.c.
5741 3. In the check_special_case() function in daemon.c I had used "errno" as an
5742 argument name, which causes warnings on some systems. This was basically a
5743 typo, since it was named "eno" in the comments!
5745 4. The code that waits for the clock to tick (at a resolution of some fraction
5746 of a second) so as to ensure message-id uniqueness was always waiting for
5747 at least one whole tick, when it could have waited for less. [This is
5748 almost certainly not relevant at current processor speeds, where it is
5749 unlikely to ever wait at all. But we try to future-proof.]
5751 5. The function that sleeps for a time interval that includes fractions of a
5752 second contained a race. It did not block SIGALRM between setting the
5753 timer, and suspending (a couple of lines later). If the interval was short
5754 and the sigsuspend() was delayed until after it had expired, the suspension
5755 never ended. On busy systems this could lead to processes getting stuck for
5758 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5759 process, before it forks any delivery processes. The open lookup caching
5760 mechanism meant that the open file or database connection was passed into
5761 the delivery process. The problem was that delivery processes always tidy
5762 up cached lookup data. This could cause a problem for the next delivery
5763 process started by the queue runner, because the external queue runner
5764 process does not know about the closure. So the next delivery process
5765 still has data in the lookup cache. In the case of a file lookup, there was
5766 no problem because closing a file descriptor in a subprocess doesn't affect
5767 the parent. However, if the lookup was caching a connection to a database,
5768 the connection was closed, and the second delivery process was likely to
5769 see errors such as "PGSQL: query failed: server closed the connection
5770 unexpectedly". The problem has been fixed by closing all cached lookups
5771 in a queue runner before running a delivery process.
5773 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5774 seem to have the "const" qualifier which it has on other OS. I've
5777 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5778 *of the same type* (client or server) with the same public name that an
5779 error should be diagnosed.
5781 9. When Exim looked up a host name for an IP address, but failed to find the
5782 original IP address when looking up the host name (a safety check), it
5783 output the message "<ip address> does not match any IP for NULL", which was
5784 confusing, to say the least. The bug was that the host name should have
5785 appeared instead of "NULL".
5787 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5788 user that is built into the binary, and the -C or -D options is used, root
5789 privilege is dropped before the configuration file is read. In addition,
5790 logging is switched to stderr instead of the normal log files. If the
5791 configuration then re-defines the Exim user, the unprivileged environment
5792 is probably not what is expected, so Exim logs a panic warning message (but
5795 However, if deliver_drop_privilege is set, the unprivileged state may well
5796 be exactly what is intended, so the warning has been cut out in that case,
5797 and Exim is allowed to try to write to its normal log files.
5803 1. smtp_return_error_details was not giving details for temporary sender
5804 or receiver verification errors.
5806 2. Diagnose a configuration error if two authenticators have the same public
5809 3. Exim used not to create the message log file for a message until the first
5810 delivery attempt. This could be confusing when incoming messages were held
5811 for policy or load reasons. The message log file is now created at the time
5812 the message is received, and an initial "Received" line is written to it.
5814 4. The automatically generated man page for command line options had a minor
5815 bug that caused no ill effects; however, a more serious problem was that
5816 the procedure for building the man page automatically didn't always
5817 operate. Consequently, release 4.20 contains an out-of-date version. This
5818 shouldn't happen again.
5820 5. When building Exim with embedded Perl support, the script that builds the
5821 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5822 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5824 6. The freeze_tell option was not being used for messages that were frozen on
5825 arrival, either by an ACL or by local_scan().
5827 7. Added the smtp_incomplete_transaction log selector.
5829 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5830 was accepting AUTH without a new EHLO.
5832 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5833 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5836 10. Logging of TCP/IP connections (when configured) now happens in the main
5837 daemon process instead of the child process, so that the TCP/IP connection
5838 count is more accurate (but it can never be perfect).
5840 11. The use of "drop" in a nested ACL was not being handled correctly in the
5841 outer ACL. Now, if condition failure induced by the nested "drop" causes
5842 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5843 or "require"), the connection is dropped.
5845 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5846 that yield "discard" can now be used with an "accept" or a "discard" verb,
5847 but an error is generated for any others (because I can't see a useful way
5848 to define what should happen).
5850 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5851 are now processed in the same way as lines in the Exim configuration file.
5852 In particular, continuation lines are supported.
5854 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5856 15. Added -ti meaning -t -i.
5858 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5859 domains, and warn by logging if others are found.
5861 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5862 returned from GIFCONF when Exim is trying to find the list of interfaces on
5863 a host. The code in os.c has been modified to copy each ifreq to an aligned
5864 structure in all cases.
5866 Also, in some cases, the returned ifreq's were being copied to a 'struct
5867 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5868 means the last couple of bytes of an IPv6 address could be chopped if the
5869 ifreq contained only a normal sockaddr (14 bytes storage).
5871 18. Named domain lists were not supported in the hosts_treat_as_local option.
5872 An entry such as +xxxx was not recognized, and was treated as a literal
5875 19. Ensure that header lines added by a DATA ACL are included in the reject log
5876 if the ACL subsequently rejects the message.
5878 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5879 MD5 (which is deprecated).
5881 21. When testing a filter file using -bf, Exim was writing a message when it
5882 took the sender from a "From " line in the message, but it was not doing so
5883 when it took $return_path from a Return-Path: header line. It now does.
5885 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5886 with a valid header line field name (a series of printing characters
5887 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5889 23. Changed "disc" in the source to "disk" to conform to the documentation and
5890 the book and for uniformity.
5892 24. Ignore Sendmail's -Ooption=value command line item.
5894 25. When execve() failed while trying to run a command in a pipe transport,
5895 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5896 could be confused with a return value of 69 from the command itself. This
5897 has been changed to 127, the value the shell returns if it is asked to run
5898 a non-existent command. The wording for the related log line suggests a
5899 non-existent command as the problem.
5901 26. If received_header_text expands to an empty string, do not add a Received:
5902 header line to the message. (Well, it adds a token one on the spool, but
5903 marks it "old" so that it doesn't get used or transmitted.)
5905 27. Installed eximstats 1.28 (addition of -nt option).
5907 28. There was no check for failure on the call to getsockname() in the daemon
5908 code. This can fail if there is a shortage of resources on the system, with
5909 ENOMEM, for example. A temporary error is now given on failure.
5911 29. Contrary to the C standard, it seems that in some environments, the
5912 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5913 program. Exim now does this explicitly; it affects the formatting of
5914 timestamps using strftime().
5916 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5917 complaints. I've now initialized all the variables, to avoid this.
5919 32. Header lines added by a system filter were not being "seen" during
5920 transport-time rewrites.
5922 33. The info_callback() function passed to OpenSSL is set up with type void
5923 (*)(SSL *, int, int), as described somewhere. However, when calling the
5924 function (actually a macro) that sets it up, the type void(*)() is
5925 expected. I've put in a cast to prevent warnings from picky compilers.
5927 34. If a DNS black list lookup found a CNAME record, but there were no A
5928 records associated with the domain it pointed at, Exim crashed.
5930 35. If a DNS black list lookup returned more than one A record, Exim ignored
5931 all but the first. It now scans all returned addresses if a particular IP
5932 value is being sought. In this situation, the contents of the
5933 $dnslist_value variable are a list of all the addresses, separated by a
5936 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5937 to accept a host name and all its aliases if the forward lookup for any of
5938 them yielded the IP address of the incoming connection. Now it accepts only
5939 those names whose forward lookup yields the correct IP address. Any other
5940 names are discarded. This closes a loophole whereby a rogue DNS
5941 administrator could create reverse DNS records to break through a
5942 wildcarded host restriction in an ACL.
5944 37. If a user filter or a system filter that ran in a subprocess used any of
5945 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5946 the wrong values were passed to the pipe command ($thisaddress had the
5947 value of $0, $0 had the value of $1, etc). This bug was introduced by
5948 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5950 38. Improved the line breaking for long SMTP error messages from ACLs.
5951 Previously, if there was no break point between 40 and 75 characters, Exim
5952 left the rest of the message alone. Two changes have been made: (a) I've
5953 reduced the minimum length to 35 characters; (b) if it can't find a break
5954 point between 35 and 75 characters, it looks ahead and uses the first one
5955 that it finds. This may give the occasional overlong line, but at least the
5956 remaining text gets split now.
5958 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5959 file descriptors might be low, and that setting 1000 would always raise it.
5960 It turns out that in some environments, the limit is already over 1000 and
5961 that lowering it causes trouble. So now Exim takes care not to decrease it.
5963 40. When delivering a message, the value of $return_path is set to $sender_
5964 address at the start of routing (routers may change the value). By an
5965 oversight, this default was not being set up when an address was tested by
5966 -bt or -bv, which affected the outcome if any router or filter referred to
5969 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5970 log only when "message" or "log_message" are set. However, if one of the
5971 conditions was an address verification, or a call to a nested ACL, the
5972 messages generated by the underlying test were being passed through. This
5973 no longer happens. The underlying message is available in $acl_verify_
5974 message for both "message" and "log_message" expansions, so it can be
5975 passed through if needed.
5977 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5978 new expansion $bh_ to give the encoded byte string without charset
5979 translation. Translation happens only if iconv() is available; HAVE_ICONV
5980 indicates this at build time. HEADERS_CHARSET gives the charset to
5981 translate to; headers_charset can change it in the configuration, and
5982 "headers charset" can change it in an individual filter file.
5984 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5985 that creates RFC 2047 encoded "words" labels them as that charset instead
5986 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5987 expansion operator; (ii) when Exim creates a From: line for a local
5988 message; (iii) when a header line is rewritten to include a "phrase" part.
5990 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5991 buggy, causing it to skip the first lines of messages whose message ID
5992 ended in 'D'. This would not have bitten before Exim release 4.14, because
5993 message IDs were unlikely to end in 'D' before then. The effect was to have
5994 incorrect size information for certain domains.
5996 45. #include "config.h" was missing at the start of the crypt16.c module. This
5997 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6000 46. If there was a timeout during a "random" callout check, Exim treated it as
6001 a failure of the random address, and carried on sending RSET and the real
6002 address. If the delay was just some slowness somewhere, the response to the
6003 original RCPT would be taken as a response to RSET and so on, causing
6004 mayhem of various kinds.
6006 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6007 when I implemented it. It didn't allow for the fact that some option values
6008 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6009 the right test for positive values.
6011 48. Domain names in DNS records are case-independent. Exim always looks them up
6012 in lower case. Some resolvers return domain names in exactly the case they
6013 appear in the zone file, that is, they may contain uppercase letters. Not
6014 all resolvers do this - some return always lower case. Exim was treating a
6015 change of case by a resolver as a change of domain, similar to a widening
6016 of a domain abbreviation. This triggered its re-routing code and so it was
6017 trying to route what was effectively the same domain again. This normally
6018 caused routing to fail (because the router wouldn't handle the domain
6019 twice). Now Exim checks for this case specially, and just changes the
6020 casing of the domain that it ultimately uses when it transmits the message
6023 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6026 50. If a filter generated a file delivery with a non-absolute name (possible if
6027 no home directory exists for the router), the forbid_file option was not
6030 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6031 the existing equality matching.
6033 52. Exim was using ints instead of ino_t variables in some places where it was
6034 dealing with inode numbers.
6036 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6037 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6038 called TMPDIR, and if it finds it is different, it changes its value.
6040 54. The smtp_printf() function is now made available to local_scan() so
6041 additional output lines can be written before returning. There is also an
6042 smtp_fflush() function to enable the detection of a dropped connection.
6043 The variables smtp_input and smtp_batched_input are exported to
6046 55. Changed the default runtime configuration: the message "Unknown user"
6047 has been removed from the ACL, and instead placed on the localuser router,
6048 using the cannot_route_message feature. This means that any verification
6049 failures that generate their own messages won't get overridden. Similarly,
6050 the "Unrouteable address" message that was in the ACL for unverifiable
6051 relay addresses has also been removed.
6053 56. Added hosts_avoid_esmtp to the smtp transport.
6055 57. The exicyclog script was not checking for the esoteric option
6056 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6057 will work only if exicyclog is run under the appropriate euid.
6059 58. Following a discussion on the list, the rules by which Exim recognises line
6060 endings on incoming messages have been changed. The -dropcr and drop_cr
6061 options are now no-ops, retained only for backwards compatibility. The
6062 following line terminators are recognized: LF CRLF CR. However, special
6063 processing applies to CR:
6065 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6066 nor a local message in the state where . is a terminator.
6068 (ii) If a bare CR is encountered in a header line, an extra space is added
6069 after the line terminator so as not to end the header. The reasoning
6070 behind this is that bare CRs in header lines are most likely either
6071 to be mistakes, or people trying to play silly games.
6073 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6074 was being incorrectly given as 18 bytes larger than it should have been.
6075 This is a VOB (very old bug).
6077 60. This may never have affected anything current, but just in case it has:
6078 When the local host is found other than at the start of a list of hosts,
6079 the local host, those with the same MX, and any that follow, are discarded.
6080 When the list in question was part of a longer list of hosts, the following
6081 hosts (not currently being processed) were also being discarded. This no
6082 longer happens. I'm not sure if this situation could ever has previously
6085 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6088 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6089 header. This is something that is recommended in a new Internet Draft, and
6090 is something that is documented as being done by Sendmail. There are two
6091 possible values. For messages generated by the autoreply transport, Exim
6094 Auto-Submitted: auto-replied
6096 whereas for all other generated messages (e.g. bounces) it adds
6098 Auto-Submitted: auto-generated
6100 63. The "personal" condition in filters now includes a test for the
6101 Auto-Submitted: header. If it contains the string "auto-" the message it
6102 not considered personal.
6104 64. Added rcpt_include_affixes as a generic transport option.
6106 65. Added queue_only_override (default true).
6108 66. Added the syslog_duplication option.
6110 67. If what should have been the first header line of a message consisted of
6111 a space followed by a colon, Exim was mis-interpreting it as a header line.
6112 It isn't of course - it is syntactically invalid and should therefore be
6113 treated as the start of the message body. The misbehaviour could have
6114 caused a number of strange effects, including loss of data in subsequent
6115 header lines, and spool format errors.
6117 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6118 client host had authenticated. This control can now be exercised by an ACL
6119 for more flexibility.
6121 69. By default, callouts do not happen when testing with -bh. There is now a
6122 variant, -bhc, which does actually run the callout code, including
6123 consulting and updating the callout cache.
6125 70. Added support for saslauthd authentication, courtesy of Alexander
6128 71. If statvfs() failed on the spool or log directories while checking their
6129 size for availability, Exim confusingly gave the error "space shortage".
6130 Furthermore, in debugging mode it crashed with a floating point exception.
6131 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6132 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6133 serious problem, Exim now writes to the main and panic logs when this
6134 happens, with details of the failure. It then refuses to accept the
6135 incoming message, giving the message "spool directory problem" or "log
6136 directory problem" with a 421 code for SMTP messages.
6138 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6139 0, 1, and 2 exist, because some OS complain for execs without them (see
6140 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6141 descriptors. However, the code omitted to check that the open succeeded,
6142 causing mysterious errors if for some reason the permissions on /dev/null
6143 got screwed. Now Exim writes a message to the main and panic logs, and
6144 bombs out if it can't open /dev/null.
6146 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6147 interact so that it is all more flexible. It is supposed to remain
6148 backwards compatible. Also added extra_local_interfaces.
6150 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6151 to bomb out with an assertion failure - to the client this appears as a
6152 connection drop. This problem occurs in the part of the code that was taken
6153 from the Samba project. Fortunately, the assertion is in a very simple
6154 function, so I have fixed this by reproducing the function inline in the
6155 one place where it is called, and arranging for authentication to fail
6156 instead of killing the process with assert().
6158 75. The SPA client code was not working when the server requested OEM rather
6159 than Unicode encoding.
6161 76. Added code to make require_files with a specific uid setting more usable in
6162 the case where statting the file as root fails - usually a non-root-mounted
6163 NFS file system. When this happens and the failure is EACCES, Exim now
6164 forks a subprocess and does the per-uid checking as the relevant uid.
6166 77. Added process_log_path.
6168 78. If log_file_path was not explicitly set, a setting of check_log_space or
6169 check_log_inodes was ignored.
6171 79. If a space check for the spool or log partitions fails, the incident is now
6172 logged. Of course, in the latter case the data may get lost...
6174 80. Added the %p formatting code to string_format() so that it can be used to
6175 print addresses in debug_print(). Adjusted all the address printing in the
6176 debugging in store.c to use %p rather than %d.
6178 81. There was a concern that a line of code in smtp_in.c could overflow a
6179 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6180 initially expressed, the concern was not well-founded, because trailing
6181 spaces are removed early. However, if the trailing spaces were followed by
6182 a NULL, they did not get removed, so the overflow was possible. Two fixes
6185 (a) I re-wrote the offending code in a cleaner fashion.
6186 (b) If an incoming SMTP command contains a NULL character, it is rejected
6189 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6190 runs initgroups(), so that if the Exim user is in any additional groups,
6191 they will be used during message reception.
6197 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob