1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and backup-only
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
68 TL/01 When running the test suite, the README says that variables such as
69 no_msglog_check are global and can be placed anywhere in a specific
70 test's script, however it was observed that placement needed to be near
71 the beginning for it to behave that way. Changed the runtest perl
72 script to read through the entire script once to detect and set these
73 variables, reset to the beginning of the script, and then run through
74 the script parsing/test process like normal.
76 TL/02 The BSD's have an arc4random API. One of the functions to induce
77 adding randomness was arc4random_stir(), but it has been removed in
78 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
79 function when detected.
81 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
82 cause callback expansion.
84 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
85 syntax errors in an expansion can be treated as a string instead of
86 logging or causing an error, due to the internal use of bool_lax
87 instead of bool when processing it.
89 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
90 server certificates when making smtp deliveries.
92 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
94 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
96 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
98 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
99 Merged patch from Sebastian Wiedenroth.
101 JH/05 Fix results-pipe from transport process. Several recipients, combined
102 with certificate use, exposed issues where response data items split
103 over buffer boundaries were not parsed properly. This eventually
104 resulted in duplicates being sent. This issue only became common enough
105 to notice due to the introduction of conection certificate information,
106 the item size being so much larger. Found and fixed by Wolfgang Breyha.
108 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
109 size buffer was used, resulting in syntax errors when an expansion
112 JH/07 Add support for directories of certificates when compiled with a GnuTLS
113 version 3.3.6 or later.
115 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
116 is EXPERIMENTAL_EVENT, the main-configuration and transport options
117 both become "event_action", the variables become $event_name, $event_data
118 and $event_defer_errno. There is a new variable $verify_mode, usable in
119 routers, transports and related events. The tls:cert event is now also
120 raised for inbound connections, if the main configuration event_action
123 TL/06 In test suite, disable OCSP for old versions of openssl which contained
124 early OCSP support, but no stapling (appears to be less than 1.0.0).
126 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
127 server certificate names available under the smtp transport option
128 "tls_verify_cert_hostname" now do not permit multi-component wildcard
131 JH/10 Time-related extraction expansions from certificates now use the main
132 option "timezone" setting for output formatting, and are consistent
133 between OpenSSL and GnuTLS compilations. Bug 1541.
135 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
136 encoded parameter in the incoming message. Bug 1558.
138 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
139 include certificate info, eximon was claiming there were spoolfile
142 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
144 JH/14 Log delivery-related information more consistently, using the sequence
145 "H=<name> [<ip>]" wherever possible.
147 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
148 are problematic for Debian distribution, omit them from the release
151 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
153 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
156 JH/17 Fix a null-indirection in certextract expansions when a nondefault
157 output list separator was used.
162 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
163 checkers that were complaining about end of non-void function with no
166 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
167 This was a regression intruduced in 4.83 by another bugfix.
169 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
171 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
172 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
178 TF/01 Correctly close the server side of TLS when forking for delivery.
180 When a message was received over SMTP with TLS, Exim failed to clear up
181 the incoming connection properly after forking off the child process to
182 deliver the message. In some situations the subsequent outgoing
183 delivery connection happened to have the same fd number as the incoming
184 connection previously had. Exim would try to use TLS and fail, logging
185 a "Bad file descriptor" error.
187 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
188 utilities have not been installed.
190 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
191 temporary space as the ACL may create new global variables.
193 TL/01 LDAP support uses per connection or global context settings, depending
194 upon the detected version of the libraries at build time.
196 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
197 to extract and use the src ip:port in logging and expansions as if it
198 were a direct connection from the outside internet. PPv2 support was
199 updated based on HAProxy spec change in May 2014.
201 JH/02 Add ${listextract {number}{list}{success}{fail}}.
203 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
204 Properly escape header and check for NULL return.
206 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
209 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
211 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
212 characters in header names, implemented as a verify condition.
213 Contributed by Michael Fischer v. Mollard.
215 TL/05 Rename SPF condition results err_perm and err_temp to standardized
216 results permerror and temperror. Previous values are deprecated but
217 still accepted. In a future release, err_perm and err_temp will be
218 completely removed, which will be a backward incompatibility if the
219 ACL tests for either of these two old results. Patch contributed by
220 user bes-internal on the mailing list.
222 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
224 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
225 selectors, in both main and reject logs.
227 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
230 JH/07 Add malware type "sock" for talking to simple daemon.
232 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
234 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
235 routers/transports under cutthrough routing.
237 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
238 numbers. Touch up "bool" conditional to keep the same definition.
240 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
242 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
244 JH/12 Expand items in router/transport headers_add or headers_remove lists
245 individually rather than the list as a whole. Bug 1452.
247 Required for reasonable handling of multiple headers_ options when
248 they may be empty; requires that headers_remove items with embedded
249 colons must have them doubled (or the list-separator changed).
251 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
252 view the policy declared in the DMARC record. Currently, $dmarc_status
253 is a combined value of both the record presence and the result of the
256 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
258 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
259 dnslookup router and the smtp transport (applying to the forward
262 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
263 of ldap servers used for a specific lookup. Patch provided by Heiko
266 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
267 New variable $lookup_dnssec_authenticated for observability.
269 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
270 Patch submitted by Lars Timman.
272 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
274 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
275 Requires trusted mode and valid format message id, aborts otherwise.
276 Patch contributed by Heiko Schlichting.
278 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
279 certextract with support for various fields. Bug 1358.
281 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
282 is requested by default, modifiable by smtp transport option
285 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
286 operate on certificate variables to give certificate fingerprints
287 Also new ${sha256:cert_variable}.
289 JH/23 The PRDR feature is moved from being Experimental into the mainline.
291 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
292 Christian Aistleitner.
294 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
296 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
297 file. Patch from Wolfgang Breyha.
299 JH/25 Expand the coverage of the delivery $host and $host_address to
300 client authenticators run in verify callout. Bug 1476.
302 JH/26 Port service names are now accepted for tls_on_connect_ports, to
303 align with daemon_smtp_ports. Bug 72.
305 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
306 support and error reporting did not work properly.
308 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
309 and is readable. Patch from Andrew Colin Kissa.
311 TL/14 Enhance documentation of ${run expansion and how it parses the
312 commandline after expansion, particularly in the case when an
313 unquoted variable expansion results in an empty value.
315 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
317 PP/02 Fix internal collision of T_APL on systems which support RFC3123
318 by renaming away from it. Addresses GH issue 15, reported by
321 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
323 TL/15 SECURITY: prevent double expansion in math comparison functions
324 (can expand unsanitized data). Not remotely exploitable.
331 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
333 PP/02 Make -n do something, by making it not do something.
334 When combined with -bP, the name of an option is not output.
336 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
339 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
340 $sender_host_name and config options to manage this, and basic check
343 PP/05 DSCP support for outbound connections and control modifier for inbound.
345 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
346 (Only plugin which currently uses this is kerberos4, which nobody should
347 be using, but we should make it available and other future plugins might
348 conceivably use it, even though it would break NAT; stuff *should* be
349 using channel bindings instead).
351 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
352 name; added for Sendmail compatibility; requires admin caller.
353 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
354 just ignore it); requires trusted caller.
355 Also parse but ignore: -Ac -Am -X<logfile>
358 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
360 TL/02 Add +smtp_confirmation as a default logging option.
362 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
363 Patch by Magnus Holmgren from 2007-02-20.
365 TL/04 Bugzilla 1281 - Spec typo.
366 Bugzilla 1283 - Spec typo.
367 Bugzilla 1290 - Spec grammar fixes.
369 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
371 TL/06 Add Experimental DMARC support using libopendmarc libraries.
373 TL/07 Fix an out of order global option causing a segfault. Reported to dev
374 mailing list by by Dmitry Isaikin.
376 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
378 JH/02 Support "G" suffix to numbers in ${if comparisons.
380 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
382 NM/01 Bugzilla 1197 - Spec typo
383 Bugzilla 1196 - Spec examples corrections
385 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
387 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
388 gnutls_enable_pkcs11, but renamed to more accurately indicate its
391 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
392 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
394 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
395 "acl {{name}{arg}...}", and optional args on acl condition
398 JH/05 Permit multiple router/transport headers_add/remove lines.
400 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
402 JH/07 Avoid using a waiting database for a single-message-only transport.
403 Performance patch from Paul Fisher. Bugzilla 1262.
405 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
408 JH/09 Add $headers_added variable, with content from use of ACL modifier
409 add_header (but not yet added to the message). Bugzilla 199.
411 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
412 Pulled from Bugzilla 817 by Wolfgang Breyha.
414 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
416 (nb: this is the same fix as in Exim 4.80.1)
418 JH/11 Add A= logging on delivery lines, and a client_set_id option on
421 JH/12 Add optional authenticated_sender logging to A= and a log_selector
424 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
426 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
427 advertise SMTP AUTH mechanism to us, instead of a generic
428 protocol violation error. Also, make Exim more robust to bad
429 data from the Dovecot auth socket.
431 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
433 When a queue runner is handling a message, Exim first routes the
434 recipient addresses, during which it prunes them based on the retry
435 hints database. After that it attempts to deliver the message to
436 any remaining recipients. It then updates the hints database using
439 So if a recipient address works intermittently, it can get repeatedly
440 deferred at routing time. The retry hints record remains fresh so the
441 address never reaches the final cutoff time.
443 This is a fairly common occurrence when a user is bumping up against
444 their storage quota. Exim had some logic in its local delivery code
445 to deal with this. However it did not apply to per-recipient defers
446 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
448 This change adds a proper retry rule check during routing so that the
449 final cutoff time is checked against the message's age. We only do
450 this check if there is an address retry record and there is not a
451 domain retry record; this implies that previous attempts to handle
452 the address had the retry_use_local_parts option turned on. We use
453 this as an approximation for the destination being like a local
454 delivery, as in LMTP.
456 I suspect this new check makes the old local delivery cutoff check
457 redundant, but I have not verified this so I left the code in place.
459 TF/02 Correct gecos expansion when From: is a prefix of the username.
461 Test 0254 submits a message to Exim with the header
465 When I ran the test suite under the user fanf2, Exim expanded
466 the header to contain my full name, whereas it should have added
467 a Resent-Sender: header. It erroneously treats any prefix of the
468 username as equal to the username.
470 This change corrects that bug.
472 GF/01 DCC debug and logging tidyup
473 Error conditions log to paniclog rather than rejectlog.
474 Debug lines prefixed by "DCC: " to remove any ambiguity.
476 TF/03 Avoid unnecessary rebuilds of lookup-related code.
478 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
479 Bug spotted by Jeremy Harris; was flawed since initial commit.
480 Would have resulted in OCSP responses post-SNI triggering an Exim
481 NULL dereference and crash.
483 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
485 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
486 Bug detection, analysis and fix by Samuel Thibault.
487 Bugzilla 1331, Debian bug #698092.
489 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
491 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
492 Server implementation by Todd Lyons, client by JH.
493 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
494 config variable "prdr_enable" controls whether the server
495 advertises the facility. If the client requests PRDR a new
496 acl_data_smtp_prdr ACL is called once for each recipient, after
497 the body content is received and before the acl_smtp_data ACL.
498 The client is controlled by bolth of: a hosts_try_prdr option
499 on the smtp transport, and the server advertisement.
500 Default client logging of deliveries and rejections involving
501 PRDR are flagged with the string "PRDR".
503 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
504 fclose(). Diagnosis by Todd Lyons.
506 PP/17 Update configure.default to handle IPv6 localhost better.
507 Patch by Alain Williams (plus minor tweaks).
510 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
511 This is now consistent with GnuTLS, and is now documented: the
512 previous undocumented portable approach to treating the option as
513 unset was to force an expansion failure. That still works, and
514 an empty string is now equivalent.
516 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
517 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
518 not performing validation itself.
520 PP/20 Added force_command boolean option to pipe transport.
521 Patch from Nick Koston, of cPanel Inc.
523 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
526 TF/04 Added udpsend ACL modifer and hexquote expansion operator
528 PP/21 Fix eximon continuous updating with timestamped log-files.
529 Broken in a format-string cleanup in 4.80, missed when I repaired the
530 other false fix of the same issue.
531 Report and fix from Heiko Schlichting.
534 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
535 Report from Prashanth Katuri.
537 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
538 It's SecureTransport, so affects any MacOS clients which use the
539 system-integrated TLS libraries, including email clients.
541 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
542 using a MIME ACL for non-SMTP local injection.
543 Report and assistance in diagnosis by Warren Baker.
545 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
547 JH/16 Fix comparisons for 64b. Bugzilla 1385.
549 TL/09 Add expansion variable $authenticated_fail_id to keep track of
550 last id that failed so it may be referenced in subsequent ACL's.
552 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
555 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
556 ldap library initialization, allowing self-signed CA's to be
557 used. Also properly sets require_cert option later in code by
558 using NULL (global ldap config) instead of ldap handle (per
559 session). Bug diagnosis and testing by alxgomz.
561 TL/12 Enhanced documentation in the ratelimit.pl script provided in
562 the src/util/ subdirectory.
564 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
565 renamed to Transport Post Delivery Action by Jeremy Harris, as
568 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
569 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
570 redis_servers = needs to be configured which will be used by the redis
571 lookup. Patch from Warren Baker, of The Packet Hub.
573 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
575 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
576 hostname or reverse DNS when processing a host list. Used suggestions
577 from multiple comments on this bug.
579 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
581 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
582 Missed a few lines, added it to make the runtest require no keyboard
585 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
586 contains upper case chars. Make router use caseful_local_part.
588 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
589 support when GnuTLS has been built with p11-kit.
595 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
597 This, or similar/improved, will also be change PP/11 of 4.82.
603 PP/01 Handle short writes when writing local log-files.
604 In practice, only affects FreeBSD (8 onwards).
605 Bugzilla 1053, with thanks to Dmitry Isaikin.
607 NM/01 Bugzilla 949 - Documentation tweak
609 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
612 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
614 PP/02 Implemented gsasl authenticator.
616 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
618 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
619 `pkg-config foo` for cflags/libs.
621 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
622 with rest of GSASL and with heimdal_gssapi.
624 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
625 `pkg-config foo` for cflags/libs for the TLS implementation.
627 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
628 properties get this fed in as external SSF. A number of robustness
629 and debugging improvements to the cyrus_sasl authenticator.
631 PP/08 cyrus_sasl server now expands the server_realm option.
633 PP/09 Bugzilla 1214 - Log authentication information in reject log.
634 Patch by Jeremy Harris.
636 PP/10 Added dbmjz lookup type.
638 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
640 PP/12 MAIL args handles TAB as well as SP, for better interop with
641 non-compliant senders.
642 Analysis and variant patch by Todd Lyons.
644 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
645 Bug report from Lars Müller <lars@samba.org> (via SUSE),
646 Patch from Dirk Mueller <dmueller@suse.com>
648 PP/13 tls_peerdn now print-escaped for spool files.
649 Observed some $tls_peerdn in wild which contained \n, which resulted
650 in spool file corruption.
652 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
653 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
654 or write after TLS renegotiation, which otherwise led to messages
657 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
658 as a tracking header (ie: a signed header comes before the signature).
659 Patch from Wolfgang Breyha.
661 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
662 comma-sep list; embedded commas doubled.
664 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
666 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
668 Report and patch from Dmitry Banschikov.
670 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
671 Removed SSL_clear() after SSL_new() which led to protocol negotiation
672 failures. We appear to now support TLS1.1+ with Exim.
674 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
675 lets Exim select keys and certificates based upon TLS SNI from client.
676 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
677 before an outbound SMTP session. New log_selector, +tls_sni.
679 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
680 NULL dereference. Report and patch from Alun Jones.
682 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
683 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
684 Not seeing resolver debug output on NetBSD, but suspect this is a
685 resolver implementation change.
687 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
688 Left warnings. Added "eximon gdb" invocation mode.
690 PP/21 Defaulting "accept_8bitmime" to true, not false.
692 PP/22 Added -bw for inetd wait mode support.
694 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
695 locate the relevant includes and libraries. Made this the default.
697 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
698 Bugzilla 1246, report and most of solution from Tomasz Kusy.
700 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
701 This may cause build issues on older platforms.
703 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
704 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
705 gnutls_require_mac & gnutls_require_protocols (no longer supported).
706 Added SNI support via GnuTLS too.
707 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
709 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
711 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
712 Applied second patch from Janne, implementing suggestion to default
713 multiple-strings-in-record handling to match SPF spec.
715 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
717 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
718 read-only, out of scope).
719 Patch from Wolfgang Breyha, report from Stuart Northfield.
721 PP/29 Fix three issues highlighted by clang analyser static analysis.
722 Only crash-plausible issue would require the Cambridge-specific
723 iplookup router and a misconfiguration.
724 Report from Marcin Mirosław.
726 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
728 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
729 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
730 As part of this, removing so much warning spew let me fix some minor
731 real issues in debug logging.
733 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
734 assignment on my part. Fixed.
736 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
737 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
738 Janne Snabb (who went above and beyond: thank you).
740 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
741 string otherwise requires a connection and a bunch more work and it's
742 relatively easy to get wrong. Should also expose TLS library linkage
745 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
746 64-bit ${eval} (JH/03).
748 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
749 GNU libc to support some of the 64-bit stuff, should not lead to
750 conflicts. Defined before os.h is pulled in, so if a given platform
751 needs to override this, it can.
753 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
754 protection layer was required, which is not implemented.
755 Bugzilla 1254, patch from Wolfgang Breyha.
757 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
758 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
759 tls_dhparam take prime identifiers. Also unbreak combination of
760 OpenSSL+DH_params+TLSSNI.
762 PP/39 Disable SSLv2 by default in OpenSSL support.
768 PP/01 Solaris build fix for Oracle's LDAP libraries.
769 Bugzilla 1109, patch from Stephen Usher.
771 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
773 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
776 TF/02 Fix a couple more cases where we did not log the error message
777 when unlink() failed. See also change 4.74-TF/03.
779 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
780 lock up or crash if it happened to be inside a call to libc when it
781 got a SIGUSR1 from exiwhat.
783 The SIGUSR1 handler appends the current process status to the process
784 log which is later printed by exiwhat. It used to use the general
785 purpose logging code to do this, but several functions it calls are
786 not safe for signals.
788 The new output code in the SIGUSR1 handler is specific to the process
789 log, and simple enough that it's easy to inspect for signal safety.
790 Removing some special cases also simplifies the general logging code.
791 Removing the spurious timestamps from the process log simplifies
794 TF/04 Improved ratelimit ACL condition.
796 The /noupdate option has been deprecated in favour of /readonly which
797 has clearer semantics. The /leaky, /strict, and /readonly update modes
798 are mutually exclusive. The update mode is no longer included in the
799 database key; it just determines when the database is updated. (This
800 means that when you upgrde Exim will forget old rate measurements.)
802 Exim now checks that the per_* options are used with an update mode that
803 makes sense for the current ACL. For example, when Exim is processing a
804 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
805 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
806 must specify per_mail/readonly. If you omit the update mode it defaults to
807 /leaky where that makes sense (as before) or /readonly where required.
809 The /noupdate option is now undocumented but still supported for
810 backwards compatibility. It is equivalent to /readonly except that in
811 ACLs where /readonly is required you may specify /leaky/noupdate or
812 /strict/noupdate which are treated the same as /readonly.
814 A useful new feature is the /count= option. This is a generalization
815 of the per_byte option, so that you can measure the throughput of other
816 aggregate values. For example, the per_byte option is now equivalent
817 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
819 The per_rcpt option has been generalized using the /count= mechanism
820 (though it's more complicated than the per_byte equivalence). When it is
821 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
822 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
823 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
824 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
825 Note that using per_rcpt with a non-readonly update mode in more than
826 one ACL will cause the recipients to be double-counted. (The per_mail
827 and per_byte options don't have this problem.)
829 The handling of very low rates has changed slightly. If the computed rate
830 is less than the event's count (usually one) then this event is the first
831 after a long gap. In this case the rate is set to the same as this event's
832 count, so that the first message of a spam run is counted properly.
834 The major new feature is a mechanism for counting the rate of unique
835 events. The new per_addr option counts the number of different
836 recipients that someone has sent messages to in the last time period. It
837 behaves like per_rcpt if all the recipient addresses are different, but
838 duplicate recipient addresses do not increase the measured rate. Like
839 the /count= option this is a general mechanism, so the per_addr option
840 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
841 example, measure the rate that a client uses different sender addresses
842 with the options per_mail/unique=$sender_address. There are further
843 details in the main documentation.
845 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
847 TF/06 Removed a few PCRE remnants.
849 TF/07 Automatically extract Exim's version number from tags in the git
850 repository when doing development or release builds.
852 PP/02 Raise smtp_cmd_buffer_size to 16kB.
853 Bugzilla 879. Patch from Paul Fisher.
855 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
856 Heavily based on revision 40f9a89a from Simon Arlott's tree.
859 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
861 PP/05 Variable $av_failed, true if the AV scanner deferred.
862 Bugzilla 1078. Patch from John Horne.
864 PP/06 Stop make process more reliably on build failure.
865 Bugzilla 1087. Patch from Heiko Schlittermann.
867 PP/07 Make maildir_use_size_file an _expandable_ boolean.
868 Bugzilla 1089. Patch from Heiko Schlittermann.
870 PP/08 Handle ${run} returning more data than OS pipe buffer size.
871 Bugzilla 1131. Patch from Holger Weiß.
873 PP/09 Handle IPv6 addresses with SPF.
874 Bugzilla 860. Patch from Wolfgang Breyha.
876 PP/10 GnuTLS: support TLS 1.2 & 1.1.
878 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
881 PP/11 match_* no longer expand right-hand-side by default.
882 New compile-time build option, EXPAND_LISTMATCH_RHS.
883 New expansion conditions, "inlist", "inlisti".
885 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
887 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
889 PP/14 fix log_write() format string regression from TF/03.
890 Bugzilla 1152. Patch from Dmitry Isaikin.
896 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
898 PP/02 Harmonised TLS library version reporting; only show if debugging.
899 Layout now matches that introduced for other libraries in 4.74 PP/03.
901 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
903 PP/04 New "dns_use_edns0" global option.
905 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
908 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
909 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
911 TK/01 Updated PolarSSL code to 0.14.2.
912 Bugzilla 1097. Patch from Andreas Metzler.
914 PP/07 Catch divide-by-zero in ${eval:...}.
917 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
920 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
921 format-string attack -- SECURITY: remote arbitrary code execution.
923 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
924 time unintentionally subject to list matching rules, letting the header
925 cause arbitrary Exim lookups (of items which can occur in lists, *not*
926 arbitrary string expansion). This allowed for information disclosure.
928 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
929 INT_MIN/-1 -- value coerced to INT_MAX.
935 NM/01 Workround for PCRE version dependency in version reporting
938 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
939 This fixes portability to compilers other than gcc, notably
940 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
942 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
943 makefiles for portability to HP-UX and POSIX correctness.
945 PP/01 Permit LOOKUP_foo enabling on the make command-line.
946 Also via indented variable definition in the Makefile.
947 (Debugging by Oliver Heesakkers).
949 PP/02 Restore caching of spamd results with expanded spamd_address.
950 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
952 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
953 Improves build reliability. Fix from: Frank Elsner
955 NM/02 Fix wide character breakage in the rfc2047 coding
956 Fixes bug 1064. Patch from Andrey N. Oktyabrski
958 NM/03 Allow underscore in dnslist lookups
959 Fixes bug 1026. Patch from Graeme Fowler
961 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
962 Code patches from Adam Ciarcinski of NetBSD.
964 NM/04 Fixed exiqgrep to cope with mailq missing size issue
967 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
968 is logged, to avoid truncation. Patch from John Horne.
970 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
971 Patch from Jakob Hirsch.
973 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
974 SQL string expansion failure details.
975 Patch from Andrey Oktyabrski.
977 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
978 Patch from Simon Arlott.
980 PP/09 New lookups functionality failed to compile on old gcc which rejects
981 extern declarations in function scope.
982 Patch from Oliver Fleischmann
984 PP/10 Use sig_atomic_t for flags set from signal handlers.
985 Check getgroups() return and improve debugging.
986 Fixed developed for diagnosis in bug 927 (which turned out to be
989 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
990 Patch from Mark Zealey.
992 PP/12 Bugzilla 1056: Improved spamd server selection.
993 Patch from Mark Zealey.
995 PP/13 Bugzilla 1086: Deal with maildir quota file races.
996 Based on patch from Heiko Schlittermann.
998 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
999 Patch from Uwe Doering, sign-off by Michael Haardt.
1001 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1002 variable declaration deep within a block. Bug and patch from
1005 PP/15 lookups-Makefile IRIX compatibilty coercion.
1007 PP/16 Make DISABLE_DKIM build knob functional.
1009 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1010 Patch by Simon Arlott
1012 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1013 variable argument macros. Our copy now differs from upstream.
1019 TF/01 Failure to get a lock on a hints database can have serious
1020 consequences so log it to the panic log.
1022 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1023 controlled using the smtp_confirmation log selector.
1025 TF/03 Include the error message when we fail to unlink a spool file.
1027 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1028 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1029 for maintaining out-of-tree patches for some time.
1031 PP/01 Bugzilla 139: Documentation and portability issues.
1032 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1033 Handle per-OS dynamic-module compilation flags.
1035 PP/02 Let /dev/null have normal permissions.
1036 The 4.73 fixes were a little too stringent and complained about the
1037 permissions on /dev/null. Exempt it from some checks.
1038 Reported by Andreas M. Kirchwitz.
1040 PP/03 Report version information for many libraries, including
1041 Exim version information for dynamically loaded libraries. Created
1042 version.h, now support a version extension string for distributors
1043 who patch heavily. Dynamic module ABI change.
1045 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1046 privilege escalation vulnerability whereby the Exim run-time user
1047 can cause root to append content of the attacker's choosing to
1050 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1053 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1054 If dropping privileges for untrusted macros, we disabled normal logging
1055 on the basis that it would fail; for the Exim run-time user, this is not
1056 the case, and it resulted in successful deliveries going unlogged.
1057 Fixed. Reported by Andreas Metzler.
1063 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1064 only prepend for the Resent-* case. Fixes regression introduced in
1065 Exim 4.70 by NM/22 for Bugzilla 607.
1067 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1068 increasing numbers of administrators be bitten by this.
1070 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1072 PP/03 Bugzilla 994: added openssl_options main configuration option.
1074 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1076 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1078 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1080 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1081 without a peer certificate, leading to a segfault because of an
1082 assumption that peers always have certificates. Be a little more
1083 paranoid. Problem reported by Martin Tscholak.
1085 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1086 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1087 NB: ClamAV planning to remove STREAM in "middle of 2010".
1088 CL also introduces -bmalware, various -d+acl logging additions and
1089 more caution in buffer sizes.
1091 PP/09 Implemented reverse_ip expansion operator.
1093 PP/10 Bugzilla 937: provide a "debug" ACL control.
1095 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1097 PP/12 Bugzilla 973: Implement --version.
1099 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1101 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1103 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1105 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1106 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1107 ignore trailing whitespace.
1109 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1111 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1114 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1115 Notification from Dr Andrew Aitchison.
1117 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1118 ExtendedDetectionInfo response format.
1119 Notification from John Horne.
1121 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1124 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1125 XSL and documented dependency on system catalogs, with examples of how
1128 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1131 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1132 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1133 configuration file which is writeable by the Exim user or group.
1135 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1136 of configuration files to cover files specified with the -C option if
1137 they are going to be used with root privileges, not just the default
1140 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1141 option (effectively making it always true).
1143 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1144 files to be used while preserving root privileges.
1146 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1147 that rogue child processes cannot use them.
1149 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1150 run-time user, instead of root.
1152 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1153 Exim run-time user without dropping privileges.
1155 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1156 result string, instead of calling string_vformat() twice with the same
1159 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1160 for other users. Others should always drop root privileges if they use
1161 -C on the command line, even for a whitelisted configure file.
1163 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1165 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1171 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1172 $data_path, and $header_path variables; fixed documentation bugs and
1175 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1176 exipick to access non-standard spools, including the "frozen" queue
1179 NM/01 Bugzilla 965: Support mysql stored procedures.
1180 Patch from Alain Williams
1182 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1184 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1185 Patch from Andreas Metzler
1187 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1188 Patch from Kirill Miazine
1190 NM/05 Bugzilla 671: Added umask to procmail example.
1192 JJ/03 installed exipick 20100323.0, fixing doc bug
1194 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1195 directory. Notification and patch from Dan Rosenberg.
1197 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1199 TK/02 Improve log output when DKIM signing operation fails.
1201 MH/01 Treat the transport option dkim_domain as a colon separated
1202 list, not as a single string, and sign the message with each element,
1203 omitting multiple occurences of the same signer.
1205 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1206 Bugzilla 985, 986. Patch by Simon Arlott
1208 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1209 Patch by Simon Arlott
1211 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1212 MBX locking. Notification from Dan Rosenberg.
1218 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1220 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1222 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1224 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1226 NM/04 Bugzilla 915: Flag broken perl installation during build.
1232 TK/01 Added patch by Johannes Berg that expands the main option
1233 "spamd_address" if it starts with a dollar sign.
1235 TK/02 Write list of recipients to X-Envelope-Sender header when building
1236 the mbox-format spool file for content scanning (suggested by Jakob
1239 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1240 (http://www.dcc-servers.net/) support via dccifd. Activated by
1241 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1243 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1244 by Mark Daniel Reidel <mr@df.eu>.
1246 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1247 When building exim an external PCRE library is now needed -
1248 PCRE is a system library on the majority of modern systems.
1249 See entry on PCRE_LIBS in EDITME file.
1251 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1252 conversation. Added nologin parameter to request.
1253 Patch contributed by Kirill Miazine.
1255 TF/01 Do not log submission mode rewrites if they do not change the address.
1257 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1259 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1260 log files in place. Contributed by Roberto Lima.
1262 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1264 TF/03 Bugzilla 615: When checking the local_parts router precondition
1265 after a local_part_suffix or local_part_prefix option, Exim now
1266 does not use the address's named list lookup cache, since this
1267 contains cached lookups for the whole local part.
1269 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1270 Robert Millan. Documentation is in experimental-spec.txt.
1272 TF/04 Bugzilla 668: Fix parallel build (make -j).
1274 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1276 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1277 Patch provided by Jan Srzednicki.
1279 TF/05 Leading white space used to be stripped from $spam_report which
1280 wrecked the formatting. Now it is preserved.
1282 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1283 that they are available at delivery time.
1285 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1287 TF/08 TLS error reporting now respects the incoming_interface and
1288 incoming_port log selectors.
1290 TF/09 Produce a more useful error message if an SMTP transport's hosts
1291 setting expands to an empty string.
1293 NM/06 Bugzilla 744: EXPN did not work under TLS.
1294 Patch provided by Phil Pennock.
1296 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1297 Patch provided by Richard Godbee.
1299 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1300 acl_smtp_notquit, added index entry.
1302 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1303 Patch provided by Eugene Bujak.
1305 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1306 accept(). Patch provided by Maxim Dounin.
1308 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1309 Patch provided by Phil Pennock.
1311 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1313 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1314 Patch provided by Brad "anomie" Jorsch.
1316 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1317 Patch provided by Dean Brooks.
1319 TK/05 Add native DKIM support (does not depend on external libraries).
1321 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1322 Patch provided by Graeme Fowler.
1324 NM/16 Bugzilla 851: Documentation example syntax fix.
1326 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1328 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1331 NM/19 Bugzilla 745: TLS version reporting.
1332 Patch provided by Phil Pennock.
1334 NM/20 Bugzilla 167: bool: condition support.
1335 Patch provided by Phil Pennock.
1337 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1338 clients. Patch provided by Phil Pennock.
1340 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1341 Patch provided by Brad "anomie" Jorsch.
1343 NM/23 Bugzilla 687: Fix misparses in eximstats.
1344 Patch provided by Heiko Schlittermann.
1346 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1347 Patch provided by Heiko Schlittermann.
1349 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1350 plus update to original patch.
1352 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1354 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1355 Patch provided by David Brownlee.
1357 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1359 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1361 NM/30 Bugzilla 888: TLS documentation bugfixes.
1363 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1365 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1366 Unlike the original bugzilla I have changed all shell scripts in src tree.
1368 NM/33 Bugzilla 898: Transport filter timeout fix.
1369 Patch by Todd Rinaldo.
1371 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1372 Patch by Serge Demonchaux.
1374 NM/35 Bugzilla 39: Base64 decode bug fixes.
1375 Patch by Jakob Hirsch.
1377 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1379 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1381 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1383 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1389 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1390 ALT-N's libdkim that I have put here:
1391 http://duncanthrax.net/exim-experimental/
1393 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1394 were called 'true' and it seems that C99 defines that as a reserved
1395 keyword to be used with 'bool' variable types. That means you could
1396 not include C99-style headers which use bools without triggering
1397 build errors in sieve.c.
1399 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1400 as mailq or other aliases. Changed the --help handling significantly
1401 to do whats expected. exim_usage() emits usage/help information.
1403 SC/01 Added the -bylocaldomain option to eximstats.
1405 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1407 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1409 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1415 PH/01 Another patch from the Sieve maintainer.
1417 PH/02 When an IPv6 address is converted to a string for single-key lookup
1418 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1419 dots are used instead of colons so that keys in lsearch files need not
1420 contain colons. This was done some time before quoting was made available
1421 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1422 (notated using the quote facility) so as to distinguish them from IPv4
1423 keys. This meant that lookups for IP addresses in host lists did not work
1424 for iplsearch lookups.
1426 This has been fixed by arranging for IPv6 addresses to be expressed with
1427 colons if the lookup type is iplsearch. This is not incompatible, because
1428 previously such lookups could never work.
1430 The situation is now rather anomolous, since one *can* have colons in
1431 ordinary lsearch keys. However, making the change in all cases is
1432 incompatible and would probably break a number of configurations.
1434 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1437 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1438 conversion specification without a maximum field width, thereby enabling
1439 a rogue spamd server to cause a buffer overflow. While nobody in their
1440 right mind would setup Exim to query an untrusted spamd server, an
1441 attacker that gains access to a server running spamd could potentially
1442 exploit this vulnerability to run arbitrary code as the Exim user.
1444 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1445 $primary_hostname instead of what libspf2 thinks the hosts name is.
1447 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1448 a directory entry by the name of the lookup key. Previously, if a
1449 symlink pointed to a non-existing file or a file in a directory that
1450 Exim lacked permissions to read, a lookup for a key matching that
1451 symlink would fail. Now it is enough that a matching directory entry
1452 exists, symlink or not. (Bugzilla 503.)
1454 PH/03 The body_linecount and body_zerocount variables are now exported in the
1457 PH/04 Added the $dnslist_matched variable.
1459 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1460 This means they are set thereafter only if the connection becomes
1463 PH/06 Added the client_condition to authenticators so that some can be skipped
1464 by clients under certain conditions.
1466 PH/07 The error message for a badly-placed control=no_multiline_responses left
1467 "_responses" off the end of the name.
1469 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1471 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1472 (without spaces) instead of just copying the configuration text.
1474 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1476 PH/11 Added $max_received_linelength.
1478 PH/12 Added +ignore_defer and +include_defer to host lists.
1480 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1481 way in which PCRE > 7.0 is built.
1483 PH/14 Implemented queue_only_load_latch.
1485 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1486 MAIL command. The effect was to mangle the value on 64-bit systems.
1488 PH/16 Another patch from the Sieve maintainer.
1490 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1492 PH/18 If a system quota error occurred while trying to create the file for
1493 a maildir delivery, the message "Mailbox is full" was not appended to the
1494 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1495 only to a quota excession during the actual writing of the file.
1497 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1498 characters?) which causes problems in log lines. The DN values are now
1499 passed through string_printing() before being added to log lines.
1501 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1502 and InterBase are left for another time.)
1504 PH/21 Added message_body_newlines option.
1506 PH/22 Guard against possible overflow in moan_check_errorcopy().
1508 PH/23 POSIX allows open() to be a macro; guard against that.
1510 PH/24 If the recipient of an error message contained an @ in the local part
1511 (suitably quoted, of course), incorrect values were put in $domain and
1512 $local_part during the evaluation of errors_copy.
1518 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1519 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1522 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1523 issue a MAIL command.
1525 PH/02 In an ACL statement such as
1527 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1529 if a client was not listed at all, or was listed with a value other than
1530 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1531 the condition was not true (as it should be), so access was not denied.
1532 The bug was that the ! inversion was incorrectly passed on to the second
1533 item. This has been fixed.
1535 PH/03 Added additional dnslists conditions == and =& which are different from
1536 = and & when the dns lookup returns more than one IP address.
1538 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1539 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1541 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1542 FSYNC, which compiles an option called disable_fsync that allows for
1543 bypassing fsync(). The documentation is heavily laced with warnings.
1545 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1547 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1548 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1549 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1550 including adding "make clean"; (3) Added -fPIC when compiling the test
1551 dynamically loaded module, to get rid of a warning.
1553 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1554 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1555 = 0s. The bug is otherwise harmless.
1557 PH/07 There was a bug in the dovecot authenticator such that the value of
1558 $auth1 could be overwritten, and so not correctly preserved, after a
1559 successful authentication. This usually meant that the value preserved by
1560 the server_setid option was incorrect.
1562 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1564 PH/09 Installed PCRE release 7.0.
1566 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1567 run for batched SMTP input. It is now run at the start of every message
1568 in the batch. While fixing this I discovered that the process information
1569 (output by running exiwhat) was not always getting set for -bs and -bS
1570 input. This is fixed, and it now also says "batched" for BSMTP.
1572 PH/11 Added control=no_pipelining.
1574 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1575 patch, slightly modified), and move the expansion of helo_data till after
1576 the connection is made in the smtp transport (so it can use these
1579 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1581 PH/14 Added log_selector = +pid.
1583 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1585 PH/16 Add ${if forany and ${if forall.
1587 PH/17 Added dsn_from option to vary the From: line in DSNs.
1589 PH/18 Flush SMTP output before performing a callout, unless control =
1590 no_callout_flush is set.
1592 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1593 was true (the default) a successful delivery failed to delete the retry
1594 item, thus causing premature timeout of the address. The bug is now
1597 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1599 PH/21 Long custom messages for fakedefer and fakereject are now split up
1600 into multiline reponses in the same way that messages for "deny" and
1601 other ACL rejections are.
1603 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1604 with slight modification.
1606 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1607 draft, changing the syntax and factoring some duplicate code".
1609 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1610 for deliveries of the second and subsequent messages over the same SMTP
1613 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1614 ${reduce, with only minor "tidies".
1616 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1618 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1619 expansion side effects.
1621 PH/27 When a message times out after an over-quota error from an Exim-imposed
1622 quota, the bounce message says "mailbox is full". This message was not
1623 being given when it was a system quota that was exceeded. It now should
1626 MH/03 Made $recipients available in local_scan(). local_scan() already has
1627 better access to the recipient list through recipients_list[], but
1628 $recipients can be useful in postmaster-provided expansion strings.
1630 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1631 in the case of a MAIL command with additional options following the
1632 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1633 were accidentally chopped off.
1635 PH/29 SMTP synchronization checks are implemented when a command is read -
1636 there is a check that no more input is waiting when there shouldn't be
1637 any. However, for some commands, a delay in an ACL can mean that it is
1638 some time before the response is written. In this time, more input might
1639 arrive, invalidly. So now there are extra checks after an ACL has run for
1640 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1641 pipelining has not been advertised.
1643 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1645 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1646 returned in the SMTP response when smtp_return_error_details was false.
1647 This has been fixed.
1649 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1650 instead of the C I/O that was originally supplied, because problems were
1651 reported on Solaris.
1653 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1654 Exim which did not show up earlier: it was assuming that a call to
1655 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1656 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1657 was a macro that became an assignment, so it seemed to work. This has
1658 changed to a proper function call with a void return, hence the compile
1659 error. Exim's code has been fixed.
1661 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1664 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1666 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1668 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1669 because it turns out to be too big or there is a timeout) there is no
1670 "Completed" line in the log. When some messages of this type were
1671 selected by exigrep, they were listed as "not completed". Others were
1672 picked up by some special patterns. I have improved the selection
1673 criteria to be more general.
1675 PH/38 The host_find_failed option in the manualroute router can now be set
1676 to "ignore", to completely ignore a host whose IP address cannot be
1677 found. If all hosts are ignored, the behaviour is controlled by the new
1678 host_all_ignored option.
1680 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1681 homing or because of multiple MX records with /mx) generated more than
1682 one IP address, and the following item turned out to be the local host,
1683 all the secondary addresses of the first item were incorrectly removed
1684 from the list, along with the local host and any following hosts (which
1685 is what is supposed to happen).
1687 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1688 whoever called Exim into the -H file. In the case of the daemon it was
1689 behaving confusingly. When first started, it used values for whoever
1690 started the daemon, but after a SIGHUP it used the Exim user (because it
1691 calls itself on a restart). I have changed the code so that it now always
1694 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1695 message are rejected with the same error (e.g. no authentication or bad
1696 sender address), and a DATA command is nevertheless sent (as can happen
1697 with PIPELINING or a stupid MUA), the error message that was given to the
1698 RCPT commands is included in the rejection of the DATA command. This is
1699 intended to be helpful for MUAs that show only the final error to their
1702 PH/42 Another patch from the Sieve maintainer.
1704 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1705 Eximstats - Fixed some broken HTML links and added missing column headers
1707 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1708 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1710 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1712 PH/43 Yet another patch from the Sieve maintainer.
1714 PH/44 I found a way to check for a TCP/IP connection going away before sending
1715 the response to the final '.' that terminates a message, but only in the
1716 case where the client has not sent further data following the '.'
1717 (unfortunately, this is allowed). However, in many cases there won't be
1718 any further data because there won't be any more messages to send. A call
1719 to select() can be used: if it shows that the input is "ready", there is
1720 either input waiting, or the socket has been closed. An attempt to read
1721 the next input character can distinguish the two cases. Previously, Exim
1722 would have sent an OK response which the client would never have see.
1723 This could lead to message repetition. This fix should cure that, at
1724 least in a lot of common cases.
1726 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1727 advertised in response to EHLO.
1733 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1734 fixed by 4.65/MH/01 (is this a record?) are fixed:
1736 (i) An empty string was always treated as zero by the numeric comparison
1737 operators. This behaviour has been restored.
1739 (ii) It is documented that the numeric comparison operators always treat
1740 their arguments as decimal numbers. This was broken in that numbers
1741 starting with 0 were being interpreted as octal.
1743 While fixing these problems I realized that there was another issue that
1744 hadn't been noticed. Values of message_size_limit (both the global option
1745 and the transport option) were treated as octal if they started with 0.
1746 The documentation was vague. These values are now always treated as
1747 decimal, and I will make that clear in the documentation.
1753 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1754 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1757 MH/01 Don't check that the operands of numeric comparison operators are
1758 integers when their expansion is in "skipping" mode (fixes bug
1759 introduced by 4.64-PH/07).
1761 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1762 child addresses, Exim now panics and dies. Previously, because the count
1763 is held in a short int, deliveries were likely to be lost. As such a
1764 large number of recipients for a single message is ridiculous
1765 (performance will be very, very poor), I have chosen to impose a limit
1766 rather than extend the field.
1772 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1773 leftover -K file (the existence of which was triggered by #402).
1774 While we were at it, introduced process PID as part of the -K
1775 filename. This should rule out race conditions when creating
1778 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1779 processing considerably. Previous code took too long for large mails,
1780 triggering a timeout which in turn triggers #401.
1782 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1783 in the DK code in transports.c. sendfile() is not really portable,
1784 hence the _LINUX specificness.
1786 TF/01 In the add_headers option to the mail command in an Exim filter,
1787 there was a bug that Exim would claim a syntax error in any
1788 header after the first one which had an odd number of characters
1791 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1792 callout verification, Exim cached a "reject" for the entire domain. This
1793 is correct for most verifications, but it is not correct for a recipient
1794 verification with use_sender or use_postmaster set, because in that case
1795 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1796 case of MAIL FROM:<> rejection from other early rejections (e.g.
1797 rejection of HELO). When verifying a recipient using a non-null MAIL
1798 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1799 Whatever the result of the callout, the value of the domain cache is
1800 left unchanged (for any other kind of callout, getting as far as trying
1801 RCPT means that the domain itself is ok).
1803 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1806 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1807 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1808 session, the daemon ignores ECONNECT errors and logs others; it now
1809 ignores EPIPE as well.
1811 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1812 (quoted-printable decoding).
1814 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1815 later the small subsequent patch to fix an introduced bug.
1817 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1819 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1821 PH/08 An error is now given if message_size_limit is specified negative.
1823 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1824 to be given (somewhat) arbitrary names.
1826 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1829 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1830 miscellaneous code fixes
1832 PH/10 Added the log_reject_target ACL modifier to specify where to log
1835 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1836 hostname. This is wrong, because it relates to the incoming message (and
1837 probably the interface on which it is arriving) and not to the outgoing
1838 callout (which could be using a different interface). This has been
1839 changed to use the value of the helo_data option from the smtp transport
1840 instead - this is what is used when a message is actually being sent. If
1841 there is no remote transport (possible with a router that sets up host
1842 addresses), $smtp_active_hostname is used.
1844 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1845 tweaks were necessary in order to get it to work (see also 21 below):
1846 (a) The code assumed that strncpy() returns a negative number on buffer
1847 overflow, which isn't the case. Replaced with Exim's string_format()
1849 (b) There were several signed/unsigned issues. I just did the minimum
1850 hacking in of casts. There is scope for a larger refactoring.
1851 (c) The code used strcasecmp() which is not a standard C function.
1852 Replaced with Exim's strcmpic() function.
1853 (d) The code set only $1; it now sets $auth1 as well.
1854 (e) A simple test gave the error "authentication client didn't specify
1855 service in request". It would seem that Dovecot has changed its
1856 interface. Fortunately there's a specification; I followed it and
1857 changed what the client sends and it appears to be working now.
1859 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1862 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1863 address A is aliased to B and C, where B exists and C does not. Without
1864 -v the output is "A verified" because verification stops after a
1865 successful redirection if more than one address is generated. However,
1866 with -v the child addresses are also verified. Exim was outputting "A
1867 failed to verify" and then showing the successful verification for C,
1868 with its parentage. It now outputs "B failed to verify", showing B's
1869 parentage before showing the successful verification of C.
1871 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1872 look up a TXT record in a specific list after matching in a combined
1875 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1876 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1877 they consult the DNS. I had assumed they would set it the way they
1878 wanted; and indeed my experiments on Linux seem to show that in some
1879 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1880 To be on the safe side, however, I have now made the interface to
1881 host_find_byname() similar to host_find_bydns(), with an argument
1882 containing the DNS resolver options. The host_find_byname() function now
1883 sets these options at its start, just as host_find_bydns() does. The smtp
1884 transport options dns_qualify_single and dns_search_parents are passed to
1885 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1886 of host_find_byname() use the default settings of RES_DEFNAMES
1887 (qualify_single) but not RES_DNSRCH (search_parents).
1889 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1890 spool_read_header() do less string testing, by means of a preliminary
1891 switch on the second character of optional "-foo" lines. (This is
1892 overdue, caused by the large number of possibilities that now exist.
1893 Originally there were few.) While I was there, I also converted the
1894 str(n)cmp tests so they don't re-test the leading "-" and the first
1895 character, in the hope this might squeeze out yet more improvement.
1897 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1898 flag allowing group syntax was set by the header_syntax check but not
1899 turned off, possible causing trouble later; (2) The flag was not being
1900 set at all for the header_verify test, causing "group"-style headers to
1901 be rejected. I have now set it in this case, and also caused header_
1902 verify to ignore an empty address taken from a group. While doing this, I
1903 came across some other cases where the code for allowing group syntax
1904 while scanning a header line wasn't quite right (mostly, not resetting
1905 the flag correctly in the right place). These bugs could have caused
1906 trouble for malformed header lines. I hope it is now all correct.
1908 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1909 with the "reply" argument non-NULL. The code, however (which originally
1910 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1911 but it didn't always do it. This confused somebody who was copying the
1912 code for some other use. I have removed all the tests.
1914 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1915 feature that was used to support insecure browsers during the U.S. crypto
1916 embargo. It requires special client support, and Exim is probably the
1917 only MTA that supported it -- and would never use it because real RSA is
1918 always available. This code has been removed, because it had the bad
1919 effect of slowing Exim down by computing (never used) parameters for the
1920 RSA_EXPORT functionality.
1922 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1923 authenticator to fail if there's a tab character in the incoming data
1924 (there should never be unless someone is messing about, as it's supposed
1925 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1926 if the connection is using TLS or if the remote IP is the same as the
1927 local IP, and the "valid-client-cert option" if a client certificate has
1930 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1931 authenticators. This can be used for authorization after authentication
1932 succeeds. (In the case of plaintext, it servers for both authentication
1935 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1936 if any retry times were supplied.
1938 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1939 connection, where there is no client IP address to check. In this
1940 situation, the verify now always succeeds.
1942 PH/25 Applied John Jetmore's -Mset patch.
1944 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1946 PH/27 In a string expansion for a processed (not raw) header when multiple
1947 headers of the same name were present, leading whitespace was being
1948 removed from all of them, but trailing whitespace was being removed only
1949 from the last one. Now trailing whitespace is removed from each header
1950 before concatenation. Completely empty headers in a concatenation (as
1951 before) are ignored.
1953 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1954 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1956 PH/29 [Removed. This was a change that I later backed out, and forgot to
1957 correct the ChangeLog entry (that I had efficiently created) before
1958 committing the later change.]
1960 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1961 address errors (4xx response to RCPT) over the same connection as other
1962 messages routed to the same hosts. Such deliveries are always "forced",
1963 so retry times are not inspected. This resulted in far too many retries
1964 for the affected addresses. The effect occurred only when there were more
1965 hosts than the hosts_max_try setting in the smtp transport when it had
1966 the 4xx errors. Those hosts that it had tried were not added to the list
1967 of hosts for which the message was waiting, so if all were tried, there
1968 was no problem. Two fixes have been applied:
1970 (i) If there are any address or message errors in an SMTP delivery, none
1971 of the hosts (tried or untried) are now added to the list of hosts
1972 for which the message is waiting, so the message should not be a
1973 candidate for sending over the same connection that was used for a
1974 successful delivery of some other message. This seems entirely
1975 reasonable: after all the message is NOT "waiting for some host".
1976 This is so "obvious" that I'm not sure why it wasn't done
1977 previously. Hope I haven't missed anything, but it can't do any
1978 harm, as the worst effect is to miss an optimization.
1980 (ii) If, despite (i), such a delivery is accidentally attempted, the
1981 routing retry time is respected, so at least it doesn't keep
1982 hammering the server.
1984 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1985 in ${readsocket because some servers need this prod.
1987 PH/32 Added some extra debug output when updating a wait-xxx database.
1989 PH/33 The hint "could be header name not terminated by colon", which has been
1990 given for certain expansion errors for a long time, was not being given
1991 for the ${if def:h_colon_omitted{... case.
1993 PH/34 The spec says: "With one important exception, whenever a domain list is
1994 being scanned, $domain contains the subject domain." There was at least
1995 one case where this was not true.
1997 PH/35 The error "getsockname() failed: connection reset by peer" was being
1998 written to the panic log as well as the main log, but it isn't really
1999 panic-worthy as it just means the connection died rather early on. I have
2000 removed the panic log writing for the ECONNRESET error when getsockname()
2003 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2004 runs only) independently of the message's sender address. This meant
2005 that, if the 4xx error was in fact related to the sender, a different
2006 message to the same recipient with a different sender could confuse
2007 things. In particualar, this can happen when sending to a greylisting
2008 server, but other circumstances could also provoke similar problems.
2009 I have changed the default so that the retry time for these errors is now
2010 based a combination of the sender and recipient addresses. This change
2011 can be overridden by setting address_retry_include_sender=false in the
2014 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2015 remote server are returned as part of bounce messages. This was not
2016 happening for LMTP over a pipe (the lmtp transport), but now it is the
2017 same for both kinds of LMTP.
2019 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2020 in header lines that were in fact CNAMEs. This is no longer the case.
2022 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2023 and queue runs started by the daemon processed all messages. This has
2024 been fixed so that -R and -S can now usefully be given with -q<time>.
2026 PH/40 Import PCRE release 6.7 (fixes some bugs).
2028 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2030 PH/42 Give an error if -q is specified more than once.
2032 PH/43 Renamed the variables $interface_address and $interface_port as
2033 $received_ip_address and $received_port, to make it clear that these
2034 values apply to message reception, and not to the outgoing interface when
2035 a message is delivered. (The old names remain recognized, of course.)
2037 PH/44 There was no timeout on the connect() call when using a Unix domain
2038 socket in the ${readsocket expansion. There now is.
2040 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2041 be meaningful with "accept".
2043 SC/01 Eximstats V1.43
2044 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2046 SC/02 Eximstats V1.44
2047 Use a glob alias rather than an array ref in the generated
2048 parser. This improves both readability and performance.
2050 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2051 Collect SpamAssassin and rejection statistics.
2052 Don't display local sender or destination tables unless
2053 there is data to show.
2054 Added average volumes into the top table text output.
2056 SC/04 Eximstats V1.46
2057 Collect data on the number of addresses (recipients)
2058 as well as the number of messages.
2060 SC/05 Eximstats V1.47
2061 Added 'Message too big' to the list of mail rejection
2062 reasons (thanks to Marco Gaiarin).
2064 SC/06 Eximstats V1.48
2065 Mainlog lines which have GMT offsets and are too short to
2066 have a flag are now skipped.
2068 SC/07 Eximstats V1.49 (Alain Williams)
2069 Added the -emptyok flag.
2071 SC/08 Eximstats V1.50
2072 Fixes for obtaining the IP address from reject messages.
2074 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2075 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2076 whitesspace changes from 4.64-PH/27
2078 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2081 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2082 are found, allow negative numbers in numeric criteria)
2084 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2086 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2087 to match changes made in 4.64-PH/43
2089 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2091 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2092 "domain" is actually an IP address, and force a failure. This locks out
2093 those revolvers/nameservers that support "A-for-A" lookups, in
2094 contravention of the specifications.
2096 PH/48 When a host name was looked up from an IP address, and the subsequent
2097 forward lookup of the name timed out, the host name was left in
2098 $sender_host_name, contrary to the specification.
2100 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2101 restricted to single-key lookups, Exim was not diagnosing an error if
2102 * or *@ was used with a query-style lookup.
2104 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2106 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2107 long ago, but noone interested enough thought of it. Let's just say that
2108 the "1.1" means that there are some new functions that weren't there at
2109 some point in the past.
2111 PH/51 Error processing for expansion failure of helo_data from an smtp
2112 transport during callout processing was broken.
2114 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2115 tested/used via the -bh/-bhc/-bs options.
2117 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2118 bug, fixed in subsequent PCRE releases).
2120 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2121 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2123 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2129 SC/01 Use a glob alias rather than an array ref in eximstats generated
2130 parser. This improves both readability and performance.
2132 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2133 Don't display local sender or destination tables in eximstats unless
2134 there is data to show.
2135 Added average volumes into the eximstats top table text output.
2137 SC/03 Collect data on the number of addresses (recipients) as well
2138 as the number of messages in eximstats.
2140 TF/01 Correct an error in the documentation for the redirect router. Exim
2141 does (usually) call initgroups() when daemonizing.
2143 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2144 with consistent privilege compared to when running as a daemon.
2146 TF/03 Note in the spec that $authenticated_id is not set for local
2147 submissions from trusted users.
2149 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2150 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2152 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2153 by adding some example configuration directives to the default
2154 configuration file. A little bit of work is required to uncomment the
2155 directives and define how usernames and passwords are checked, but
2156 there is now a framework to start from.
2158 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2159 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2160 without this. I don't know how relevant this is to other LDAP libraries.
2162 PH/02 Add the verb name to the "unknown ACL verb" error.
2164 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2166 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2168 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2169 directory not expanded when it should be if an expanded home directory
2170 was set for the address (which is overridden by the transport).
2172 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2175 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2176 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2177 because it is too late at that time, and has no effect.
2179 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2180 security issue with \' (bugzilla #107). I could not use the
2181 PQescapeStringConn() function, because it needs a PGconn value as one of
2184 PH/08 When testing addresses using -bt, indicate those final addresses that
2185 are duplicates that would not cause an additional delivery. At least one
2186 person was confused, thinking that -bt output corresponded to deliveries.
2187 (Suppressing duplicates isn't a good idea as you lose the information
2188 about possibly different redirections that led to the duplicates.)
2190 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2191 systems where poll() doesn't work, in particular OS X.
2193 PH/10 Added more information to debugging output for retry time not reached.
2195 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2196 operations in malware.c.
2198 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2201 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2202 syslog_duplication set false, log lines that would normally be written
2203 both the the main log and to the reject log were not written to syslog at
2206 PH/14 In the default configuration, change the use of "message" in ACL warn
2207 statements to "add_header".
2209 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2210 not followed by a command (e.g. "seen endif").
2212 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2213 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2216 PH/17 Added extra conditions to the default value of delay_warning_condition
2220 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2221 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2222 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2225 The Auto-Submitted: and various List- headers are standardised, whereas I
2226 don't think Precedence: ever was.
2228 PH/18 Refactored debugging code in route_finduser() to show more information,
2229 in particular, the error code if getpwnam() issues one.
2231 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2232 This is apparently needed in addition to the PH/07 change above to avoid
2233 any possible encoding problems.
2235 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2236 but not after initializing Perl.
2238 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2239 output them only if debugging. By default they are written stderr,
2240 apparently, which is not desirable.
2242 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2245 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2248 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2250 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2251 authenticated or an ident call has been made. Suppress the default
2252 values for $authenticated_id and $authenticated_sender (but permit -oMai
2253 and -oMas) when testing with -bh.
2255 PH/24 Re-jigged the order of the tests in the default configuration so that the
2256 tests for valid domains and recipients precede the DNS black list and CSA
2257 tests, on the grounds that those ones are more expensive.
2259 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2260 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2261 This bug exists in every version of Exim that I still have, right back to
2264 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2265 However, an attempt to turn on case-sensitivity in a regex key by
2266 including (?-i) didn't work because the subject string was already
2267 lowercased, and the effects were non-intuitive. It turns out that a
2268 one-line patch can be used to allow (?-i) to work as expected.
2274 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2275 other effects) broke the use of negated acl sub-conditions.
2277 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2280 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2281 "Deny" causes Exim to reject the incoming connection with a 554 error.
2282 Unfortunately, if there is a major crisis, such as a disk failure,
2283 tcp-wrappers gives "deny", whereas what one would like would be some
2284 kind of temporary error. A kludge has been added to help with this.
2285 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2286 554 error is used if errno is still zero or contains ENOENT (which occurs
2287 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2290 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2292 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2293 errors. Otherwise a message that provokes a temporary error (when other
2294 messages do not) can cause a whole host to time out.
2296 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2297 addresses were routed directly to files or pipes from a redirect router.
2298 File deliveries just didn't batch; pipe deliveries might have suffered
2301 PH/06 A failure to get a lock for a hints database would erroneously always say
2302 "Failed to get write lock", even when it was really a read lock.
2304 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2305 of 0600. This has been changed to use the value of the lockfile_mode
2306 option (which defaults to 0600).
2308 PH/08 Applied small patch from the Sieve maintainer.
2310 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2311 folder from quota calculations, a direct delivery into this folder messed
2312 up the contents of the maildirsize file. This was because the regex was
2313 used only to exclude .Trash (or whatever) when the size of the mailbox
2314 was calculated. There was no check that a delivery was happening into an
2315 excluded directory. This bug has been fixed by ignoring all quota
2316 processing for deliveries into excluded directories.
2318 PH/10 Added the maildirfolder_create_regex option to appendfile.
2324 PH/01 The code for finding all the local interface addresses on a FreeBSD
2325 system running IPv6 was broken. This may well have applied to all BSD
2326 systems, as well as to others that have similar system calls. The broken
2327 code found IPv4 interfaces correctly, but gave incorrect values for the
2328 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2329 that it would not match correctly against @[] and not recognize the IPv6
2332 PH/02 The ipliteral router was not recognizing addresses of the form user@
2333 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2335 PH/03 Added disable_ipv6.
2337 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2338 decoded form, because it is most often used to construct To: headers
2339 lines in autoreplies, and the decoded form may well be syntactically
2340 invalid. However, $reply_address has leading white space removed, and all
2341 newlines turned into spaces so that the autoreply transport does not
2344 PH/05 If group was specified without a user on a router, and no group or user
2345 was specified on a transport, the group from the router was ignored.
2347 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2348 for visible compile-time settings that can be used to change these
2349 numbers, for those that want even more. Backwards compatibility with old
2350 spool files has been maintained. However, going back to a previous Exim
2351 release will lost any variables that are in spool files.
2353 PH/07 Two small changes when running in the test harness: increase delay when
2354 passing a TCP/IP connection to a new process, in case the original
2355 process has to generate a bounce, and remove special handling of
2356 127.0.0.2 (sic), which is no longer necessary.
2358 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2359 be the same on different OS.
2361 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2364 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2365 whether --show-vars was specified or not
2367 JJ/02 exipick: Added support for new ACL variable spool format introduced
2370 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2371 syntactically invalid From: or Reply-to: line, and a filter used this to
2372 generate an autoreply, and therefore failed to obtain an address for the
2373 autoreply, Exim could try to deliver to a non-existent relative file
2374 name, causing unrelated and misleading errors. What now happens is that
2375 it logs this as a hard delivery error, but does not attempt to create a
2378 PH/11 The exinext utility has a -C option for testing purposes, but although
2379 the given file was scanned by exinext itself; it wasn't being passed on
2380 when Exim was called.
2382 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2383 an end-of-file indication when reading a command response.
2385 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2386 compiled. In many other places in Exim, IPv6 addresses are always
2387 recognized, so I have changed this. It also means that IPv4 domain
2388 literals of the form [IPV4:n.n.n.n] are now always recognized.
2390 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2391 used if the router is not running as root, for example, when verifying at
2392 ACL time, or when using -bh. The debugging output from this situation was
2393 non-existent - all you got was a failure to exec. I have made two
2396 (a) Failures to set uid/gid, the current directory, or a process leader
2397 in a subprocess such as that created by queryprogram now generate
2398 suitable debugging ouput when -d is set.
2400 (b) The queryprogram router detects when it is not running as root,
2401 outputs suitable debugging information if -d is set, and then runs
2402 the subprocess without attempting to change uid/gid.
2404 PH/15 Minor change to Makefile for building test_host (undocumented testing
2407 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2408 additional section of a DNS packet that returns MX or SRV records.
2409 Instead, it always explicitly searches for A/AAAA records. This avoids
2410 major problems that occur when a DNS server includes only records of one
2411 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2412 fixed another bug: if SRV records were looked up and the corresponding
2413 address records were *not* found in the additional section, the port
2414 values from the SRV records were lost.
2416 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2417 using the correct key (the original address) when searching the retry
2418 rules in order to find which one to use for generating the retry hint.
2420 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2421 adding the default one. Similarly, if it contains a Reply-To: header, the
2422 errors_reply_to option, if set, is not used.
2424 PH/19 When calculating a retry time, Exim used to measure the "time since
2425 failure" by looking at the "first failed" field in the retry record. Now
2426 it does not use this if it is later than than the arrival time of the
2427 message. Instead it uses the arrival time. This makes for better
2428 behaviour in cases where some deliveries succeed, thus re-setting the
2429 "first failed" field. An example is a quota failure for a huge message
2430 when small messages continue to be delivered. Without this change, the
2431 "time since failure" will always be short, possible causing more frequent
2432 delivery attempts for the huge message than are intended.
2433 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2435 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2436 $1, $2, $3) because the numerical variables can be reset during some
2437 expansion items (e.g. "match"), thereby losing the authentication data.
2439 PH/21 Make -bV show the size of off_t variables so that the test suite can
2440 decide whether to run tests for quotas > 2G.
2442 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2443 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2444 filecount value is greater than 2G or if a quota value is greater than 2G
2445 on a system where the size of off_t is not greater than 4, a panic error
2448 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2449 never match. The debug and -bh output now contains an explicit error
2450 message indicating a malformed IPv4 address or mask.
2452 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2453 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2454 PH/23 above applies.
2456 PH/25 Do not write to syslog when running in the test harness. The only
2457 occasion when this arises is a failure to open the main or panic logs
2458 (for which there is an explicit test).
2460 PH/26 Added the /no_tell option to "control=freeze".
2462 PH/27 If a host name lookup failed very early in a connection, for example, if
2463 the IP address matched host_lookup and the reverse lookup yielded a name
2464 that did not have a forward lookup, an error message of the form "no IP
2465 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2466 could be logged. Now it outputs the IP address instead of "NULL".
2468 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2469 allows the sender and the authenticated sender to be set when
2470 submitting a message from within Exim. Since child_open_exim() is
2471 documented for local_scan(), the new function should be too.
2473 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2474 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2475 results in an empty string is now treated as unset.
2477 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2479 PH/31 Added sender_verify_fail logging option.
2481 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2482 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2483 tidied the source and removed it altogether.
2485 PH/33 When a queue run was abandoned because the load average was too high, a
2486 log line was always written; now it is written only if the queue_run log
2487 selector is set. In addition, the log line for abandonment now contains
2488 information about the queue run such as the pid. This is always present
2489 in "start" and "stop" lines but was omitted from the "abandon" line.
2491 PH/34 Omit spaces between a header name and the colon in the error message that
2492 is given when verify = headers_syntax fails (if there are lots of them,
2493 the message gets confusing).
2495 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2496 names, as there are now some PTR records that contain slashes. This check
2497 is only to protect against broken name servers that fall over on strange
2498 characters, so the fact that it applies to all lookups doesn't matter.
2500 PH/36 Now that the new test suite is complete, we can remove some of the
2501 special code in Exim that was needed for the old test suite. For example,
2502 sorting DNS records because real resolvers return them in an arbitrary
2503 order. The new test suite's fake resolver always returns records in the
2506 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2507 bounces) except when queue_only is set, to avoid logging races between
2508 the different processes.
2510 PH/38 Panic-die if .include specifies a non-absolute path.
2512 PH/39 A tweak to the "H" retry rule from its user.
2514 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2515 a label. They prevented compilation on older perls.
2517 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2518 a warning to be raised on newish perls.
2520 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2521 on queue. Changes to match documented behaviour of showing count of
2522 messages matching specified criteria.
2524 PH/40 Changed the default ident timeout from 30s to 5s.
2526 PH/41 Added support for the use of login_cap features, on those BSD systems
2527 that have them, for controlling the resources used by pipe deliveries.
2529 PH/42 The content-scanning code uses fopen() to create files in which to put
2530 message data. Previously it was not paying any attention to the mode of
2531 the files. Exim runs with umask(0) because the rest of the code creates
2532 files with open(), and sets the required mode explicitly. Thus, these
2533 files were ending up world-writeable. This was not a big issue, because,
2534 being within the spool directory, they were not world-accessible. I have
2535 created a function called modefopen, which takes an additional mode
2536 argument. It sets umask(777), creates the file, chmods it to the required
2537 mode, then resets the umask. All the relevant calls to fopen() in the
2538 content scanning code have been changed to use this function.
2540 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2541 to 24 hours. This avoids potential overflow problems when processing G
2542 and H retry rules. I suspect nobody ever tinkers with this value.
2544 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2546 PH/45 When the plaintext authenticator is running as a client, the server's
2547 challenges are checked to ensure they are valid base64 strings. By
2548 default, the authentication attempt is cancelled if an invalid string is
2549 received. Setting client_ignore_invalid_base64 true ignores these errors.
2550 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2551 they are received. Thus, the responses can be made to depend on the
2552 challenges. If an invalid string is ignored, an empty string is placed in
2555 PH/46 Messages that are created by the autoreply transport now contains a
2556 References: header, in accordance with RFCs 2822 and 3834.
2558 PH/47 Added authenticated_sender_force to the smtp transport.
2560 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2562 PH/49 Installed latest patch from the Sieve maintainer.
2564 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2565 was also set, the appendfile transport was unnecessarily scanning a
2566 directory of message files (e.g. for maildir delivery) to find the count
2567 of files (along with the size), even though it did not need this
2568 information. It now does the scan only if it needs to find either the
2569 size of the count of files.
2571 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2573 PH/52 Two bugs concerned with error handling when the smtp transport is
2576 (i) Exim was not creating retry information for temporary errors given
2577 for individual recipients after the DATA command when the smtp transport
2578 was used in LMTP mode. This meant that they could be retried too
2579 frequently, and not timed out correctly.
2581 (ii) Exim was setting the flag that allows error details to be returned
2582 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2583 recipients that were returned after the DATA command.
2585 PH/53 This is related to PH/52, but is more general: for any failing address,
2586 when detailed error information was permitted to be returned to the
2587 sender, but the error was temporary, then after the final timeout, only
2588 "retry timeout exceeded" was returned. Now it returns the full error as
2589 well as "retry timeout exceeded".
2591 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2592 do this, and (what is worse) MTAs that accept it.
2594 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2595 will now be deprecated.
2597 PH/56 New os.c-cygwin from the Cygwin maintainer.
2599 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2600 formats (previously only available in exim formats via -bpr, -bpru,
2601 and -bpra. Now also available in native and exiqgrep formats)
2603 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2604 with very large, slow to parse queues
2606 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2608 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2610 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2611 responses to authentication challenges, though it was showing the
2612 challenges; (ii) I've removed the CR characters from the debug output for
2615 PH/58 Allow for the insertion of a newline as well as a space when a string
2616 is turned into more than one encoded-word during RFC 2047 encoding. The
2617 Sieve code now uses this.
2619 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2620 data_4xx, lost_connection, tls_required.
2622 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2623 message was being sent as an SMTP response.
2625 PH/61 Add -l and -k options to exicyclog.
2627 PH/62 When verifying, if an address was redirected to one new address, so that
2628 verification continued, and the new address failed or deferred after
2629 having set something in $address_data, the value of $address_data was not
2630 passed back to the ACL. This was different to the case when no
2631 redirection occurred. The value is now passed back in both cases.
2633 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2634 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2635 use login_cap.h, so on its own it isn't the distinguishing feature. The
2636 new name refers directly to the setclassresources() function.
2638 PH/65 Added configuration files for NetBSD3.
2640 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2642 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2643 is preferred over IPv4.
2645 PH/68 The bounce_return_message and bounce_return_body options were not being
2646 honoured for bounces generated during the reception of non-SMTP messages.
2647 In particular, this applied to messages rejected by the ACL. This bug has
2648 been fixed. However, if bounce_return_message is true and bounce_return_
2649 body is false, the headers that are returned for a non-SMTP message
2650 include only those that have been read before the error was detected.
2651 (In the case of an ACL rejection, they have all been read.)
2653 PH/69 The HTML version of the specification is now built in a directory called
2654 spec_html instead of spec.html, because the latter looks like a path with
2655 a MIME-type, and this confuses some software.
2657 PH/70 Catch two compiler warnings in sieve.c.
2659 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2660 function verify_get_ident() calls ip_connect() to connect a socket, but
2661 if the "connect()" function timed out, ip_connect() used to close the
2662 socket. However, verify_get_ident() also closes the socket later, and in
2663 between Exim writes to the log, which may get opened at this point. When
2664 the socket was closed in ip_connect(), the log could get the same file
2665 descriptor number as the socket. This naturally causes chaos. The fix is
2666 not to close the socket in ip_connect(); the socket should be closed by
2667 the function that creates it. There was only one place in the code where
2668 this was missing, in the iplookup router, which I don't think anybody now
2669 uses, but I've fixed it anyway.
2671 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2672 well as to direct DNS lookups. Otherwise the handling of names in host
2673 lists is inconsistent and therefore confusing.
2679 PH/01 Two changes to the default runtime configuration:
2681 (1) Move the checks for relay_from_hosts and authenticated clients from
2682 after to before the (commented out) DNS black list checks.
2684 (2) Add control=submission to the relay_from_hosts and authenticated
2685 clients checks, on the grounds that messages accepted by these
2686 statements are most likely to be submissions.
2688 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2690 (1) Generate an error if the third argument for the ${prvs expansion is
2693 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2696 (3) Reset the variables that are obtained from the first argument of
2697 ${prvscheck and used in the second argument before leaving the code,
2698 because their memory is reclaimed, so using them afterwards may do
2701 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2702 one (it's much easier than Tom thought :-).
2704 (5) Because of (4), we can now allow for the use of $prvscheck_result
2705 inside the third argument.
2707 PH/03 For some reason, the default setting of PATH when running a command from
2708 a pipe transport was just "/usr/bin". I have changed it to
2711 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2712 anything to be listed in the output from -bV.
2714 PH/05 When a filter generated an autoreply, the entire To: header line was
2715 quoted in the delivery log line, like this:
2717 => >A.N.Other <ano@some.domain> <original@ddress> ...
2719 This has been changed so that it extracts the operative address. There
2720 may be more than one such address. If so, they are comma-separated, like
2723 => >ano@some.domain,ona@other.domain <original@ddress> ...
2725 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2726 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2727 not being looked up in the rDNS to get a host name, Exim was showing the
2728 IP address twice in Received: lines, even though the IP addresses were
2729 identical. For example:
2731 Received: from [1.2.3.4] (helo=[1.2.3.4])
2733 However, if the real host name was known, it was omitting the HELO data
2734 if it matched the actual IP address. This has been tidied up so that it
2735 doesn't show the same IP address twice.
2737 PH/07 When both +timestamp and +memory debugging was on, the value given by
2738 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2739 called by the debug printing, thereby overwriting the timestamp buffer.
2740 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2742 PH/08 When the original message was included in an autoreply transport, it
2743 always said "this is a copy of the message, including all the headers",
2744 even if body_only or headers_only was set. It now gives an appropriate
2747 PH/09 Applied a patch from the Sieve maintainer which:
2749 o fixes some comments
2750 o adds the (disabled) notify extension core
2751 o adds some debug output for the result of if/elsif tests
2752 o points to the current vacation draft in the documentation
2753 and documents the missing references header update
2757 o fixes a bug in processing the envelope test (when testing
2758 multiple envelope elements, the last element determinted the
2761 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2762 Electronic Mail") by including:
2764 Auto-submitted: auto-generated
2766 in the messages that it generates (bounce messages and others, such as
2767 warnings). In the case of bounce messages for non-SMTP mesages, there was
2768 also a typo: it was using "Auto_submitted" (underscore instead of
2769 hyphen). Since every message generated by Exim is necessarily in response
2770 to another message, thes have all been changed to:
2772 Auto-Submitted: auto-replied
2774 in accordance with these statements in the RFC:
2776 The auto-replied keyword:
2778 - SHOULD be used on messages sent in direct response to another
2779 message by an automatic process,
2781 - MUST NOT be used on manually-generated messages,
2783 - MAY be used on Delivery Status Notifications (DSNs) and Message
2784 Disposition Notifications (MDNs),
2786 - MUST NOT be used on messages generated by automatic or periodic
2787 processes, except for messages which are automatic responses to
2790 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2791 to the default Received: header definition.
2793 PH/12 Added log selector acl_warn_skipped (default on).
2795 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2796 variables because (a) they are in the wrong storage pool and (b) even if
2797 they were copied, it wouldn't work properly because of the caching.
2799 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2800 checking when decoding. Apparently there are clients that generate
2801 overlong encoded strings. Why am I not surprised?
2803 PH/15 If the first argument of "${if match_address" was not empty, but did not
2804 contain an "@" character, Exim crashed. Now it writes a panic log message
2805 and treats the condition as false.
2807 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2809 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2810 extension "envelope-auth". The code is finished and in agreement with
2811 other implementations, but there is no documentation so far and in fact,
2812 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2813 not changing the active code.
2815 Print executed "if" and "elsif" statements when debugging is used. This
2816 helps a great deal to understand what a filter does.
2818 Document more things not specified clearly in RFC3028. I had all this
2819 sorted out, when out of a sudden new issues came to my mind. Oops."
2821 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2824 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2825 canonical form (as documented). However, after a host name lookup from
2826 the IP address, check_host() was doing a simple string comparison with
2827 addresses acquired from the DNS when checking that the found name did
2828 have the original IP as one of its addresses. Since any found IPv6
2829 addresses are likely to be in abbreviated form, the comparison could
2830 fail. Luckily, there already exists a function for doing the comparison
2831 by converting both addresses to binary, so now that is used instead of
2832 the text comparison.
2834 PH/20 There was another similar case to PH/19, when a complete host name was
2835 given in a host list; looking up its IP address could give an abbreviated
2836 form, whereas the current host's name might or might not be abbreviated.
2837 The same fix has been applied.
2843 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2844 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2847 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2848 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2850 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2852 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2853 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2854 Also, for "Auto-Submitted", treat anything other than "no" as
2855 non-personal, in accordance with RFC 3834. (Previously it treated
2856 anything starting "auto-" as non-personal.)
2858 TF/01 The control=submission/name=... option had a problem with syntax
2859 errors if the name included a slash character. The /name= option
2860 now slurps the rest of the string, so it can include any characters
2861 but it must come last in the list of options (after /sender_retain
2864 PH/05 Some modifications to the interface to the fake nameserver for the new
2872 TK/01 Added the "success_on_redirect" address verification option. See
2873 NewStuff for rationale and an example.
2875 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2877 PH/02 Patch to exigrep to allow it to work on syslog lines.
2879 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2880 fread() to skip over the body file's header line, because in Cygwin the
2881 header line is locked and is inaccessible.
2883 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2884 co-exist for some time) to make it clear that it is the Exim ID that is
2885 referenced, not the Message-ID: header line.
2887 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2888 string_format() function, because snprintf() does not exist on all
2891 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2892 ${stat: expansion item.
2894 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2895 protocol synchronization error", to keep the pedants happy.
2897 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2898 well as for IRIX systems, when gcc is being used. See the host.c source
2901 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2903 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2906 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2907 transport and to the smtp transport in LMTP mode.
2909 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2911 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2912 run a filter in a subprocess. This could lead to confusion in subsequent
2913 lookups in the parent process. There should also be a search_tidyup() at
2914 the end of the subprocess.
2916 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2917 only if the host matched helo_try_verify_hosts, which caused the
2918 verification to occur when the EHLO/HELO command was issued. The ACL just
2919 tested the remembered result. Now, if a previous verification attempt has
2920 not happened, "verify = helo" does it there and then.
2922 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2924 TK/03 Fix log output including CR from clamd.
2926 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2927 exist provoked a memory error which could cause a segfault.
2929 PH/15 Installed PCRE 6.2
2931 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2933 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2934 of the problem. Specifically, suggested +O2 rather than +O1 for the
2937 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2939 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2940 record, the port was not correctly logged when the outgoing_port log
2941 selector was set (it logged the transort's default port).
2943 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2944 fallback_hosts, and "hosts" in the smtp transport.
2946 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2947 host errors such as "Connection refused".
2949 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2950 authentication with radiusclient 0.4.9:
2952 - Error returned from rc_read_config was caught wrongly
2953 - Username/password not passed on to radius server due to wrong length.
2955 The presumption is that some radiusclient API changes for 4.51/PH/17
2956 were not taken care of correctly. The code is still untested by me (my
2957 Linux distribution still has 0.3.2 of radiusclient), but it was
2958 contributed by a Radius user.
2960 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2961 expanding the "port" option of the smtp transport.
2963 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2964 while reading a MIME header. Thanks to Tom Hughes for a patch.
2966 PH/24 Include config.h inside local_scan.h so that configuration settings are
2969 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2970 that in an ACL for RCPT (for example), you can examine exactly what was
2973 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2974 commands, but it was not correctly comparing the address with the actual
2975 client host address. Thus, it would show the EHLO address in Received:
2976 header lines when this was not necessary.
2978 PH/27 Added the % operator to ${eval:}.
2980 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2981 it should be ignoring failures (because with -C, for example, it has lost
2982 privilege). It wasn't ignoring creation failures other than "already
2985 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2988 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2989 because an input line was too long, either on its own, or by virtue of
2990 too many continuations, the temporary file was not being removed, and the
2991 return code was incorrect.
2993 PH/31 Missing "BOOL" in function definition in filtertest.c.
2995 PH/32 Applied Sieve patches from the maintainer.
2997 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2999 PH/33 Added "verify = not_blind".
3001 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3002 Local/Makefile (with some defaults set). These are used in built scripts
3003 such as exicyclog, but they have never been used in the exim_install
3004 script (though there are many overriding facilities there). I have
3005 arranged that the exim_install script now takes note of these two
3008 PH/35 Installed configuration files for Dragonfly.
3010 PH/36 When a locally submitted message by a trusted user did not contain a
3011 From: header, and the sender address was obtained from -f or from an SMTP
3012 MAIL command, and the trusted user did not use -F to supply a sender
3013 name, $originator_name was incorrectly used when constructing a From:
3014 header. Furthermore, $originator_name was used for submission mode
3015 messages from external hosts without From: headers in a similar way,
3016 which is clearly wrong.
3018 PH/37 Added control=suppress_local_fixups.
3020 PH/38 When log_selector = +received_sender was set, and the addition of the
3021 sender made the log line's construction buffer exactly full, or one byte
3022 less than full, an overflow happened when the terminating "\n" was
3025 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3026 when the result of a list match is failure because a DNS lookup failed.
3028 PH/40 RM_COMMAND is now used in the building process.
3030 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3031 the "build-* directories that it finds.
3033 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3034 address in a domain literal was a prefix of an interface address.
3036 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3037 when verifying a sender address, unless rewrite_headers is false.
3039 PH/44 Wrote a long comment about why errors_to addresses are verified as
3040 recipients, not senders.
3042 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3043 the ratelimit ACL was added.
3045 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3047 PH/46 Added extra information about PostgreSQL errors to the error string.
3049 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3050 avoiding the need to install special zones in a real server. This is
3051 backwards compatible; if it can't find the fake resolver, it drops back.
3052 Thus, both old and new test suites can be run.
3054 TF/02 Added util/ratelimit.pl
3056 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3057 clock is set back in time.
3059 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3060 Candler <B.Candler@pobox.com>.
3062 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3063 OK for addresses that are the result of redirections.
3065 PH/48 A number of further additions for the benefit of the new test suite,
3066 including a fake gethostbyname() that interfaces to the fake DNS resolver
3069 TF/06 The fix for widen_domains has also been applied to qualify_single and
3070 search_parents which are the other dnslookup options that can cause
3073 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3076 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3078 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3079 service" if it encountered an MX record with an empty target hostname.
3080 The message is now "an MX or SRV record indicated no SMTP service".
3082 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3083 if the DNS of the sending site is misconfigured. This is quite a
3084 common situation. This change restores the behaviour of treating a
3085 helo verification defer as a failure.
3087 PH/51 If self=fail was set on a router, the bounce message did not include the
3088 actual error message.
3094 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3096 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3097 command itself ended in error, the underlying message about the transport
3098 filter timeout was being overwritten with the pipe command error. Now the
3099 underlying error message should be appended to the second error message.
3101 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3103 PH/02 Reduce the amount of output that "make" produces by default. Full output
3104 can still be requested.
3106 PH/03 The warning log line about a condition test deferring for a "warn" verb
3107 was being output only once per connection, rather than after each
3108 occurrence (because it was using the same function as for successful
3109 "warn" verbs). This seems wrong, so I have changed it.
3111 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3112 it should not have, which might have caused a crash in the right
3113 circumstances, but probably never did.
3115 PH/04 Installed a modified version of Tony Finch's patch to make submission
3116 mode fix the return path as well as the Sender: header line, and to
3117 add a /name= option so that you can make the user's friendly name appear
3120 TF/03 Added the control = fakedefer ACL modifier.
3122 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3123 Mark Lowes for thorough testing.
3125 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3127 TK/03 Merged latest SRS patch from Miles Wilton.
3129 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3130 with the definition in sysexits.h (which is #included earlier).
3131 Fortunately, Exim does not actually use EX_OK. The code used to try to
3132 preserve the sysexits.h value, by assumimg that macro definitions were
3133 scanned for macro replacements. I have been disabused of this notion,
3134 so now the code just undefines EX_OK before #including unistd.h.
3136 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3137 in the smtp transport. When a block could not be written in a single
3138 write() function, the timeout was being re-applied to each part-write.
3139 This seems wrong - if the receiver was accepting one byte at a time it
3140 would take for ever. The timeout is now adjusted when this happens. It
3141 doesn't have to be particularly precise.
3143 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3144 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3146 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3147 without a domain if the check to <postmaster@domain> fails.
3149 SC/01 Eximstats: added -xls and the ability to specify output files
3150 (patch written by Frank Heydlauf).
3152 SC/02 Eximstats: use FileHandles for outputing results.
3154 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3156 SC/04 Eximstats: fixed display of large numbers with -nvr option
3158 SC/05 Eximstats: fixed merging of reports with empty tables.
3160 SC/06 Eximstats: added the -include_original_destination flag
3162 SC/07 Eximstats: removed tabs and trailing whitespace.
3164 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3166 TK/06 MBOX spool code: Add real "From " MBOX separator line
3167 so the .eml file is really in mbox format (even though
3168 most programs do not really care). Patch from Alex Miller.
3170 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3171 The latter is generated from $received_to and is only set if the
3172 message has one envelope recipient. SA can use these headers,
3173 obviously out-of-the-box. Patch from Alex Miller.
3175 PH/08 The ${def test on a variable was returning false if the variable's
3176 value was "0", contrary to what the specification has always said!
3177 The result should be true unless the variable is empty.
3179 PH/09 The syntax error of a character other than { following "${if
3180 def:variable_name" (after optional whitespace) was not being diagnosed.
3181 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3182 accidental colon was present, for example, could give incorrect results.
3184 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3185 result is used (not including appendfile, where other changes are about
3188 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3189 This involved changing a lot of size variables from int to off_t. It
3190 should work with maildirs and everything.
3192 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3193 spamd dying while we are connected to it.
3195 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3198 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3199 function for BDB 4.3.
3201 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3203 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3204 the sources for PCRE have been re-arranged and more files are now
3207 PH/15 The code I had for printing potentially long long variables in PH/11
3208 above was not the best (it lost precision). The length of off_t variables
3209 is now inspected at build time, and an appropriate printing format (%ld
3210 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3211 to be "long long int" or "long int". This is needed for the internal
3212 formatting function string_vformat().
3214 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3215 the configuration file to be ":syslog", then the script "guesses" where
3216 the logs files are, rather than using the compiled in default. In our
3217 case the guess is not the same as the compiled default, so the script
3218 suddenly stopped working when I started to use syslog. The patch checks
3219 to see if log_file_path is "". If so, it attempts to read it from exim
3220 with no configuration file to get the compiled in version, before it
3221 falls back to the previous guessing code."
3223 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3224 implementing BATV in an Exim configuration. See NewStuff for the gory
3227 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3228 Makefile that are specific to HP-UX.
3230 PH/18 If the "use_postmaster" option was set for a recipient callout together
3231 with the "random" option, the postmaster address was used as the MAIL
3232 FROM address for the random test, but not for the subsequent recipient
3233 test. It is now used for both.
3235 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3236 patch removes a few documentation additions to RFC 3028, because the
3237 latest draft now contains them. It adds the new en;ascii-case comparator
3238 and a new error check for 8bit text in MIME parts. Comparator and
3239 require names are now matched exactly. I enabled the subaddress
3240 extension, but it is not well tested yet (read: it works for me)."
3242 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3243 rework some of the code of TK/09 above to avoid the hardwired use of
3244 "%lld" and "long long". Replaced the call to snprintf() with a call to
3247 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3248 records point to non-existent hosts", "retry timeout exceeded", and
3249 "retry time not reached for any host after a long failure period".
3251 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3252 experimental DomainKeys support:
3254 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3255 (2) On an error such as an illegally used "control", the wrong name for
3256 the control was given.
3258 These problems did NOT occur unless DomainKeys support was compiled.
3260 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3262 PH/24 Added ${if match_ip condition.
3264 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3265 databases so that it will be absolutely obvious if a crash occurs in the
3266 DB library. This is a regular occurrence (often caused by mis-matched
3269 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3270 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3271 picked up on a user's system that detects such things. There doesn't seem
3272 to be a gcc warning option for this - only an attribute that has to be
3273 put on the function's prototype. It seems that in Fedora Core 4 they have
3274 set this on a number of new functions. No doubt there will be more in due
3277 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3278 specify a transport. However, if an address that was verified by such a
3279 router was the subject of a callout, Exim crashed because it tried to
3280 read the rcpt_include_affixes from the non-existent transport. Now it
3281 just assumes that the setting of that option is false. This bug was
3282 introduced by 4.51/PH/31.
3284 PH/28 Changed -d+all to exclude +memory, because that information is very
3285 rarely of interest, but it makes the output a lot bigger. People tend to
3286 do -d+all out of habit.
3288 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3289 code in os-type was giving problems when libc.so lives in lib64, like on
3292 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3293 aren't the modern standard, and it seems that some systems' include files
3294 don't always have them. Exim was already checking for some of the newer
3295 ones like T_AAAA, and defining it itself. I've added checks for all the
3296 record types that Exim uses.
3298 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3299 not automatically generating a new one, as it is supposed to. This
3300 prevented TLS from working. If the file did exist, but contained invalid
3301 data, a new version was generated, as expected. It was only the case of a
3302 non-existent file that was broken.
3304 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3305 with a change in libdomainkeys > 0.64.
3307 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3308 from DNS. If the selector record carries the flag, it now has
3309 precedence over the domain-wide flag.
3311 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3313 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3314 the use of a port name that isn't defined in /etc/services) occurred, the
3315 message was deferred as in a normal delivery, and thus remained on the
3316 spool, instead of being failed because of the mua_wrapper setting. This
3317 is now fixed, and I tidied up some of the mua_wrapper messages at the
3320 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3321 the messages in a hash of arrays rather than using individual hashes.
3322 This is a bit cleaner and results in dramatic memory savings, albeit
3323 at a slight CPU cost.
3325 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3326 as requested by Marc Sherman.
3328 SC/10 Eximstats: added histograms for user specified patterns as requested
3331 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3333 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3334 fopen() in the content-scanning modules that did not already have it.
3340 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3341 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3343 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3345 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3347 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3348 to test Sieve filters that use "vacation".
3350 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3351 that changes the way the GnuTLS parameters are stored in the cache file.
3352 The new format can be generated externally. For backward compatibility,
3353 if the data in the cache doesn't make sense, Exim assumes it has read an
3354 old-format file, and it generates new data and writes a new file. This
3355 means that you can't go back to an older release without removing the
3358 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3359 work if there are any delivery delays because "one_time" forces the
3360 parent to be marked "delivered", so its unseen clone is never tried
3361 again. For this reason, Exim now forbids the simultaneous setting of
3364 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3365 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3366 entry if you want to know the details. The fix, however, neglected to
3367 consider the case when local delivery batching is involved. The test for
3368 "previously delivered" was not happening when checking to see if an
3369 address could be batched with a previous (undelivered) one; under
3370 certain circumstances this could lead to multiple deliveries to the same
3373 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3374 in its include files, and this causes problems building Exim.
3376 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3377 header_syntax) but Exim was just ignoring anything given after a slash.
3378 In particular, this caused confusion with an attempt to use "verify =
3379 reverse_host_lookup/defer_ok". An error is now given when options are
3380 supplied for verify items that do not have them. (Maybe reverse_host_
3381 lookup should have a defer_ok option, but that's a different point.)
3383 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3384 defined by RFC 821) to 2048, because there were problems with some AUTH
3385 commands, and RFC 1869 says the size should be increased for extended
3386 SMTP commands that take arguments.
3388 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3391 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3392 "unknown" error; now it says that the functionality isn't in the binary.
3394 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3395 an address' error message when a string expansion fails (syntax or
3396 whatever). Otherwise the password may appear in the log. Following change
3397 PH/42 below, there is no longer a chance of it appearing in a bounce
3400 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3402 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3403 of its addresses was ever tried. (Bugzilla bug #2.)
3405 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3406 the result incorrectly in the debug output. (It correctly added a newline
3407 to what was transported.)
3409 TF/01 Added $received_time.
3411 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3412 commented out examples of how to interface to a virus scanner and to
3413 SpamAssassin. Also added commented examples of av_scanner and
3414 spamd_address settings.
3416 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3417 and controls are allowed in which ACLs. There were a couple of minor
3418 errors. Some of the entries in the conditions table (which is a table of
3419 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3420 negation of where the condition IS allowed.
3422 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3424 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3425 header file does not have a version number, so I've had to invent a new
3426 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3427 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3428 radiusclient), but it was contributed by a Radius user.
3430 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3431 files or directories, for OpenSSL.
3433 PH/19 When an Exim process that is running as root has to create an Exim log
3434 file, it does so in a subprocess that runs as exim:exim so as to get the
3435 ownership right at creation (otherwise, other Exim processes might see
3436 the file with the wrong ownership). There was no test for failure of this
3437 fork() call, which would lead to the process getting stuck as it waited
3438 for a non-existent subprocess. Forks do occasionally fail when resources
3439 run out. I reviewed all the other calls to fork(); they all seem to check
3442 PH/20 When checking for unexpected SMTP input at connect time (before writing
3443 the banner), Exim was not dealing correctly with a non-positive return
3444 from the read() function. If the client had disconnected by this time,
3445 the result was a log entry for a synchronization error with an empty
3446 string after "input=" when read() returned zero. If read() returned -1
3447 (an event I could not check), uninitialized data bytes were printed.
3448 There were reports of junk text (parts of files, etc) appearing after
3451 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3453 PH/22 Added support for macro redefinition, and (re)definition in between
3454 driver and ACL definitions.
3456 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3457 forgetting to use the resulting value; it was using the unexpanded value.
3459 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3460 hadn't been configured. The fix is from Juergen Kreileder, who
3461 understands it better than I do:
3463 "Here's what I see happening with three configured cyrus_sasl
3464 authenticators configured (plain, login, cram-md5):
3466 On startup auth_cyrus_sasl_init() gets called for each of these.
3467 This means three calls to sasl_listmech() without a specified mech_list.
3468 => SASL tests which mechs of all available mechs actually work
3469 => three warnings about OTP not working
3470 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3472 With the patch, sasl_listmech() also gets called three times. But now
3473 SASL's mech_list option is set to the server_mech specified in the the
3474 authenticator. Or in other words, the answer from sasl_listmech()
3475 gets limited to just the mech you're testing for (which is different
3477 => the return list contains just 'plain' or 'login', 'cram-md5' or
3478 nothing depending on the value of ob->server_mech.
3480 I've just tested the patch: Authentication still works fine,
3481 unavailable mechs specified in the exim configuration are still
3482 caught, and the auth.log warnings about OTP are gone."
3484 PH/25 When debugging is enabled, the contents of the command line are added
3485 to the debugging output, even when log_selector=+arguments is not
3488 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3489 answer is "GNU", and only if the return is "GNU/something" is the answer
3492 PH/27 $acl_verify_message is now set immediately after the failure of a
3493 verification in an ACL, and so is available in subsequent modifiers. In
3494 particular, the message can be preserved by coding like this:
3496 warn !verify = sender
3497 set acl_m0 = $acl_verify_message
3499 Previously, $acl_verify_message was set only while expanding "message"
3500 and "log_message" when a very denied access.
3502 PH/28 Modified OS/os.c-Linux with
3504 -#ifndef OS_LOAD_AVERAGE
3505 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3507 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3508 nomenclature these days.)
3510 PH/29 Installed patch from the Sieve maintainer that adds the options
3511 sieve_useraddress and sieve_subaddress to the redirect router.
3513 PH/30 In these circumstances:
3514 . Two addresses routed to the same list of hosts;
3515 . First host does not offer TLS;
3516 . First host accepts first address;
3517 . First host gives temporary error to second address;
3518 . Second host offers TLS and a TLS session is established;
3519 . Second host accepts second address.
3520 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3521 and peerdn, if requested) that were in fact used only for the second
3524 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3525 attention to any local part prefix or suffix that was matched by the
3526 router that accepted the address. It now behaves in the same way as it
3527 does for delivery: the affixes are removed from the local part unless
3528 rcpt_include_affixes is set on the transport.
3530 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3531 timeout during the DATA phase of an incoming message.
3533 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3534 applied a patch sanctioned by the Sieve maintainer.
3536 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3537 the uid or gid is negative. A case of a negative gid caused this to be
3538 noticed. The fix allows for either to be negative.
3540 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3541 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3544 PH/36 The change PH/12 above was broken. Fixed it.
3546 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3547 the grounds that routing the same address twice would always produce the
3548 same answer. This might have been true once, but it is certainly no
3549 longer true now. Routing a child address may depend on the previous
3550 routing that produced that child. Some complicated redirection strategies
3551 went wrong when messages had multiple recipients, and made Exim's
3552 behaviour dependent on the order in which the addresses were given.
3554 I have moved the duplicate checking until after the routing is complete.
3555 Exim scans the addresses that are assigned to local and remote
3556 transports, and removes any duplicates. This means that more work will be
3557 done, as duplicates will always all be routed, but duplicates are
3558 presumably rare, so I don't expect this is of any significance.
3560 For deliveries to pipes, files, and autoreplies, the duplicate checking
3561 still happens during the routing process, since they are not going to be
3564 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3565 It corrects a timeout issue with spamd. This is Ian's comment: "The
3566 background is that sometimes spamd either never reads data from a
3567 connection it has accepted, or it never writes response data. The exiscan
3568 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3569 blindly assumes that writes won't block so it may never time out."
3571 PH/39 Allow G after quota size as well as K and M.
3573 PH/40 The value set for $authenticated_id in an authenticator may not contain
3574 binary zeroes or newlines because the value is written to log lines and
3575 to spool files. There was no check on this. Now the value is run through
3576 the string_printing() function so that such characters are converted to
3577 printable escape sequences.
3579 PH/41 $message_linecount is a new variable that contains the total number of
3580 lines in the message. Compare $body_linecount, which is the count for the
3583 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3584 bounce and delay warning messages, except in certain special cases, which
3587 (a) An SMTP error message from a remote host;
3588 (b) A message specified in a :fail: redirection;
3589 (c) A message specified in a "fail" command in a system filter;
3590 (d) A message specified in a FAIL return from the queryprogram router;
3591 (e) A message specified by the cannot_route_message router option.
3593 In these cases only, Exim does include the error details in bounce and
3594 warning messages. There are also a few cases where bland messages such
3595 as "unrouteable address" or "local delivery error" are given.
3597 PH/43 $value is now also set for the "else" part of a ${run expansion.
3599 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3600 being worked on, but at least Exim now implements the latest version to
3603 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3604 process to complete was treated as a delivery failure, a timeout while
3605 writing the message to the pipe was logged, but erroneously treated as a
3606 successful delivery. Such timeouts include transport filter timeouts. For
3607 consistency with the overall process timeout, these timeouts are now
3608 treated as errors, giving rise to delivery failures by default. However,
3609 there is now a new Boolean option for the pipe transport called
3610 timeout_defer, which, if set TRUE, converts the failures into defers for
3611 both kinds of timeout. A transport filter timeout is now identified in
3614 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3615 systems where "make" and "gmake" are different, calling "gmake" at top
3616 level broke things. I've arranged for the value of $(MAKE) to be passed
3617 from the Makefile to this script so that it can call the same version of
3621 A note about Exim versions 4.44 and 4.50
3622 ----------------------------------------
3624 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3625 changes of various kinds. As a consequence, a big documentation update was
3626 needed. This delayed the release for rather longer than seemed good, especially
3627 in the light of a couple of (minor) security issues. Therefore, the changes
3628 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3629 release. So 4.44 and 4.50 are in effect two different branches that both start
3632 I have left the 4.50 change log unchanged; it contains all the changes since
3633 4.43. The change log for 4.44 is below; many of its items are identical to
3634 those for 4.50. This seems to be the most sensible way to preserve the
3635 historical information.
3641 1. Minor wording change to the doc/README.SIEVE file.
3643 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3644 computation of the current number of files was incorrect.
3646 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3647 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3650 4. Give more explanation in the error message when the command for a transport
3651 filter fails to execute.
3653 5. There are several places where Exim runs a non-Exim command in a
3654 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3655 was being done only for the command run by the queryprogram router. It is
3656 now done for all such subprocesses. The other cases are: ${run, transport
3657 filters, and the commands run by the lmtp and pipe transports.
3659 6. Added CONFIGURE_GROUP build-time option.
3661 7. Some older OS have a limit of 256 on the maximum number of file
3662 descriptors. Exim was using setrlimit() to set 1000 as a large value
3663 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3664 systems. I've change it so that if it can't get 1000, it tries for 256.
3666 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3667 was an oversight, and furthermore, ever since the addition of extra
3668 controls (e.g. 4.43/32), the checks on when to allow different forms of
3669 "control" were broken. There should now be diagnostics for all cases when a
3670 control that does not make sense is encountered.
3672 9. Added the /retain_sender option to "control=submission".
3674 10. $recipients is now available in the predata ACL (oversight).
3676 11. Tidy the search cache before the fork to do a delivery from a message
3677 received from the command line. Otherwise the child will trigger a lookup
3678 failure and thereby defer the delivery if it tries to use (for example) a
3679 cached ldap connection that the parent has called unbind on.
3681 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3682 of $address_data from the recipient verification was clobbered by the
3683 sender verification.
3685 13. The value of address_data from a sender verification is now available in
3686 $sender_address_data in subsequent conditions in the ACL statement.
3688 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3690 15. Added a new option "connect=<time>" to callout options, to set a different
3693 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3694 was its contents. (It was OK if the option was not defined at all.)
3696 17. A "Completed" log line is now written for messages that are removed from
3697 the spool by the -Mrm option.
3699 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3700 information about exactly what failed.
3702 19. Added -dd to debug only the daemon process.
3704 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3705 handles timeouts, both on the server side and network timeouts. Renamed the
3706 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3708 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3709 It is now set to "smtps".
3711 22. $host_address is now set to the target address during the checking of
3712 ignore_target_hosts.
3714 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3715 being passed; this would have caused $sender_host_name to have been used if
3716 matching the list had actually called for a host name (not very likely,
3717 since this list is usually IP addresses). A host name is now passed as
3720 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3721 code that specifies a non-restarting handler (typically sigaction() in
3722 modern systems) in an attempt to fix a rare and obscure crash bug.
3724 25. Narrowed the window for a race in the daemon that could cause it to ignore
3725 SIGCHLD signals. This is not a major problem, because they are used only to
3726 wake it up if nothing else does.
3728 26. A malformed maildirsize file could cause Exim to calculate negative values
3729 for the mailbox size or file count. Odd effects could occur as a result.
3730 The maildirsize information is now recalculated if the size or filecount
3733 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3734 support for a long time. Removed HAVE_SYS_VFS_H.
3736 28. Installed the latest version of exipick from John Jetmore.
3738 29. In an address list, if the pattern was not a regular expression, an empty
3739 subject address (from a bounce message) matched only if the pattern was an
3740 empty string. Non-empty patterns were not even tested. This was the wrong
3741 because it is perfectly reasonable to use an empty address as part of a
3742 database query. An empty address is now tested by patterns that are
3743 lookups. However, all the other forms of pattern expect the subject to
3744 contain a local part and a domain, and therefore, for them, an empty
3745 address still always fails if the pattern is not itself empty.
3747 30. Exim went into a mad DNS loop when attempting to do a callout where the
3748 host was specified on an smtp transport, and looking it up yielded more
3749 than one IP address.
3751 31. Re-factored the code for checking spool and log partition space into a
3752 function that finds that data and another that does the check. The former
3753 is then used to implement four new variables: $spool_space, $log_space,
3754 $spool_inodes, and $log_inodes.
3756 32. The RFC2047 encoding function was originally intended for short strings
3757 such as real names; it was not keeping to the 75-character limit for
3758 encoded words that the RFC imposes. It now respects the limit, and
3759 generates multiple encoded words if necessary. To be on the safe side, I
3760 have increased the buffer size for the ${rfc2047: expansion operator from
3763 33. It is now permitted to omit both strings after an "if" condition; if the
3764 condition is true, the result is "true". As before, when the second string
3765 is omitted, a false condition yields an empty string. This makes it less
3766 cumbersome to write custom ACL and router conditions.
3768 34. Failure to deliver a bounce message always caused it to be frozen, even if
3769 there was an errors_to setting on the router. The errors_to setting is now
3772 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3773 canonical form (fully expanded) before being placed in
3774 $sender_host_address.
3776 36. The table in the code that translates DNS record types into text (T_A to
3777 "A" for instance) was missing entries for NS and CNAME. It is just possible
3778 that this could have caused confusion if both these types were looked up
3779 for the same domain, because the text type is used as part of Exim's
3780 per-process caching. But the chance of anyone hitting this buglet seems
3783 37. The dnsdb lookup has been extended in a number of ways.
3785 (1) There is a new type, "zns", which walks up the domain tree until it
3786 finds some nameserver records. It should be used with care.
3788 (2) There is a new type, "mxh", which is like "mx" except that it returns
3789 just the host names, not the priorities.
3791 (3) It is now possible to give a list of domains (or IP addresses) to be
3792 looked up. The behaviour when one of the lookups defers can be
3793 controlled by a keyword.
3795 (4) It is now possible to specify the separator character for use when
3796 multiple records are returned.
3798 38. The dnslists ACL condition has been extended: it is now possible to supply
3799 a list of IP addresses and/or domains to be looked up in a particular DNS
3802 39. Added log_selector=+queue_time_overall.
3804 40. When running the queue in the test harness, wait just a tad after forking a
3805 delivery process, to get repeatability of debugging output.
3807 41. Include certificate and key file names in error message when GnuTLS fails
3808 to set them up, because the GnuTLS error message doesn't include the name
3809 of the failing file when there is a problem reading it.
3811 42. Allow both -bf and -bF in the same test run.
3813 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3815 44. The "Exiscan patch" is now merged into the mainline Exim source.
3817 45. Sometimes the final signoff response after QUIT could fail to get
3818 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3819 < 0 before doing a fflush(). This bug looks as though it goes back to the
3820 introduction of TLS in release 3.20, but "sometimes" must have been rare
3821 because the tests only now provoked it.
3823 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3824 (this can affect the format of dates).
3826 47. exim_tidydb, when checking for the continued existence of a message for
3827 which it has found a message-specific retry record, was not finding
3828 messages that were in split spool directories. Consequently, it was
3829 deleting retry records that should have stayed in existence.
3831 48. Steve fixed some bugs in eximstats.
3833 49. The SPA authentication driver was not abandoning authentication and moving
3834 on to the next authenticator when an expansion was forced to fail,
3835 contradicting the general specification for all authenticators. Instead it
3836 was generating a temporary error. It now behaves as specified.
3838 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3839 (the order specifies the preference for clients). The order is now AES256,
3840 AES128, 3DES, ARCFOUR128.
3842 51. Small patch to Sieve code - explicitly set From: when generating an
3845 52. Exim crashed if a remote delivery caused a very long error message to be
3846 recorded - for instance if somebody sent an entire SpamAssassin report back
3847 as a large number of 550 error lines. This bug was coincidentally fixed by
3848 increasing the size of one of Exim's internal buffers (big_buffer) that
3849 happened as part of the Exiscan merge. However, to be on the safe side, I
3850 have made the code more robust (and fixed the comments that describe what
3853 53. Now that there can be additional text after "Completed" in log lines (if
3854 the queue_time_overall log selector is set), a one-byte patch to exigrep
3855 was needed to allow it to recognize "Completed" as not the last thing in
3858 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3859 patch that reportedly fixes this has been added. I am not expert enough to
3860 create a test for it. This is what the patch creator wrote:
3862 "I found a little strange behaviour of ldap code when working with
3863 Windows 2003 AD Domain, where users was placed in more than one
3864 Organization Units. When I tried to give exim partial DN, the exit code
3865 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3866 But simultaneously result of request was absolutely normal ldap result,
3867 so I produce this patch..."
3869 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3870 REFERENCE, so I have modified the code to exclude the patch when that macro
3873 55. Some experimental protocols are using DNS PTR records for new purposes. The
3874 keys for these records are domain names, not reversed IP addresses. The
3875 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3876 leaves it alone. Component reversal etc. now happens only for IP addresses.
3879 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3881 57. Double the size of the debug message buffer (to 2048) so that more of very
3882 long debug lines gets shown.
3884 58. The exicyclog utility now does better if the number of log files to keep
3885 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3887 59. Two changes related to the smtp_active_hostname option:
3889 (1) $smtp_active_hostname is now available as a variable.
3890 (2) The default for smtp_banner uses $smtp_active_hostname instead
3891 of $primary_hostname.
3893 60. The host_aton() function is supposed to be passed a string that is known
3894 to be a valid IP address. However, in the case of IPv6 addresses, it was
3895 not checking this. This is a hostage to fortune. Exim now panics and dies
3896 if the condition is not met. A case was found where this could be provoked
3897 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3898 components; fortuitously, this particular loophole had already been fixed
3899 by change 4.50/55 above.
3901 If there are any other similar loopholes, the new check in host_aton()
3902 itself should stop them being exploited. The report I received stated that
3903 data on the command line could provoke the exploit when Exim was running as
3904 exim, but did not say which command line option was involved. All I could
3905 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3906 running as the user.
3909 61. There was a buffer overflow vulnerability in the SPA authentication code
3910 (which came originally from the Samba project). I have added a test to the
3911 spa_base64_to_bits() function which I hope fixes it.
3914 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3915 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3917 63. The daemon start-up calls getloadavg() while still root for those OS that
3918 need the first call to be done as root, but it missed one case: when
3919 deliver_queue_load_max is set with deliver_drop_privilege. This is
3920 necessary for the benefit of the queue runner, because there is no re-exec
3921 when deliver_drop_privilege is set.
3923 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3924 This has been fixed.
3926 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3927 was in use, was not putting the data itself into the right store pool;
3928 consequently, it could be overwritten for a subsequent message in the same
3929 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3932 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3934 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3935 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3936 the function were treating the return as a boolean value, which happened to
3937 work because 0=false and not-0=true, but is not correct code.
3939 68. The host_aton() function was not handling scoped IPv6 addresses (those
3940 with, for example, "%eth0" on the end) correctly.
3942 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3943 negated items (that is, ~something) in unsigned ints. Some compilers
3944 apparently mutter when there is no cast.
3946 70. If an address verification called from an ACL failed, and did not produce a
3947 user-specific message (i.e. there was only a "system" message), nothing was
3948 put in $acl_verify_message. In this situation, it now puts the system
3951 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3952 session; change 4.31/43 added the unwanted input to the log line - except
3953 that it did not do this in the start of session case. It now does.
3955 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3956 This is wrong and can cause the other end to generate a synchronization
3957 error if it is another Exim or anything else that does the synchronization
3958 check. A QUIT command is no longer sent after a timeout.
3960 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3961 during host lookups.
3963 74. The defer_ok option of callout verification was not working if it was used
3964 when verifying addresses in header lines, that is, for this case:
3966 verify = header_sender/callout=defer_ok
3968 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3969 those file descriptors could be used for SMTP connections. If anything
3970 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3971 could be sent to the SMTP client, causing chaos. The daemon now opens
3972 stdin, stdout, and stderr to /dev/null when it puts itself into the
3975 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3976 log by default. The user can override this with suitable Perl magic.
3978 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3979 the log message when discard triggers, was not working for the DATA ACL or
3980 for the non-SMTP ACL.
3982 78. Error message wording change in sieve.c.
3984 79. If smtp_accept_max_per_host was set, the number of connections could be
3985 restricted to fewer than expected, because the daemon was trying to set up
3986 a new connection before checking whether the processes handling previous
3987 connections had finished. The check for completed processes is now done
3988 earlier. On busy systems, this bug wouldn't be noticed because something
3989 else would have woken the daemon, and it would have reaped the completed
3992 80. If a message was submitted locally by a user whose login name contained one
3993 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3994 It caused a spool format error. I have fixed the spool reading code. A
3995 related problem was that the "from" clause in the Received: line became
3996 illegal because of the space(s). It is now covered by ${quote_local_part.
3998 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4001 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4003 83. Patch from Sieve maintainer to fix unterminated string problem in
4004 "vacation" handling.
4006 84. Some minor changes to the Linux configuration files to help with other
4007 OS variants using glibc.
4009 85. One more patch for Sieve to update vacation handling to latest spec.
4012 ----------------------------------------------------
4013 See the note above about the 4.44 and 4.50 releases.
4014 ----------------------------------------------------
4020 1. Change 4.43/35 introduced a bug that caused file counts to be
4021 incorrectly computed when quota_filecount was set in an appendfile
4024 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4025 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4028 3. Give more explanation in the error message when the command for a transport
4029 filter fails to execute.
4031 4. There are several places where Exim runs a non-Exim command in a
4032 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4033 was being done only for the command run by the queryprogram router. It is
4034 now done for all such subprocesses. The other cases are: ${run, transport
4035 filters, and the commands run by the lmtp and pipe transports.
4037 5. Some older OS have a limit of 256 on the maximum number of file
4038 descriptors. Exim was using setrlimit() to set 1000 as a large value
4039 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4040 systems. I've change it so that if it can't get 1000, it tries for 256.
4042 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4043 was an oversight, and furthermore, ever since the addition of extra
4044 controls (e.g. 4.43/32), the checks on when to allow different forms of
4045 "control" were broken. There should now be diagnostics for all cases when a
4046 control that does not make sense is encountered.
4048 7. $recipients is now available in the predata ACL (oversight).
4050 8. Tidy the search cache before the fork to do a delivery from a message
4051 received from the command line. Otherwise the child will trigger a lookup
4052 failure and thereby defer the delivery if it tries to use (for example) a
4053 cached ldap connection that the parent has called unbind on.
4055 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4056 of $address_data from the recipient verification was clobbered by the
4057 sender verification.
4059 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4060 was its contents. (It was OK if the option was not defined at all.)
4062 11. A "Completed" log line is now written for messages that are removed from
4063 the spool by the -Mrm option.
4065 12. $host_address is now set to the target address during the checking of
4066 ignore_target_hosts.
4068 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4069 being passed; this would have caused $sender_host_name to have been used if
4070 matching the list had actually called for a host name (not very likely,
4071 since this list is usually IP addresses). A host name is now passed as
4074 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4075 code that specifies a non-restarting handler (typically sigaction() in
4076 modern systems) in an attempt to fix a rare and obscure crash bug.
4078 15. Narrowed the window for a race in the daemon that could cause it to ignore
4079 SIGCHLD signals. This is not a major problem, because they are used only to
4080 wake it up if nothing else does.
4082 16. A malformed maildirsize file could cause Exim to calculate negative values
4083 for the mailbox size or file count. Odd effects could occur as a result.
4084 The maildirsize information is now recalculated if the size or filecount
4087 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4088 support for a long time. Removed HAVE_SYS_VFS_H.
4090 18. Updated exipick to current release from John Jetmore.
4092 19. Allow an empty sender to be matched against a lookup in an address list.
4093 Previously the only cases considered were a regular expression, or an
4096 20. Exim went into a mad DNS lookup loop when doing a callout where the
4097 host was specified on the transport, if the DNS lookup yielded more than
4100 21. The RFC2047 encoding function was originally intended for short strings
4101 such as real names; it was not keeping to the 75-character limit for
4102 encoded words that the RFC imposes. It now respects the limit, and
4103 generates multiple encoded words if necessary. To be on the safe side, I
4104 have increased the buffer size for the ${rfc2047: expansion operator from
4107 22. Failure to deliver a bounce message always caused it to be frozen, even if
4108 there was an errors_to setting on the router. The errors_to setting is now
4111 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4112 canonical form (fully expanded) before being placed in
4113 $sender_host_address.
4115 24. Updated eximstats to version 1.33
4117 25. Include certificate and key file names in error message when GnuTLS fails
4118 to set them up, because the GnuTLS error message doesn't include the name
4119 of the failing file when there is a problem reading it.
4121 26. Expand error message when OpenSSL has problems setting up cert/key files.
4124 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4125 (this can affect the format of dates).
4127 28. exim_tidydb, when checking for the continued existence of a message for
4128 which it has found a message-specific retry record, was not finding
4129 messages that were in split spool directories. Consequently, it was
4130 deleting retry records that should have stayed in existence.
4132 29. eximstats updated to version 1.35
4133 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4134 1.35 - bugfix such that pie charts by volume are generated correctly
4136 30. The SPA authentication driver was not abandoning authentication and moving
4137 on to the next authenticator when an expansion was forced to fail,
4138 contradicting the general specification for all authenticators. Instead it
4139 was generating a temporary error. It now behaves as specified.
4141 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4142 (the order specifies the preference for clients). The order is now AES256,
4143 AES128, 3DES, ARCFOUR128.
4145 31. Small patch to Sieve code - explicitly set From: when generating an
4148 32. Exim crashed if a remote delivery caused a very long error message to be
4149 recorded - for instance if somebody sent an entire SpamAssassin report back
4150 as a large number of 550 error lines. This bug was coincidentally fixed by
4151 increasing the size of one of Exim's internal buffers (big_buffer) that
4152 happened as part of the Exiscan merge. However, to be on the safe side, I
4153 have made the code more robust (and fixed the comments that describe what
4156 33. Some experimental protocols are using DNS PTR records for new purposes. The
4157 keys for these records are domain names, not reversed IP addresses. The
4158 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4159 leaves it alone. Component reversal etc. now happens only for IP addresses.
4162 34. The host_aton() function is supposed to be passed a string that is known
4163 to be a valid IP address. However, in the case of IPv6 addresses, it was
4164 not checking this. This is a hostage to fortune. Exim now panics and dies
4165 if the condition is not met. A case was found where this could be provoked
4166 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4167 components; fortuitously, this particular loophole had already been fixed
4168 by change 4.50/55 or 4.44/33 above.
4170 If there are any other similar loopholes, the new check in host_aton()
4171 itself should stop them being exploited. The report I received stated that
4172 data on the command line could provoke the exploit when Exim was running as
4173 exim, but did not say which command line option was involved. All I could
4174 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4175 running as the user.
4178 35. There was a buffer overflow vulnerability in the SPA authentication code
4179 (which came originally from the Samba project). I have added a test to the
4180 spa_base64_to_bits() function which I hope fixes it.
4183 36. The daemon start-up calls getloadavg() while still root for those OS that
4184 need the first call to be done as root, but it missed one case: when
4185 deliver_queue_load_max is set with deliver_drop_privilege. This is
4186 necessary for the benefit of the queue runner, because there is no re-exec
4187 when deliver_drop_privilege is set.
4189 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4190 was in use, was not putting the data itself into the right store pool;
4191 consequently, it could be overwritten for a subsequent message in the same
4192 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4195 38. Sometimes the final signoff response after QUIT could fail to get
4196 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4197 < 0 before doing a fflush(). This bug looks as though it goes back to the
4198 introduction of TLS in release 3.20, but "sometimes" must have been rare
4199 because the tests only now provoked it.
4205 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4206 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4207 Now it returns an integer. A number of calls were still expecting a T/F
4208 return. Fortuitously, in all cases, the tests worked in OK situations,
4209 which is the norm. However, things would have gone wrong on any write
4210 failures on the smtp file descriptor. This function is used when sending
4211 messages over SMTP and also when doing verify callouts.
4213 2. When Exim is called to do synchronous delivery of a locally submitted
4214 message (the -odf or -odi options), it no longer closes stderr before doing
4217 3. Implemented the mua_wrapper option.
4219 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4221 5. Implemented the functions header_remove(), header_testname(),
4222 header_add_at_position(), and receive_remove_recipient(), and exported them
4225 6. If an ACL "warn" statement specified the addition of headers, Exim already
4226 inserted X-ACL-Warn: at the start if there was no header name. However, it
4227 was not making this test for the second and subsequent header lines if
4228 there were newlines in the string. This meant that an invalid header could
4229 be inserted if Exim was badly configured.
4231 7. Allow an ACL "warn" statement to add header lines at the start or after all
4232 the Received: headers, as well as at the end.
4234 8. Added the rcpt_4xx retry error code.
4236 9. Added postmaster_mailfrom=xxx to callout verification option.
4238 10. Added mailfrom=xxxx to the callout verification option, for verify=
4241 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4242 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4244 12. Inserted some casts to stop certain compilers warning when using pointer
4245 differences as field lengths or precisions in printf-type calls (mostly
4246 affecting debugging statements).
4248 13. Added optional readline() support for -be (dynamically loaded).
4250 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4251 same clock tick as a message's arrival, so that its received time was the
4252 same as the "first fail" time on the retry record, and that message
4253 remained on the queue past the ultimate address timeout, every queue runner
4254 would try a delivery (because it was past the ultimate address timeout) but
4255 after another failure, the ultimate address timeout, which should have then
4256 bounced the address, did not kick in. This was a "< instead of <=" error;
4257 in most cases the first failure would have been in the next clock tick
4258 after the received time, and all would be well.
4260 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4261 being recognized when the domain list was tested by the match_domain
4262 condition in an expansion string.
4264 16. Added the ${str2b64: operator.
4266 17. Exim was always calling setrlimit() to set a large limit for the number of
4267 processes, without checking whether the existing limit was already
4268 adequate. (It did check for the limit on file descriptors.) Furthermore,
4269 errors from getrlimit() and setrlimit() were being ignored. Now they are
4270 logged to the main and panic logs, but Exim does carry on, to try to do its
4271 job under whatever limits there are.
4273 18. Imported PCRE 5.0.
4275 19. Trivial typo in log message " temporarily refused connection" (the leading
4278 20. If the log selector return_path_on_delivery was set and an address was
4279 redirected to /dev/null, the delivery process crashed because it assumed
4280 that a return path would always be set for a "successful" delivery. In this
4281 case, the whole delivery is bypassed as an optimization, and therefore no
4284 21. Internal re-arrangement: the function for sending a challenge and reading
4285 a response while authentication was assuming a zero-terminated challenge
4286 string. It's now changed to take a pointer and a length, to allow for
4287 binary data in such strings.
4289 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4291 23. Exim was not respecting finduser_retries when seeking the login of the
4292 uid under which it was called; it was always trying 10 times. (The default
4293 setting of finduser_retries is zero.) Also, it was sleeping after the final
4294 failure, which is pointless.
4296 24. Implemented tls_on_connect_ports.
4298 25. Implemented acl_smtp_predata.
4300 26. If the domain in control=submission is set empty, Exim assumes that the
4301 authenticated id is a complete email address when it generates From: or
4302 Sender: header lines.
4304 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4305 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4306 chown and chgrp in /bin and hostname in /usr/bin.
4308 28. Exim was keeping the "process log" file open after each use, just as it
4309 does for the main log. This opens the possibility of it remaining open for
4310 long periods when the USR1 signal hits a daemon. Occasional processlog
4311 errors were reported, that could have been caused by this. Anyway, it seems
4312 much more sensible not to leave this file open at all, so that is what now
4315 29. The long-running daemon process does not normally write to the log once it
4316 has entered its main loop, and it closes the log before doing so. This is
4317 so that log files can straightforwardly be renamed and moved. However,
4318 there are a couple of unusual error situations where the daemon does write
4319 log entries, and I had neglected to close the log afterwards.
4321 30. The text of an SMTP error response that was received during a remote
4322 delivery was being truncated at 512 bytes. This is too short for some of
4323 the long messages that one sometimes sees. I've increased the limit to
4326 31. It is now possible to make retry rules that apply only when a message has a
4327 specific sender, in particular, an empty sender.
4329 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4330 it possible to be selective about when SMTP synchronization is enforced.
4332 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4334 32. Implemented hosts_connection_nolog.
4336 33. Added an ACL for QUIT.
4338 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4341 35. Added mailbox_size and mailbox_filecount to appendfile.
4343 36. Added control = no_multiline_responses to ACLs.
4345 37. There was a bug in the logic of the code that waits for the clock to tick
4346 in the case where the clock went backwards by a substantial amount such
4347 that the microsecond fraction of "now" was more than the microsecond
4348 fraction of "then" (but the whole seconds number was less).
4350 38. Added support for the libradius Radius client library this is found on
4351 FreeBSD (previously only the radiusclient library was supported).
4357 1. When certain lookups returned multiple values in the form name=value, the
4358 quoting of the values was not always being done properly. Specifically:
4359 (a) If the value started with a double quote, but contained no whitespace,
4361 (b) If the value contained whitespace other than a space character (i.e.
4362 tabs or newlines or carriage returns) it was not quoted.
4363 This fix has been applied to the mysql and pgsql lookups by writing a
4364 separate quoting function and calling it from the lookup code. The fix
4365 should probably also be applied to nisplus, ibase and oracle lookups, but
4366 since I cannot test any of those, I have not disturbed their existing code.
4368 2. A hit in the callout cache for a specific address caused a log line with no
4369 reason for rejecting RCPT. Now it says "Previous (cached) callout
4370 verification failure".
4372 3. There was an off-by-one bug in the queryprogram router. An over-long
4373 return line was truncated at 256 instead of 255 characters, thereby
4374 overflowing its buffer with the terminating zero. As well as fixing this, I
4375 have increased the buffer size to 1024 (and made a note to document this).
4377 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4378 when Exim is waiting for an SMTP response from a remote server, Exim
4379 restarts its select() call on the socket, thereby resetting its timeout.
4380 This is not a problem when such interrupts are rare. Somebody set up a cron
4381 job to run exiwhat every 2 minutes, which is less than the normal select()
4382 timeout (5 or 10 minutes). This meant that the select() timeout never
4383 kicked in because it was always reset. I have fixed this by comparing the
4384 time when an interrupt arrives with the time at the start of the first call
4385 to select(). If more time than the timeout has elapsed, the interrupt is
4386 treated as a timeout.
4388 5. Some internal re-factoring in preparation for the addition of Sieve
4389 extensions (by MH). In particular, the "personal" test is moved to a
4390 separate function, and given an option for scanning Cc: and Bcc: (which is
4391 not set for Exim filters).
4393 6. When Exim created an email address using the login of the caller as the
4394 local part (e.g. when creating a From: or Sender: header line), it was not
4395 quoting the local part when it contained special characters such as @.
4397 7. Installed new OpenBSD configuration files.
4399 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4400 try to make them clearer.
4402 9. Callout options, other than the timeout value, were being ignored when
4403 verifying sender addresses in header lines. For example, when using
4405 verify = header_sender/callout=no_cache
4407 the cache was (incorrectly) being used.
4409 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4410 only the Cygwin environment.
4412 11. When return_path_on_delivery was set as a log selector, if different remote
4413 addresses in the same message used different return paths and parallel
4414 remote delivery occurred, the wrong values would sometimes be logged.
4415 (Whenever a remote delivery process finished, the return path value from
4416 the most recently started remote delivery process was logged.)
4418 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4419 lines when AUTH and/or TLS are in use. This is the "received protocol"
4420 field. Exim used to use "asmtp" for authenticated SMTP, without any
4421 indication (in the protocol name) for TLS use. Now it follows the RFC and
4422 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4423 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4424 names appear in log lines as well as in Received: header lines.
4426 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4427 extensions, and comparison tests, and to fix some bugs.
4429 14. Changes to the "personal" filter test:
4431 (1) The test was buggy in that it was just doing the equivalent of
4432 "contains" tests on header lines. For example, if a user's address was
4433 anne@some.where, the "personal" test would incorrectly be true for
4435 To: susanne@some.where
4437 This test is now done by extracting each address from the header in turn,
4438 and checking the entire address. Other tests that are part of "personal"
4439 are now done using regular expressions (for example, to check local parts
4440 of addresses in From: header lines).
4442 (2) The list of non-personal local parts in From: addresses has been
4443 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4444 taken from the Sieve specification recommendations.
4446 (3) If the message contains any header line starting with "List-" it is
4447 treated as non-personal.
4449 (4) The test for "circular" in the Subject: header line has been removed
4450 because it now seems ill-conceived.
4452 15. Minor typos in src/EDITME comments corrected.
4454 16. Installed latest exipick from John Jetmore.
4456 17. If headers_add on a router specified a text string that was too long for
4457 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4458 of string_sprintf() is now avoided.
4460 18. $message_body_size was not set (it was always zero) when running the DATA
4461 ACL and the local_scan() function.
4463 19. For the "mail" command in an Exim filter, no default was being set for
4464 the once_repeat time, causing a random time value to be used if "once" was
4465 specified. (If the value happened to be <= 0, no repeat happened.) The
4466 default is now 0s, meaning "never repeat". The "vacation" command was OK
4467 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4468 (I found it when inspecting the code).
4470 20. There is now an overall timeout for performing a callout verification. It
4471 defaults to 4 times the callout timeout, which applies to individual SMTP
4472 commands during the callout. The overall timeout applies when there is more
4473 than one host that can be tried. The timeout is checked before trying the
4474 next host. This prevents very long delays if there are a large number of
4475 hosts and all are timing out (e.g. when the network connections are timing
4476 out). The value of the overall timeout can be changed by specifying an
4477 additional sub-option for "callout", called "maxwait". For example:
4479 verify = sender/callout=5s,maxwait=20s
4481 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4482 to the end before writing, but this should make it even safer).
4484 22. Exim was forgetting that it had advertised PIPELINING for the second and
4485 subsequent messages on an SMTP connection. It was also not resetting its
4486 memory on STARTTLS and an internal HELO.
4488 23. When Exim logs an SMTP synchronization error within a session, it now
4489 records whether PIPELINING has been advertised or not.
4491 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4492 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4493 rather than long int.
4495 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4497 26. Added the never_mail option to autoreply.
4503 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4504 crash if the getsockname() call failed; this can happen if a connection is
4505 closed very soon after it is established. The problem was simply in the
4506 order in which certain operations were done, causing Exim to try to write
4507 to the SMTP stream before it had set up the file descriptor. The bug has
4508 been fixed by making things happen in the correct order.
4514 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4515 before the connection was closed, thus losing the rejection response.
4517 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4518 some early Solaris releases, but causes trouble in current releases where
4519 socklen_t is defined.
4521 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4524 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4527 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4528 that releasing the top of it at the end releases what was used for sub-
4529 expansions (unless the block got too big). However, discard this block if
4530 the first thing is a variable or header, so that we can use its block when
4531 it is dynamic (useful for very large $message_headers, for example).
4533 6. Lookups now cache *every* query, not just the most recent. A new, separate
4534 store pool is used for this. It can be recovered when all lookup caches are
4535 flushed. Lookups now release memory at the end of their result strings.
4536 This has involved some general refactoring of the lookup sources.
4538 7. Some code has been added to the store_xxx() functions to reduce the amount
4539 of flapping under certain conditions.
4541 8. log_incoming_interface used to affect only the <= reception log lines. Now
4542 it causes the local interface and port to be added to several more SMTP log
4543 lines, for example "SMTP connection from", and rejection lines.
4545 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4547 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4549 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4550 could be overwritten at the end of the current message (or the start of a
4551 new message if it was set in a HELO ACL). The value is now preserved for
4552 the duration of the SMTP connection.
4554 12. If a transport had a headers_rewrite setting, and a matching header line
4555 contained an unqualified address, that address was qualified, even if it
4556 did not match any rewriting rules. The underlying bug was that the values
4557 of the flags that permit the existence of unqualified sender and recipient
4558 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4559 non-local messages, and by -bnq for local messages) were not being
4560 preserved with the message after it was received.
4562 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4563 "next input=" as part of the text comprising the host identity instead of
4564 the correct text. The code was using the same buffer for two different
4565 strings. However, depending on which order the printing function evaluated
4566 its arguments, the bug did not always show up. Under Linux, for example, my
4567 test suite worked just fine.
4569 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4570 doesn't work with some older versions of Perl. It has been changed to "my",
4571 which in any case is probably the better facility to use.
4573 15. A really picky compiler found some instances of statements for creating
4574 error messages that either had too many or two few arguments for the format
4577 16. The size of the buffer for calls to the DNS resolver has been increased
4578 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4579 for addresses that have a lot of PTR records. This alleviates a problem; it
4580 does not fully solve it.
4582 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4583 buffer now truncates the list and logs the incident, which is the same
4584 action as happens when Exim is looking up a host name and its aliases.
4585 Previously in this situation something unpredictable would happen;
4586 sometimes it was "internal error: store_reset failed".
4588 18. If a server dropped the connection unexpectedly when an Exim client was
4589 using GnuTLS and trying to read a response, the client delivery process
4590 crashed while trying to generate an error log message.
4592 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4593 string, for example:
4595 warn message = H1: something\nH2: something
4597 the text was added as a single header line from Exim's point of view
4598 though it ended up OK in the delivered message. However, searching for the
4599 second and subsequent header lines using $h_h2: did not work. This has been
4600 fixed. Similarly, if a system filter added multiple headers in this way,
4601 the routers could not see them.
4603 20. Expanded the error message when iplsearch is called with an invalid key to
4604 suggest using net-iplsearch in a host list.
4606 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4607 ACLs are no longer actually imposed (and a message to that effect is
4610 22. If a "gecos" field in a passwd entry contained escaped characters, in
4611 particular, if it contained a \" sequence, Exim got it wrong when building
4612 a From: or a Sender: header from that name. A second bug also caused
4613 incorrect handling when an unquoted " was present following a character
4614 that needed quoting.
4616 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4617 was not being matched caselessly.
4619 24. Arranged for all hyphens in the exim.8 source to be escaped with
4622 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4623 properly. Recipient callout cache records were still being keyed to include
4624 the sender, even when use_sender was set false. This led to far more
4625 callouts that were necessary. The sender is no longer included in the key
4626 when use_sender is false.
4628 26. Added "control = submission" modifier to ACLs.
4630 27. Added the ${base62d: operator to decode base 62 numbers.
4632 28. dnsdb lookups can now access SRV records.
4634 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4635 the configuration file.
4637 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4638 (-v) mode. This makes the output for a verbose queue run more intelligible.
4640 31. Added a use_postmaster feature to recipient callouts.
4642 32. Added the $body_zerocount variable, containing the number of binary zero
4643 bytes in the message body.
4645 33. The time of last modification of the "new" subdirectory is now used as the
4646 "mailbox time last read" when there is a quota error for a maildir
4649 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4651 35. Added +ignore_unknown as a special item in host lists.
4653 36. Code for decoding IPv6 addresses in host lists is now included, even if
4654 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4655 address was recognized as an IP address, but was then not correctly decoded
4656 into binary, causing unexpected and incorrect effects when compared with
4663 1. Very minor rewording of debugging text in manualroute to say "list of
4664 hosts" instead of "hostlist".
4666 2. If verify=header_syntax was set, and a header line with an unqualified
4667 address (no domain) and a large number of spaces between the end of the
4668 name and the colon was received, the reception process suffered a buffer
4669 overflow, and (when I tested it) crashed. This was caused by some obsolete
4670 code that should have been removed. The fix is to remove it!
4672 3. When running in the test harness, delay a bit after writing a bounce
4673 message to get a bit more predictability in the log output.
4675 4. Added a call to search_tidyup() just before forking a reception process. In
4676 theory, someone could use a lookup in the expansion of smtp_accept_max_
4677 per_host which, without the tidyup, could leave open a database connection.
4679 5. Added the variables $recipient_data and $sender_data which get set from a
4680 lookup success in an ACL "recipients" or "senders" condition, or a router
4681 "senders" option, similar to $domain_data and $local_part_data.
4683 6. Moved the writing of debug_print from before to after the "senders" test
4686 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4687 problems for message scanning, either using a data ACL, or using
4688 local_scan() because the Received: header was not generated till after they
4689 were called (in order to set the time as the time of reception completion).
4690 I have revised the way this works. The header is now generated after the
4691 body is received, but before the ACL or local_scan() are called. After they
4692 are run, the timestamp in the header is updated.
4698 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4699 before starting a queue runner without re-exec. This happened only when
4700 deliver_drop_privilege was set or when the Exim user was set to root. The
4701 effect of the bug was that timeouts during subsequent deliveries caused
4702 crashes instead of being properly handled. The handler is now left at its
4703 default (and expected) setting.
4705 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4706 message, again when deliver_drop_privilege is set or Exim is run as root.
4707 The bug described in (1) was not present in this case, but the tidying up
4708 of the other signals was missing. I have made the two cases consistent.
4710 3. The ignore_target_hosts setting on a manualroute router was being ignored
4711 for hosts that were looked up using the /MX notation.
4713 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4716 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4717 operated on the sender address. After changing the $sender_address to <>
4718 for the sender address verify, Exim was re-instated it as the original
4719 (before rewriting) address, but remembering that it had rewritten it, so it
4720 wasn't rewriting it again. This bug also had the effect of breaking the
4721 sender address verification caching when the sender address was rewritten.
4723 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4724 This has been changed so that if the ip literal address matches
4725 ignore_target_hosts, the router declines.
4727 7. Added expansion conditions match_domain, match_address, and match_local_
4728 part (NOT match_host).
4730 8. The placeholder for the Received: header didn't have a length field set.
4732 9. Added code to Exim itself and to exim_lock to test for a specific race
4733 condition that could lead to file corruption when using MBX delivery. The
4734 issue is with the lockfile that is created in /tmp. If this file is removed
4735 after a process has opened it but before that process has acquired a lock,
4736 there is the potential for a second process to recreate the file and also
4737 acquire a lock. This could lead to two Exim processes writing to the file
4738 at the same time. The added code performs the same test as UW imapd; it
4739 checks after acquiring the lock that its file descriptor still refers to
4740 the same named file.
4742 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4743 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4746 11. Added the smtp_active_hostname option. If used, this will typically be made
4747 to depend on the incoming interface address. Because $interface_address is
4748 not set up until the daemon has forked a reception process, error responses
4749 that can happen earlier (such as "too many connections") no longer contain
4752 12. If an expansion in a condition on a "warn" statement fails because a lookup
4753 defers, the "warn" statement is abandoned, and the next ACL statement is
4754 processed. Previously this caused the whole ACL to be aborted.
4756 13. Added the iplsearch lookup type.
4758 14. Added ident_timeout as a log selector.
4760 15. Added tls_certificate_verified as a log selector.
4762 16. Added a global option tls_require_ciphers (compare the smtp transport
4763 option of the same name). This controls incoming TLS connections.
4765 17. I finally figured out how to make tls_require_ciphers do a similar thing
4766 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4767 before starting the TLS session.
4769 18. Tabs are now shown as \t in -bP output.
4771 19. If the log selector return_path_on_delivery was set, Exim crashed when
4772 bouncing a message because it had too many Received: header lines.
4774 20. If two routers both had headers_remove settings, and the first one included
4775 a superfluous trailing colon, the final name in the first list and the
4776 first name in the second list were incorrectly joined into one item (with a
4777 colon in the middle).
4783 1. Added -C and -D options to the exinext utility, mainly to make it easier
4784 to include in the automated testing, but these could be helpful when
4785 multiple configurations are in use.
4787 2. The exinext utility was not formatting the output nicely when there was
4788 an alternate port involved in the retry record key, nor when there was a
4789 message id as well (for retries that were specific to a specific message
4790 and a specific host). It was also confused by IPv6 addresses, because of
4791 the additional colons they contain. I have fixed the IPv4 problem, and
4792 patched it up to do a reasonable job for IPv6.
4794 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4795 delivery, the log line now contains "pipelined" if PIPELINING was used.
4797 4. An SMTP transport process used to panic and die if the bind() call to set
4798 an explicit outgoing interface failed. This has been changed; it is now
4799 treated in the same way as a connect() failure.
4801 5. A reference to $sender_host_name in the part of a conditional expansion
4802 that was being skipped was still causing a DNS lookup. This no longer
4805 6. The def: expansion condition was not recognizing references to header lines
4806 that used bh_ and bheader_.
4808 7. Added the _cache feature to named lists.
4810 8. The code for checking quota_filecount in the appendfile transport was
4811 allowing one more file than it should have been.
4813 9. For compatibility with Sendmail, the command line option
4821 and sets the incoming protocol and host name (for trusted callers). The
4822 host name and its colon can be omitted when only the protocol is to be set.
4823 Note the Exim already has two private options, -pd and -ps, that refer to
4824 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4825 "s", but I don't think that's a major issue.
4827 10. A number of refactoring changes to the code, none of which should affect
4830 (a) The number of logging options was getting close to filling up the
4831 32-bit word that was used as a bit map. I have split them into two classes:
4832 those that are passed in the argument to log_write(), and those that are
4833 only ever tested independently outside of that function. These are now in
4834 separate 32-bit words, so there is plenty of room for expansion again.
4835 There is no change in the user interface or the logging behaviour.
4837 (b) When building, for example, log lines, the code previously used a
4838 macro that called string_cat() twice, in order to add two strings. This is
4839 not really sufficiently general. Furthermore, there was one instance where
4840 it was actually wrong because one of the argument was used twice, and in
4841 one call a function was used. (As it happened, calling the function twice
4842 did not affect the overall behaviour.) The macro has been replaced by a
4843 function that can join an arbitrary number of extra strings onto a growing
4846 (c) The code for expansion conditions now uses a table and a binary chop
4847 instead of a serial search (which was left over from when there were very
4848 few conditions). Also, it now recognizes conditions like "pam" even when
4849 the relevant support is not compiled in: a suitably worded error message is
4850 given if an attempt is made to use such a condition.
4852 11. Added ${time_interval:xxxxx}.
4854 12. A bug was causing one of the ddress fields not to be passed back correctly
4855 from remote delivery subprocesses. The field in question was not being
4856 subsequently used, so this caused to problems in practice.
4858 13. Added new log selectors queue_time and deliver_time.
4860 14. Might have fixed a bug in maildirsizefile handling that threw up
4861 "unexpected character" debug warnings, and recalculated the data
4862 unnecessarily. In any case, I expanded the warning message to give more
4865 15. Added the message "Restricted characters in address" to the statements in
4866 the default ACL that block characters like @ and % in local parts.
4868 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4869 Three changes have been made:
4871 (a) There was a serious bug; a negative response to MAIL caused the whole
4872 recipient domain to be cached as invalid, thereby blocking all messages
4873 to all local parts at the same domain, from all senders. This bug has
4874 been fixed. The domain is no longer cached after a negative response to
4875 MAIL if the sender used is not empty.
4877 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4880 (c) A new callout option, "use_sender" has been added for people who want
4881 the modified behaviour.
4887 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4890 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4891 indeed breaks things for older releases.
4893 3. Added additional logging to the case where there is a problem reading data
4894 from a filter that is running in a subprocess using a pipe, in order to
4895 try to track down a specific problem.
4897 4. Testing facility fudge: when running in the test harness and attempting
4898 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4899 getting "No route to host". Convert this to a timeout.
4901 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4904 6. Some OS don't have socklen_t but use size_t instead. This affects the
4905 fifth argument of getsockopt() amongst other things. This is now
4906 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4907 can be set for individual OS. I have set it for SunOS5, OSF1, and
4908 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4909 some earlier ones do not.
4911 7. Change 4.30/15 was not doing the test caselessly.
4913 8. The standard form for an IPv6 address literal was being rejected by address
4914 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4915 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4916 this, as well as the form without the "IPv6" on the front (but only when
4917 address literals are enabled, of course).
4919 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4921 10. Exim crashed if a message with an empty sender address specified by -f
4922 encountered a router with an errors_to setting. This could be provoked only
4923 by a command such as
4927 where an empty string was supplied; "<>" did not hit this bug.
4929 11. Installed PCRE release 4.5.
4931 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4932 remained set. It is now erased.
4934 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4935 times from message ids (which are base 36 rather than the normal 62).
4937 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4938 were being counted as actual protocol errors, and logged if the log
4939 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4940 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4941 and DATA following a set of rejected RCPTs do not count as protocol errors.
4942 In other words, Exim assumes they were pipelined, though this may not
4943 actually be the case. Of course, in all cases the client gets an
4944 appropriate error code.
4946 15. If a lookup fails in an ACL condition, a message about the failure may
4947 be available; it is used if testing the ACL cannot continue, because most
4948 such messages specify what the cause of the deferral is. However, some
4949 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4950 that caused an old message to be retained and used if a later statement
4951 caused a defer, replacing the real cause of the deferral.
4953 16. If an IP address had so many PTR records that the DNS lookup buffer
4954 was not large enough to hold them, Exim could crash while trying to process
4955 the truncated data. It now detects and logs this case.
4957 17. Further to 4.21/58, another change has been made: if (and only if) the
4958 first line of a message (the first header line) ends with CRLF, a bare LF
4959 in a subsequent header line has a space inserted after it, so as not to
4960 terminate the header.
4962 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4963 unnecessarily, used atoi() instead of strtol(), and didn't check the
4964 termination when getting file sizes from file names by regex.
4966 19. Completely re-implemented the support for maildirsize files, in the light
4967 of a number of problems with the previous contributed implementation
4968 (4.30/29). In particular:
4970 . If the quota is zero, the maildirsize file is maintained, but no quota is
4973 . If the maildir directory does not exist, it is created before any attempt
4974 to write a maildirsize file.
4976 . The quota value in the file is just a cache; if the quota is changed in
4977 the transport, the new value overrides.
4979 . A regular expression is available for excluding directories from the
4982 20. The autoreply transport checks the characters in options that define the
4983 message's headers; it allows continued headers, but it was checking with
4984 isspace() after an embedded newline instead of explicitly looking for a
4987 21. If all the "regular" hosts to which an address was routed had passed their
4988 expiry times, and had not reached their retry times, the address was
4989 bounced, even if fallback hosts were defined. Now Exim should go on to try
4992 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4993 equivalent code in the SMTP transport. Some hosts send humungous responses
4994 to HELO/EHLO, more than 1024 it seems.
4996 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4997 gives compiler warnings in some environments. I've now done it "properly",
5000 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5001 (because of problems with the built-in one) was declared to return uschar *
5002 instead of char *, causing compiler failure.
5004 25. Fixed a file descriptor leak when processing alias/forward files.
5006 26. Fixed a minor format string issue in dbfn.c.
5008 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5010 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5011 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5014 29. When a sender address is verified, it is cached, to save repeating the test
5015 when there is more than one recipient in a message. However, when the
5016 verification involves a callout, it is possible for different callout
5017 options to be set for different recipients. It is too complicated to keep
5018 track of this in the cache, so now Exim always runs a verification when a
5019 callout is required, relying on the callout cache for the optimization.
5020 The overhead is duplication of the address routing, but this should not be
5023 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5024 to be verified with callout=postmaster, and the main callout worked but the
5025 postmaster check failed, the verification correctly failed. However, if a
5026 subsequent RCPT command asked for sender verification *without* the
5027 postmaster check, incorrect caching caused this verification also to fail,
5030 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5031 it was not caching the DNS options (qualify_single, search_parents) that
5032 were used when the lookup failed. A subsequent lookup with different
5033 options therefore always gave the same answer, though there were cases
5034 where it should not have. (Example: a "domains = !$mx_any" option on a
5035 dnslookup router: the "domains" option is always processed without any
5036 widening, but the router might have qualify_single set.) Now Exim uses the
5037 cached value only when the same options are set.
5039 32. Added John Jetmore's "exipick" utility to the distribution.
5041 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5042 than a timeout (e.g. a certificate is required, and is not provided), an
5043 Exim server now closes the connection immediately. Previously it waited for
5044 the client to close - but if the client is SSL, it seems that they each
5045 wait for each other, leading to a delay before one of them times out.
5047 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5048 maintained 0.8.x compatibility because I don't think many are using it, and
5049 it is clearly obsolete.
5051 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5054 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5055 client certificate was expired. A simple patch fixes this, though I don't
5056 understand the full logic of why the verify callback is called multiple
5059 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5060 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5061 which causes problems with some clients (such as the Certicom SSL Plus
5062 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5063 disables the coutermeasure allowing Eudora to connect."
5065 38. Exim was not checking that a write() to a log file succeeded. This could
5066 lead to Bad Things if a log got too big, in particular if it hit a file
5067 size limit. Exim now panics and dies if it cannot write to a log file, just
5068 as it does if it cannot open a log file.
5070 39. Modified OS/Makefile-Linux so that it now contains
5072 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5074 The two -D definitions ensure that Exim is compiled with large file
5075 support, which makes it possible to handle log files that are bigger than
5078 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5079 instance) a domain was checked against a named list that involved a lookup,
5080 causing $domain_data to be set, then another domain was checked against the
5081 same list, then the first domain was re-checked, the value of $domain_data
5082 after the final check could be wrong. In particular, if the second check
5083 failed, it could be set empty. This bug probably also applied to
5086 41. The strip_trailing_dot option was not being applied to the address given
5087 with the -f command-line option.
5089 42. The code for reading a message's header from the spool was incrementing
5090 $received_count, but never initializing it. This meant that the value was
5091 incorrect (doubled) while delivering a message in the same process in which
5092 it was received. In the most common configuration of Exim, this never
5093 happens - a fresh exec is done - but it can happen when
5094 deliver_drop_privilege is set.
5096 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5097 it now includes up to 150 characters of the unexpected data in the log
5100 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5101 and building data strings. The size of both of these buffers was 10 000
5102 bytes - far larger than anybody would *ever* want, thought I. Needless to
5103 say, somebody hit the limit. I have increased the maximum line length to
5104 20 000 and the maximum data length of concatenated lines to 100 000. I have
5105 also fixed two bugs, because there was no checking on these buffers. Tsk,
5106 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5107 buffer is too small.
5109 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5110 lsearch lookups. Now it does.
5112 46. When parsing a route_list item in a manualroute router, a fixed-length
5113 buffer was used for the list of hosts. I made this 1024 bytes long,
5114 thinking that nobody would ever have a list of hosts that long. Wrong.
5115 Somebody had a whole pile of complicated expansion conditions, and the
5116 string was silently truncated, leading to an expansion error. It turns out
5117 that it is easier to change to an unlimited length (owing to other changes
5118 that have happened since this code was originally written) than to build
5119 structure for giving a limitation error. The length of the item that
5120 expands into the list of hosts is now unlimited.
5122 47. The lsearch lookup could not handle data where the length of text line was
5123 more than 4095 characters. Such lines were truncated, leading to shortened
5124 data being returned. It should now handle lines of any length.
5126 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5127 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5130 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5131 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5132 no longer happens in comment lines. A list of replacements is now placed
5133 at the head of all of the source files, except those whose only change is
5134 to replace PERL_COMMAND in the very first #! line.
5136 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5137 messages on the queue, with a bottom-up merge sort, using code contributed
5138 by Michael Haardt. This should make operations like -bp somewhat faster on
5139 large queues. It won't affect queue runners, except when queue_run_in_order
5142 51. Installed eximstats 1.31 in the distribution.
5144 52. Added support for SRV lookups to the dnslookup router.
5146 53. If an ACL referred to $message_body or $message_body_end, the value was not
5147 reset for any messages that followed in the same SMTP session.
5149 54. The store-handling optimization for building very long strings was not
5150 differentiating between the different store pools. I don't think this
5151 actually made any difference in practice, but I've tidied it.
5153 55. While running the routers to verify a sender address, $sender_address
5154 was still set to the sender address. This is wrong, because when routing to
5155 send a bounce to the sender, it would be empty. Therefore, I have changed
5156 it so that, while verifying a sender address, $sender_address is set to <>.
5157 (There is no change to what happens when verifying a recipient address.)
5159 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5160 target A or AAAA records (if not already returned) without resetting the
5161 qualify_single or search_parents options of the DNS resolver. These are
5162 inappropriate in this case because the targets of MX and SRV records must
5163 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5164 target that, when qualified, matched something in the local domain. These
5165 two options are now turned off when doing these lookups.
5167 57. It seems that at least some releases of Reiserfs (which does not have the
5168 concept of a fixed number of inodes) returns zero and not -1 for the
5169 number of available inodes. This interacted badly with check_spool_inodes,
5170 which assumed that -1 was the "no such thing" setting. What I have done is
5171 to check that the total number of inodes is greater than zero before doing
5172 the test of how many are available.
5174 58. When a "warn" ACL statement has a log_message modifier, the message is
5175 remembered, and not repeated. This is to avoid a lot of repetition when a
5176 message has many recipients that cause the same warning to be written.
5177 Howewer, Exim was preserving the list of already written lines for an
5178 entire SMTP session, which doesn't seem right. The memory is now reset if a
5179 new message is started.
5181 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5182 debugging output unless log_rewrite was also set.
5184 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5185 of (int)(handle) when we know that handle contains (void *)(-1).
5187 61. The Exim daemon panic-logs an error return when it closes the incoming
5188 connection. However "connection reset by peer" seems to be common, and
5189 isn't really an error worthy of noting specially, so that particular error
5192 62. When Exim is trying to find all the local interfaces, it used to panic and
5193 die if the ioctl to get the interface flags failed. However, it seems that
5194 on at least one OS (Solaris 9) it is possible to have an interface that is
5195 included in the list of interfaces, but for which you get a failure error
5196 for this call. This happens when the interface is not "plumbed" into a
5197 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5198 failure of the "get flags" call assumes that the interface is down.
5200 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5201 makes life easier for people who are doing arithmetic on fields extracted
5202 from dates, where you often get leading zeros that should not be
5203 interpreted as octal.
5205 64. Added qualify_domain to the redirect router, to override the global
5208 65. If a pathologically long header line contained very many addresses (the
5209 report of this problem mentioned 10 000) and each of them was rewritten,
5210 Exim could use up a very large amount of memory. (It kept on making new
5211 copies of the header line as it rewrote, and never released the old ones.)
5212 At the expense of a bit more processing, the header rewriting function has
5213 been changed so that it no longer eats memory in this way.
5215 66. The generation of the Received: header has been moved from the time that a
5216 message starts to be received, to the time that it finishes. The timestamp
5217 in the Received: header should now be very close to that of the <= log
5218 line. There are two side-effects of this change:
5220 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5221 logged header lines no longer include the local Received: line, because
5222 it has not yet been created. The same applies to a copy of the message
5223 that is returned to a non-SMTP sender when a message is rejected.
5225 (b) When a filter file is tested using -bf, no additional Received: header
5226 is added to the test message. After some thought, I decided that this
5229 This change does not affect the value of $received_for. It is still set
5230 after address rewriting, but before local_scan() is called.
5232 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5234 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5235 gave an unhelpful panic error message, and a defer error. I have managed to
5236 change this behaviour so that it now rejects any supplied certificate,
5237 which seems right, as the list of acceptable certificates is empty.
5239 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5240 gave an unhelpful defer error. I have not managed to make this reject any
5241 supplied certificates, but the error message it gives is "no certificate
5242 supplied", which is not helpful.
5244 70. exigrep's output now also includes lines that are not associated with any
5245 message, but which match the given pattern. Implemented by a patch from
5246 Martin Sluka, which also tidied up the Perl a bit.
5248 71. Recipient callout verification, like sender verification, was using <> in
5249 the MAIL FROM command. This isn't really the right thing, since the actual
5250 sender may affect whether the remote host accepts the recipient or not. I
5251 have changed it to use the actual sender in the callout; this means that
5252 the cache record is now keyed on a recipient/sender pair, not just the
5253 recipient address. There doesn't seem to be a real danger of callout loops,
5254 since a callout by the remote host to check the sender would use <>.
5255 [SEE ABOVE: changed after hitting problems.]
5257 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5258 temporary errors. However, in the case of such a code being given after
5259 the end of a data transmission (i.e. after ".") Exim was failing to write
5260 a retry record for the message. (Yes, there was some broken host that was
5261 actually sending 8xx at this point.)
5263 73. An unknown lookup type in a host list could cause Exim to panic-die when
5264 the list was checked. (An example that provoked this was putting <; in the
5265 middle of a list instead of at the start.) If this happened during a DATA
5266 ACL check, a -D file could be left lying around. This kind of configuration
5267 error no longer causes Exim to die; instead it causes a defer errror. The
5268 incident is still logged to the main and panic logs.
5270 74. Buglet left over from Exim 3 conversion. The message "too many messages
5271 in one connection" was written to the rejectlog but not the mainlog, except
5272 when address rewriting (yes!) was being logged.
5274 75. Added write_rejectlog option.
5276 76. When a system filter was run not as root (that is, when system_filter_user
5277 was set), the values of the $n variables were not being returned to the
5278 main process; thus, they were not subsequently available in the $sn
5281 77. Added +return_path_on_delivery log selector.
5283 78. A connection timeout was being treated differently from recipients deferred
5284 when testing hosts_max_try with a message that was older than the host's
5285 retry timeout. (The host should not be counted, thus allowing all hosts to
5286 be tried at least once before bouncing.) This may have been the cause of an
5287 occasionally reported bug whereby a message would remain on the queue
5288 longer than the retry timeout, but would be bounced if a delivery was
5289 forced. I say "may" because I never totally pinned down the problem;
5290 setting up timeout/retry tests is difficult. See also the next item.
5292 79. The ultimate address timeout was not being applied to errors that involved
5293 a combination of host plus message (for example, a timeout on a MAIL
5294 command). When an address resolved to a number of possible hosts, and they
5295 were not all tried for each delivery (e.g. because of hosts_max_try), a
5296 message could remain on the queue longer than the retry timeout.
5298 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5301 81. Fixed an obscure SMTP outgoing bug which required at least the following
5302 conditions: (a) there was another message waiting for the same server;
5303 (b) the server returned 5xx to all RCPT commands in the first message so
5304 that the message was not completed; (c) the server dropped the connection
5305 or gave a negative response to the RSET that Exim sends to abort the
5306 transaction. The observed case was a dropped connection after DATA that had
5307 been sent in pipelining mode. That is, the server had advertised PIPELINING
5308 but was not implementing it correctly. The effect of the bug was incorrect
5309 behaviour, such as trying another host, and this could lead to a crash.
5315 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5316 and daemon.c were passed as pointers to ints; they should have been
5317 pointers to socklen_t variables (which are typically unsigned ints).
5319 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5322 3. Fixed a really odd bug that affected only the testing scheme; patching a
5323 certain fixed string in the binary changed the value of another string that
5324 happened to be identical to the end of the original first string.
5326 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5327 name", it returns that address as the IP address. On some operating
5328 systems (e.g. Solaris), it also passes back the IP address string as the
5329 "host name". However, on others (e.g. Linux), it passes back an empty
5330 string. Exim wasn't checking for this, and was changing the host name to an
5331 empty string, assuming it had been canonicized.
5333 5. Although rare, it is permitted to have more than one PTR record for a given
5334 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5335 all the names associated with an address, because they do in Solaris.
5336 However, it seems that they do not in Linux for data that comes from the
5337 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5338 I found this out when I moved to a new Linux workstation and tried to run
5339 the Exim test suite.
5341 To get round this problem I have changed the code so that it now does its
5342 own call to the DNS to look up PTR records when searching for a host name.
5343 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5344 addresses that are only in /etc/hosts are still found.
5346 This behaviour is, however, controlled by an option called host_lookup_
5347 order, which defaults to "bydns:byaddr". If people want to use the other
5348 order, or indeed, just use one or the other means of lookup, they can
5349 specify it in this variable.
5351 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5352 some operating systems, this comes back from gethostbyaddr() as an empty
5353 string, and this is what Exim used to test for. However, it seems that in
5354 other systems, "." is yielded. Exim now tests for this case too.
5356 7. The values of check_spool_space and check_log_space are now held internally
5357 as a number of kilobytes instead of an absolute number of bytes. If a
5358 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5359 kilobyte. This means that much larger values can be stored.
5361 8. Exim monitor: an attempt to get the action menu when not actually pointing
5362 at a message produces an empty menu entitled "No message selected". This
5363 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5364 no entries in it ("Shell widget menu has zero width and/or height"). So I
5365 have added a single, blank menu entry in this case.
5367 9. Added ${quote_local_part.
5369 10. MIME decoding is now applied to the contents of Subject: header lines when
5372 11. Now that a reference to $sender_host_address automatically causes a reverse
5373 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5374 host lookup before query-style lookups in lists that might use this
5375 variable. This has therefore been abolished, and the "net-" prefix is no
5376 longer necessary for query-style lookups.
5378 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5379 have been a typo for LFLAGS, so it has been changed.
5381 13. The install script calls Exim with "-C /dev/null" in order to find the
5382 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5383 to be output. Howeve, since Exim outputs its version number before the
5384 error, it didn't break the script. It just looked ugly. I fixed this by
5385 always allowing "-C /dev/null" if the caller is root.
5387 14. Ignore overlarge ACL variable number when reading spool file - insurance
5388 against a later release with more variables having written the file.
5390 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5391 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5392 this, as well as the form without the "IPv6" on the front.
5394 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5395 OS/Makefile-Darwin file.
5397 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5398 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5400 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5401 variable. However, some OS use ints for inodes. I've added cast to long int
5402 to get rid of the compiler warning.
5404 19. I had forgotten to lock out "/../" in configuration file names when
5405 ALT_CONFIG_PREFIX was set.
5407 20. Routers used for verification do not need to specify transports. However,
5408 if such a router generated a host list, and callout was configured, Exim
5409 crashed, because it could not find a port number from the (non-existent)
5410 transport. It now assumes port 25 in this circumstance.
5412 21. Added the -t option to exigrep.
5414 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5415 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5416 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5417 them is set, LOOKUP_LSEARCH is forced.
5419 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5421 24. Added sender and host information to the "rejected by local_scan()" log
5422 line; previously there was no indication of these.
5424 25. Added .include_if_exists.
5426 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5427 Linux. It turns out that not all file systems support this. Apparently some
5428 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5429 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5430 not supported on the file descriptor, is now ignored when Exim is trying to
5431 sync a directory. This applies only to Linux.
5433 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5435 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5436 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5439 29. Added support for maildirsize files from supplied patch (modified a bit).
5441 30. The use of :fail: followed by an empty string could lead Exim to respond to
5442 sender verification failures with (e.g.):
5444 550 Verification failed for <xxx>
5445 550 Sender verify failed
5447 where the first response line was missing the '-' that indicates it is not
5448 the final line of the response.
5450 31. The loop for finding the name of the user that called Exim had a hardwired
5451 limit of 10; it now uses the value of finduser_retries, which is used for
5452 all other user lookups.
5454 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5457 33. Exim was neglecting to zero errno before one call of strtol() when
5458 expanding a string and expecting an integer value. On some systems this
5459 resulted in spurious "integer overflow" errors. Also, it was casting the
5460 result into an int without checking.
5462 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5463 did not work. The code looks as if it has *never* worked, though it appears
5464 to have been documented since at least releast 1.62. I have made it work.
5466 35. The "timeout_DNS" error in retry rules, also documented since at least
5467 1.62, also never worked. As it isn't clear exactly what this means, and
5468 clearly it isn't a major issue, I have abolished the feature by treating it
5469 as "timeout", and writing a warning to the main and panic logs.
5471 36. The display of retry rules for -brt wasn't always showing the error code
5474 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5475 timeout_connect_A, timeout_connect_MX.
5477 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5478 to the empty sender.
5480 39. The daemon was not analysing the content of -oX till after it had closed
5481 stderr and disconnected from the controlling terminal. This meant that any
5482 syntax errors were only noted on the panic log, and the return code from
5483 the command was 0. By re-arranging the code a little, I've made the
5484 decoding happen first, so such errors now appear on stderr, and the return
5485 code is 1. However, the actual setting up of the sockets still happens in
5486 the disconnected process, so errors there are still only recorded on the
5489 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5490 connections (as happens on some IP stacks) was logged at start up time as
5491 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5492 it from "IPv6 and IPv4", which means that two separate sockets are being
5495 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5496 says whether AF_INET or AF_INET6 was passed as an argument.
5498 42. Exiwhat output was messed up when time zones were included in log
5501 43. Exiwhat now gives more information about the daemon's listening ports,
5502 and whether -tls-on-connect was used.
5504 44. The "port" option of the smtp transport is now expanded.
5506 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5507 silently ignored. Now an error message is written to the main and panic
5510 46. There's a new ACL modifier called "logwrite" which writes to a log file
5511 as soon as it is encountered.
5513 47. Added $local_user_uid and $local_user_gid at routing time.
5515 48. Exim crashed when trying to verify a sender address that was being
5518 49. Exim was recognizing only a space character after ".include". It now also
5519 recognizes a tab character.
5521 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5522 extracting the relevant information from the specification. The man page no
5523 longer contains scrambled data for the -d option, and I've added a section
5524 at the front about calling Exim under different names.
5526 51. Added "extra_headers" argument to the "mail" command in filter files.
5528 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5531 53. Installed eximstats 1.29.
5533 54. Added transport_filter_timeout as a generic transport option.
5535 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5536 Cc: header lines. This was required by RFC 822, but it not required by RFC
5539 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5540 incoming messages that did not have them. Now it does so only if the
5541 message originates locally, that is, if there is no associated remote host
5542 address. When Resent- header lines are present, this applies to the Resent-
5543 lines rather than the non-Resent- lines.
5545 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5546 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5548 58. Messages for configuration errors now include the name of the main
5549 configuration file - useful now that there may be more than one file in a
5550 list (.included file names were always shown).
5552 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5553 for those rare installations that do not start the daemon as root or run it
5554 setuid root. I've cut out the call to initgroups() if the daemon is not
5557 60. The Exim user and group can now be bound into the binary as text strings
5558 that are looked up at the start of Exim's processing.
5560 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5562 62. Added $mailstore_basename variable.
5564 63. Installed patch to sieve.c from Michael Haardt.
5566 64. When Exim failed to open the panic log after failing to open the main log,
5567 the original message it was trying to log was written to stderr and debug
5568 output, but if they were not available (the usual case in production), it
5569 was lost. Now it is written to syslog before the two lines that record the
5570 failures to open the logs.
5572 65. Users' Exim filters run in subprocesses under the user's uid. It is
5573 possible for a "deliver" command or an alias in a "personal" command to
5574 provoke an address rewrite. If logging of address rewriting is configured,
5575 this fails because the process is not running as root or exim. There may be
5576 a better way of dealing with this, but for the moment (because 4.30 needs
5577 to be released), I have disabled address rewrite logging when running a
5578 filter in a non-root, non-exim process.
5584 1. The buildconfig auxiliary program wasn't quoting the value set for
5585 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5586 not defined. This bug was masked in 4.22 by the effect that was fixed in
5589 2. Some messages that were rejected after a message id was allocated were
5590 shown as "incomplete" by exigrep. It no longer does this for messages that
5591 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5593 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5594 have allow_domain_literals set, the ID did not get logged in the <= line.
5595 Domain literals are now always recognized in Message-ID: header lines.
5597 4. The first argument for a ${extract expansion item is the key name or field
5598 number. Leading and trailing spaces in this item were not being ignored,
5599 causing some misleading effects.
5601 5. When deliver_drop_privilege was set, single queue runner processes started
5602 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5603 same command in the process it spins off) were not dropping privilege.
5605 6. When the daemon running as "exim" started a queue runner, it always
5606 re-executed Exim in the spun-off process. This is a waste of effort when
5607 deliver_drop_privilege is set. The new process now just calls the
5608 queue-runner function directly.
5614 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5617 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5618 ignored. Though the use of -oP was forcing the writing of a pid file, it
5619 was always written to the default place.
5621 3. If the message "no IP address found for host xxxx" is generated during
5622 incoming verification, it is now followed by identification of the incoming
5623 connection (so you can more easily find what provoked it).
5625 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5627 5. Added some features to "harden" Exim a bit more against certain attacks:
5629 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5630 be put in Local/Makefile. This is like the never_users runtime option,
5631 but it cannot be overridden. The default setting is "root".
5633 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5634 prefix string with which any file named in a -C command line option
5637 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5638 is retained for -C and -D only if the caller of Exim is root. Without
5639 it, the exim user may also use -C and -D and retain privilege.
5641 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5642 command line option is disabled.
5644 6. Macro names set by the -D option must start with an upper case letter, just
5645 like macro names defined in the configuration file.
5647 7. Added "dereference=" facility to LDAP.
5649 8. Two instances of the typo "uknown" in the source files are fixed.
5651 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5652 the Configure-Makefile script screwed up while processing it.
5654 10. Incorporated PCRE 4.4.
5656 11. The SMTP synchronization check was not operating right at the start of an
5657 SMTP session. For example, it could not catch a HELO sent before the client
5658 waited for the greeting. There is now a check for outstanding input at the
5659 point when the greeting is written. Because of the duplex, asynchronous
5660 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5661 way, but not yet received, when the check is performed.
5663 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5664 on TCP/IP sockets, because this apparently causes some broken clients to
5667 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5668 unchanged) from the Cygwin maintainer.
5670 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5671 was supported instead of testing for mbx. Effectively a typo.
5673 15. The spa authenticator server code was not checking that the input it
5674 received was valid base64.
5676 16. The debug output line for the "set" modifier in ACLs was not showing the
5677 name of the variable that was being set.
5679 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5681 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5682 lookup on its own. Something else was needed to trigger the lookup. For
5683 example, a match in host_lookup or the need for a host name in a host list.
5684 Now, if $sender_host_name is referenced and the host name has not yet been
5685 looked up, a lookup is performed. If the lookup fails, the variable remains
5686 empty, and $host_lookup_failed is set to "1".
5688 19. Added "eqi" as a case-independent comparison operator.
5690 20. The saslauthd authentication condition could segfault if neither service
5691 nor realm was specified.
5693 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5694 error message that was logged was misleading, and incoming SMTP
5695 connections were dropped. The message is now more accurate, and temporary
5696 errors are given to SMTP connections.
5698 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5699 (except RSET) with a 421 error, until QUIT is received. However, it was
5700 failing to send a response to QUIT.
5702 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5703 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5704 ACL is now run after the helo_try_verify_hosts code.
5706 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5707 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5708 case-independent, so other case variants are also recognized). Apparently
5709 some systems use these upper case variants.
5711 25. If more than two messages were waiting for the same host, and a transport
5712 filter was specified for the transport, Exim sent two messages over the
5713 same TCP/IP connection, and then failed with "socket operation on non-
5714 socket" when it tried to send the third.
5716 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5718 27. The extern definition of crypt16() in expand.c was not being excluded when
5719 the OS had its own crypt16() function.
5721 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5722 as a preferred synonym for return_size_limit, both as an option and as an
5725 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5727 30. Changed the default configuration ACL to relax the local part checking rule
5728 for addresses that are not in any local domains. For these addresses,
5729 slashes and pipe symbols are allowed within local parts, but the sequence
5730 /../ is explicitly forbidden.
5732 31. SPA server authentication was not clearing the challenge buffer before
5735 32. log_message in a "warn" ACL statement was writing to the reject log as
5736 well as to the main log, which contradicts the documentation and doesn't
5737 seem right (because no rejection is happening). So I have stopped it.
5739 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5740 However, I am unable to do any testing of this.
5742 34. Fixed an infelicity in the appendfile transport. When checking directories
5743 for a mailbox, to see if any needed to be created, it was accidentally
5744 using path names with one or more superfluous leading slashes; tracing
5745 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5747 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5748 contents are added to the log line that is written for every discarded
5749 recipient. (Previously a log_message setting was ignored.)
5751 36. The ${quote: operator now quotes the string if it is empty.
5753 37. The install script runs exim in order to find its version number. If for
5754 some reason other than non-existence or emptiness, which it checks, it
5755 could not run './exim', it was installing it with an empty version number,
5756 i.e. as "exim-". This error state is now caught, and the installation is
5759 38. An argument was missing from the function that creates an error message
5760 when Exim fails to connect to the socket for saslauthd authentication.
5761 This could cause Exim to crash, or give a corrupted message.
5763 39. Added isip, isip4, and isip6 to ${if conditions.
5765 40. The ACL variables $acl_xx are now saved with the message, and can be
5766 accessed later in routers, transports, and filters.
5768 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5769 strings in the file are not string-expanded.
5771 42. If a MAIL command specified a SIZE value that was too large to fit into an
5772 int variable, the check against message_size_limit failed. Such values are
5773 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5774 day this will have to be increased, but I don't think I want to be around
5775 when emails are that large.
5782 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5783 iconv() is not standard in FreeBSD.
5785 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5786 IPv6 enabled. The observed symptom was a segmentation fault on return from
5787 the function os_common_find_running_interfaces() in src/os.c.
5789 3. In the check_special_case() function in daemon.c I had used "errno" as an
5790 argument name, which causes warnings on some systems. This was basically a
5791 typo, since it was named "eno" in the comments!
5793 4. The code that waits for the clock to tick (at a resolution of some fraction
5794 of a second) so as to ensure message-id uniqueness was always waiting for
5795 at least one whole tick, when it could have waited for less. [This is
5796 almost certainly not relevant at current processor speeds, where it is
5797 unlikely to ever wait at all. But we try to future-proof.]
5799 5. The function that sleeps for a time interval that includes fractions of a
5800 second contained a race. It did not block SIGALRM between setting the
5801 timer, and suspending (a couple of lines later). If the interval was short
5802 and the sigsuspend() was delayed until after it had expired, the suspension
5803 never ended. On busy systems this could lead to processes getting stuck for
5806 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5807 process, before it forks any delivery processes. The open lookup caching
5808 mechanism meant that the open file or database connection was passed into
5809 the delivery process. The problem was that delivery processes always tidy
5810 up cached lookup data. This could cause a problem for the next delivery
5811 process started by the queue runner, because the external queue runner
5812 process does not know about the closure. So the next delivery process
5813 still has data in the lookup cache. In the case of a file lookup, there was
5814 no problem because closing a file descriptor in a subprocess doesn't affect
5815 the parent. However, if the lookup was caching a connection to a database,
5816 the connection was closed, and the second delivery process was likely to
5817 see errors such as "PGSQL: query failed: server closed the connection
5818 unexpectedly". The problem has been fixed by closing all cached lookups
5819 in a queue runner before running a delivery process.
5821 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5822 seem to have the "const" qualifier which it has on other OS. I've
5825 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5826 *of the same type* (client or server) with the same public name that an
5827 error should be diagnosed.
5829 9. When Exim looked up a host name for an IP address, but failed to find the
5830 original IP address when looking up the host name (a safety check), it
5831 output the message "<ip address> does not match any IP for NULL", which was
5832 confusing, to say the least. The bug was that the host name should have
5833 appeared instead of "NULL".
5835 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5836 user that is built into the binary, and the -C or -D options is used, root
5837 privilege is dropped before the configuration file is read. In addition,
5838 logging is switched to stderr instead of the normal log files. If the
5839 configuration then re-defines the Exim user, the unprivileged environment
5840 is probably not what is expected, so Exim logs a panic warning message (but
5843 However, if deliver_drop_privilege is set, the unprivileged state may well
5844 be exactly what is intended, so the warning has been cut out in that case,
5845 and Exim is allowed to try to write to its normal log files.
5851 1. smtp_return_error_details was not giving details for temporary sender
5852 or receiver verification errors.
5854 2. Diagnose a configuration error if two authenticators have the same public
5857 3. Exim used not to create the message log file for a message until the first
5858 delivery attempt. This could be confusing when incoming messages were held
5859 for policy or load reasons. The message log file is now created at the time
5860 the message is received, and an initial "Received" line is written to it.
5862 4. The automatically generated man page for command line options had a minor
5863 bug that caused no ill effects; however, a more serious problem was that
5864 the procedure for building the man page automatically didn't always
5865 operate. Consequently, release 4.20 contains an out-of-date version. This
5866 shouldn't happen again.
5868 5. When building Exim with embedded Perl support, the script that builds the
5869 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5870 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5872 6. The freeze_tell option was not being used for messages that were frozen on
5873 arrival, either by an ACL or by local_scan().
5875 7. Added the smtp_incomplete_transaction log selector.
5877 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5878 was accepting AUTH without a new EHLO.
5880 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5881 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5884 10. Logging of TCP/IP connections (when configured) now happens in the main
5885 daemon process instead of the child process, so that the TCP/IP connection
5886 count is more accurate (but it can never be perfect).
5888 11. The use of "drop" in a nested ACL was not being handled correctly in the
5889 outer ACL. Now, if condition failure induced by the nested "drop" causes
5890 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5891 or "require"), the connection is dropped.
5893 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5894 that yield "discard" can now be used with an "accept" or a "discard" verb,
5895 but an error is generated for any others (because I can't see a useful way
5896 to define what should happen).
5898 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5899 are now processed in the same way as lines in the Exim configuration file.
5900 In particular, continuation lines are supported.
5902 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5904 15. Added -ti meaning -t -i.
5906 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5907 domains, and warn by logging if others are found.
5909 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5910 returned from GIFCONF when Exim is trying to find the list of interfaces on
5911 a host. The code in os.c has been modified to copy each ifreq to an aligned
5912 structure in all cases.
5914 Also, in some cases, the returned ifreq's were being copied to a 'struct
5915 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5916 means the last couple of bytes of an IPv6 address could be chopped if the
5917 ifreq contained only a normal sockaddr (14 bytes storage).
5919 18. Named domain lists were not supported in the hosts_treat_as_local option.
5920 An entry such as +xxxx was not recognized, and was treated as a literal
5923 19. Ensure that header lines added by a DATA ACL are included in the reject log
5924 if the ACL subsequently rejects the message.
5926 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5927 MD5 (which is deprecated).
5929 21. When testing a filter file using -bf, Exim was writing a message when it
5930 took the sender from a "From " line in the message, but it was not doing so
5931 when it took $return_path from a Return-Path: header line. It now does.
5933 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5934 with a valid header line field name (a series of printing characters
5935 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5937 23. Changed "disc" in the source to "disk" to conform to the documentation and
5938 the book and for uniformity.
5940 24. Ignore Sendmail's -Ooption=value command line item.
5942 25. When execve() failed while trying to run a command in a pipe transport,
5943 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5944 could be confused with a return value of 69 from the command itself. This
5945 has been changed to 127, the value the shell returns if it is asked to run
5946 a non-existent command. The wording for the related log line suggests a
5947 non-existent command as the problem.
5949 26. If received_header_text expands to an empty string, do not add a Received:
5950 header line to the message. (Well, it adds a token one on the spool, but
5951 marks it "old" so that it doesn't get used or transmitted.)
5953 27. Installed eximstats 1.28 (addition of -nt option).
5955 28. There was no check for failure on the call to getsockname() in the daemon
5956 code. This can fail if there is a shortage of resources on the system, with
5957 ENOMEM, for example. A temporary error is now given on failure.
5959 29. Contrary to the C standard, it seems that in some environments, the
5960 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5961 program. Exim now does this explicitly; it affects the formatting of
5962 timestamps using strftime().
5964 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5965 complaints. I've now initialized all the variables, to avoid this.
5967 32. Header lines added by a system filter were not being "seen" during
5968 transport-time rewrites.
5970 33. The info_callback() function passed to OpenSSL is set up with type void
5971 (*)(SSL *, int, int), as described somewhere. However, when calling the
5972 function (actually a macro) that sets it up, the type void(*)() is
5973 expected. I've put in a cast to prevent warnings from picky compilers.
5975 34. If a DNS black list lookup found a CNAME record, but there were no A
5976 records associated with the domain it pointed at, Exim crashed.
5978 35. If a DNS black list lookup returned more than one A record, Exim ignored
5979 all but the first. It now scans all returned addresses if a particular IP
5980 value is being sought. In this situation, the contents of the
5981 $dnslist_value variable are a list of all the addresses, separated by a
5984 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5985 to accept a host name and all its aliases if the forward lookup for any of
5986 them yielded the IP address of the incoming connection. Now it accepts only
5987 those names whose forward lookup yields the correct IP address. Any other
5988 names are discarded. This closes a loophole whereby a rogue DNS
5989 administrator could create reverse DNS records to break through a
5990 wildcarded host restriction in an ACL.
5992 37. If a user filter or a system filter that ran in a subprocess used any of
5993 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5994 the wrong values were passed to the pipe command ($thisaddress had the
5995 value of $0, $0 had the value of $1, etc). This bug was introduced by
5996 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5998 38. Improved the line breaking for long SMTP error messages from ACLs.
5999 Previously, if there was no break point between 40 and 75 characters, Exim
6000 left the rest of the message alone. Two changes have been made: (a) I've
6001 reduced the minimum length to 35 characters; (b) if it can't find a break
6002 point between 35 and 75 characters, it looks ahead and uses the first one
6003 that it finds. This may give the occasional overlong line, but at least the
6004 remaining text gets split now.
6006 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6007 file descriptors might be low, and that setting 1000 would always raise it.
6008 It turns out that in some environments, the limit is already over 1000 and
6009 that lowering it causes trouble. So now Exim takes care not to decrease it.
6011 40. When delivering a message, the value of $return_path is set to $sender_
6012 address at the start of routing (routers may change the value). By an
6013 oversight, this default was not being set up when an address was tested by
6014 -bt or -bv, which affected the outcome if any router or filter referred to
6017 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6018 log only when "message" or "log_message" are set. However, if one of the
6019 conditions was an address verification, or a call to a nested ACL, the
6020 messages generated by the underlying test were being passed through. This
6021 no longer happens. The underlying message is available in $acl_verify_
6022 message for both "message" and "log_message" expansions, so it can be
6023 passed through if needed.
6025 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6026 new expansion $bh_ to give the encoded byte string without charset
6027 translation. Translation happens only if iconv() is available; HAVE_ICONV
6028 indicates this at build time. HEADERS_CHARSET gives the charset to
6029 translate to; headers_charset can change it in the configuration, and
6030 "headers charset" can change it in an individual filter file.
6032 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6033 that creates RFC 2047 encoded "words" labels them as that charset instead
6034 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6035 expansion operator; (ii) when Exim creates a From: line for a local
6036 message; (iii) when a header line is rewritten to include a "phrase" part.
6038 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6039 buggy, causing it to skip the first lines of messages whose message ID
6040 ended in 'D'. This would not have bitten before Exim release 4.14, because
6041 message IDs were unlikely to end in 'D' before then. The effect was to have
6042 incorrect size information for certain domains.
6044 45. #include "config.h" was missing at the start of the crypt16.c module. This
6045 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6048 46. If there was a timeout during a "random" callout check, Exim treated it as
6049 a failure of the random address, and carried on sending RSET and the real
6050 address. If the delay was just some slowness somewhere, the response to the
6051 original RCPT would be taken as a response to RSET and so on, causing
6052 mayhem of various kinds.
6054 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6055 when I implemented it. It didn't allow for the fact that some option values
6056 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6057 the right test for positive values.
6059 48. Domain names in DNS records are case-independent. Exim always looks them up
6060 in lower case. Some resolvers return domain names in exactly the case they
6061 appear in the zone file, that is, they may contain uppercase letters. Not
6062 all resolvers do this - some return always lower case. Exim was treating a
6063 change of case by a resolver as a change of domain, similar to a widening
6064 of a domain abbreviation. This triggered its re-routing code and so it was
6065 trying to route what was effectively the same domain again. This normally
6066 caused routing to fail (because the router wouldn't handle the domain
6067 twice). Now Exim checks for this case specially, and just changes the
6068 casing of the domain that it ultimately uses when it transmits the message
6071 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6074 50. If a filter generated a file delivery with a non-absolute name (possible if
6075 no home directory exists for the router), the forbid_file option was not
6078 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6079 the existing equality matching.
6081 52. Exim was using ints instead of ino_t variables in some places where it was
6082 dealing with inode numbers.
6084 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6085 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6086 called TMPDIR, and if it finds it is different, it changes its value.
6088 54. The smtp_printf() function is now made available to local_scan() so
6089 additional output lines can be written before returning. There is also an
6090 smtp_fflush() function to enable the detection of a dropped connection.
6091 The variables smtp_input and smtp_batched_input are exported to
6094 55. Changed the default runtime configuration: the message "Unknown user"
6095 has been removed from the ACL, and instead placed on the localuser router,
6096 using the cannot_route_message feature. This means that any verification
6097 failures that generate their own messages won't get overridden. Similarly,
6098 the "Unrouteable address" message that was in the ACL for unverifiable
6099 relay addresses has also been removed.
6101 56. Added hosts_avoid_esmtp to the smtp transport.
6103 57. The exicyclog script was not checking for the esoteric option
6104 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6105 will work only if exicyclog is run under the appropriate euid.
6107 58. Following a discussion on the list, the rules by which Exim recognises line
6108 endings on incoming messages have been changed. The -dropcr and drop_cr
6109 options are now no-ops, retained only for backwards compatibility. The
6110 following line terminators are recognized: LF CRLF CR. However, special
6111 processing applies to CR:
6113 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6114 nor a local message in the state where . is a terminator.
6116 (ii) If a bare CR is encountered in a header line, an extra space is added
6117 after the line terminator so as not to end the header. The reasoning
6118 behind this is that bare CRs in header lines are most likely either
6119 to be mistakes, or people trying to play silly games.
6121 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6122 was being incorrectly given as 18 bytes larger than it should have been.
6123 This is a VOB (very old bug).
6125 60. This may never have affected anything current, but just in case it has:
6126 When the local host is found other than at the start of a list of hosts,
6127 the local host, those with the same MX, and any that follow, are discarded.
6128 When the list in question was part of a longer list of hosts, the following
6129 hosts (not currently being processed) were also being discarded. This no
6130 longer happens. I'm not sure if this situation could ever has previously
6133 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6136 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6137 header. This is something that is recommended in a new Internet Draft, and
6138 is something that is documented as being done by Sendmail. There are two
6139 possible values. For messages generated by the autoreply transport, Exim
6142 Auto-Submitted: auto-replied
6144 whereas for all other generated messages (e.g. bounces) it adds
6146 Auto-Submitted: auto-generated
6148 63. The "personal" condition in filters now includes a test for the
6149 Auto-Submitted: header. If it contains the string "auto-" the message it
6150 not considered personal.
6152 64. Added rcpt_include_affixes as a generic transport option.
6154 65. Added queue_only_override (default true).
6156 66. Added the syslog_duplication option.
6158 67. If what should have been the first header line of a message consisted of
6159 a space followed by a colon, Exim was mis-interpreting it as a header line.
6160 It isn't of course - it is syntactically invalid and should therefore be
6161 treated as the start of the message body. The misbehaviour could have
6162 caused a number of strange effects, including loss of data in subsequent
6163 header lines, and spool format errors.
6165 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6166 client host had authenticated. This control can now be exercised by an ACL
6167 for more flexibility.
6169 69. By default, callouts do not happen when testing with -bh. There is now a
6170 variant, -bhc, which does actually run the callout code, including
6171 consulting and updating the callout cache.
6173 70. Added support for saslauthd authentication, courtesy of Alexander
6176 71. If statvfs() failed on the spool or log directories while checking their
6177 size for availability, Exim confusingly gave the error "space shortage".
6178 Furthermore, in debugging mode it crashed with a floating point exception.
6179 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6180 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6181 serious problem, Exim now writes to the main and panic logs when this
6182 happens, with details of the failure. It then refuses to accept the
6183 incoming message, giving the message "spool directory problem" or "log
6184 directory problem" with a 421 code for SMTP messages.
6186 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6187 0, 1, and 2 exist, because some OS complain for execs without them (see
6188 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6189 descriptors. However, the code omitted to check that the open succeeded,
6190 causing mysterious errors if for some reason the permissions on /dev/null
6191 got screwed. Now Exim writes a message to the main and panic logs, and
6192 bombs out if it can't open /dev/null.
6194 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6195 interact so that it is all more flexible. It is supposed to remain
6196 backwards compatible. Also added extra_local_interfaces.
6198 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6199 to bomb out with an assertion failure - to the client this appears as a
6200 connection drop. This problem occurs in the part of the code that was taken
6201 from the Samba project. Fortunately, the assertion is in a very simple
6202 function, so I have fixed this by reproducing the function inline in the
6203 one place where it is called, and arranging for authentication to fail
6204 instead of killing the process with assert().
6206 75. The SPA client code was not working when the server requested OEM rather
6207 than Unicode encoding.
6209 76. Added code to make require_files with a specific uid setting more usable in
6210 the case where statting the file as root fails - usually a non-root-mounted
6211 NFS file system. When this happens and the failure is EACCES, Exim now
6212 forks a subprocess and does the per-uid checking as the relevant uid.
6214 77. Added process_log_path.
6216 78. If log_file_path was not explicitly set, a setting of check_log_space or
6217 check_log_inodes was ignored.
6219 79. If a space check for the spool or log partitions fails, the incident is now
6220 logged. Of course, in the latter case the data may get lost...
6222 80. Added the %p formatting code to string_format() so that it can be used to
6223 print addresses in debug_print(). Adjusted all the address printing in the
6224 debugging in store.c to use %p rather than %d.
6226 81. There was a concern that a line of code in smtp_in.c could overflow a
6227 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6228 initially expressed, the concern was not well-founded, because trailing
6229 spaces are removed early. However, if the trailing spaces were followed by
6230 a NULL, they did not get removed, so the overflow was possible. Two fixes
6233 (a) I re-wrote the offending code in a cleaner fashion.
6234 (b) If an incoming SMTP command contains a NULL character, it is rejected
6237 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6238 runs initgroups(), so that if the Exim user is in any additional groups,
6239 they will be used during message reception.
6245 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob