1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
90 TL/01 When running the test suite, the README says that variables such as
91 no_msglog_check are global and can be placed anywhere in a specific
92 test's script, however it was observed that placement needed to be near
93 the beginning for it to behave that way. Changed the runtest perl
94 script to read through the entire script once to detect and set these
95 variables, reset to the beginning of the script, and then run through
96 the script parsing/test process like normal.
98 TL/02 The BSD's have an arc4random API. One of the functions to induce
99 adding randomness was arc4random_stir(), but it has been removed in
100 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
101 function when detected.
103 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
104 cause callback expansion.
106 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
107 syntax errors in an expansion can be treated as a string instead of
108 logging or causing an error, due to the internal use of bool_lax
109 instead of bool when processing it.
111 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
112 server certificates when making smtp deliveries.
114 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
116 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
118 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
120 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
121 Merged patch from Sebastian Wiedenroth.
123 JH/05 Fix results-pipe from transport process. Several recipients, combined
124 with certificate use, exposed issues where response data items split
125 over buffer boundaries were not parsed properly. This eventually
126 resulted in duplicates being sent. This issue only became common enough
127 to notice due to the introduction of conection certificate information,
128 the item size being so much larger. Found and fixed by Wolfgang Breyha.
130 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
131 size buffer was used, resulting in syntax errors when an expansion
134 JH/07 Add support for directories of certificates when compiled with a GnuTLS
135 version 3.3.6 or later.
137 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
138 is EXPERIMENTAL_EVENT, the main-configuration and transport options
139 both become "event_action", the variables become $event_name, $event_data
140 and $event_defer_errno. There is a new variable $verify_mode, usable in
141 routers, transports and related events. The tls:cert event is now also
142 raised for inbound connections, if the main configuration event_action
145 TL/06 In test suite, disable OCSP for old versions of openssl which contained
146 early OCSP support, but no stapling (appears to be less than 1.0.0).
148 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
149 server certificate names available under the smtp transport option
150 "tls_verify_cert_hostname" now do not permit multi-component wildcard
153 JH/10 Time-related extraction expansions from certificates now use the main
154 option "timezone" setting for output formatting, and are consistent
155 between OpenSSL and GnuTLS compilations. Bug 1541.
157 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
158 encoded parameter in the incoming message. Bug 1558.
160 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
161 include certificate info, eximon was claiming there were spoolfile
164 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
166 JH/14 Log delivery-related information more consistently, using the sequence
167 "H=<name> [<ip>]" wherever possible.
169 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
170 are problematic for Debian distribution, omit them from the release
173 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
175 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
178 JH/17 Fix a null-indirection in certextract expansions when a nondefault
179 output list separator was used.
184 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
185 checkers that were complaining about end of non-void function with no
188 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
189 This was a regression intruduced in 4.83 by another bugfix.
191 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
193 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
194 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
200 TF/01 Correctly close the server side of TLS when forking for delivery.
202 When a message was received over SMTP with TLS, Exim failed to clear up
203 the incoming connection properly after forking off the child process to
204 deliver the message. In some situations the subsequent outgoing
205 delivery connection happened to have the same fd number as the incoming
206 connection previously had. Exim would try to use TLS and fail, logging
207 a "Bad file descriptor" error.
209 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
210 utilities have not been installed.
212 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
213 temporary space as the ACL may create new global variables.
215 TL/01 LDAP support uses per connection or global context settings, depending
216 upon the detected version of the libraries at build time.
218 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
219 to extract and use the src ip:port in logging and expansions as if it
220 were a direct connection from the outside internet. PPv2 support was
221 updated based on HAProxy spec change in May 2014.
223 JH/02 Add ${listextract {number}{list}{success}{fail}}.
225 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
226 Properly escape header and check for NULL return.
228 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
231 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
233 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
234 characters in header names, implemented as a verify condition.
235 Contributed by Michael Fischer v. Mollard.
237 TL/05 Rename SPF condition results err_perm and err_temp to standardized
238 results permerror and temperror. Previous values are deprecated but
239 still accepted. In a future release, err_perm and err_temp will be
240 completely removed, which will be a backward incompatibility if the
241 ACL tests for either of these two old results. Patch contributed by
242 user bes-internal on the mailing list.
244 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
246 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
247 selectors, in both main and reject logs.
249 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
252 JH/07 Add malware type "sock" for talking to simple daemon.
254 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
256 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
257 routers/transports under cutthrough routing.
259 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
260 numbers. Touch up "bool" conditional to keep the same definition.
262 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
264 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
266 JH/12 Expand items in router/transport headers_add or headers_remove lists
267 individually rather than the list as a whole. Bug 1452.
269 Required for reasonable handling of multiple headers_ options when
270 they may be empty; requires that headers_remove items with embedded
271 colons must have them doubled (or the list-separator changed).
273 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
274 view the policy declared in the DMARC record. Currently, $dmarc_status
275 is a combined value of both the record presence and the result of the
278 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
280 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
281 dnslookup router and the smtp transport (applying to the forward
284 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
285 of ldap servers used for a specific lookup. Patch provided by Heiko
288 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
289 New variable $lookup_dnssec_authenticated for observability.
291 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
292 Patch submitted by Lars Timman.
294 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
296 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
297 Requires trusted mode and valid format message id, aborts otherwise.
298 Patch contributed by Heiko Schlichting.
300 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
301 certextract with support for various fields. Bug 1358.
303 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
304 is requested by default, modifiable by smtp transport option
307 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
308 operate on certificate variables to give certificate fingerprints
309 Also new ${sha256:cert_variable}.
311 JH/23 The PRDR feature is moved from being Experimental into the mainline.
313 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
314 Christian Aistleitner.
316 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
318 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
319 file. Patch from Wolfgang Breyha.
321 JH/25 Expand the coverage of the delivery $host and $host_address to
322 client authenticators run in verify callout. Bug 1476.
324 JH/26 Port service names are now accepted for tls_on_connect_ports, to
325 align with daemon_smtp_ports. Bug 72.
327 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
328 support and error reporting did not work properly.
330 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
331 and is readable. Patch from Andrew Colin Kissa.
333 TL/14 Enhance documentation of ${run expansion and how it parses the
334 commandline after expansion, particularly in the case when an
335 unquoted variable expansion results in an empty value.
337 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
339 PP/02 Fix internal collision of T_APL on systems which support RFC3123
340 by renaming away from it. Addresses GH issue 15, reported by
343 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
345 TL/15 SECURITY: prevent double expansion in math comparison functions
346 (can expand unsanitized data). Not remotely exploitable.
353 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
355 PP/02 Make -n do something, by making it not do something.
356 When combined with -bP, the name of an option is not output.
358 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
361 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
362 $sender_host_name and config options to manage this, and basic check
365 PP/05 DSCP support for outbound connections and control modifier for inbound.
367 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
368 (Only plugin which currently uses this is kerberos4, which nobody should
369 be using, but we should make it available and other future plugins might
370 conceivably use it, even though it would break NAT; stuff *should* be
371 using channel bindings instead).
373 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
374 name; added for Sendmail compatibility; requires admin caller.
375 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
376 just ignore it); requires trusted caller.
377 Also parse but ignore: -Ac -Am -X<logfile>
380 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
382 TL/02 Add +smtp_confirmation as a default logging option.
384 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
385 Patch by Magnus Holmgren from 2007-02-20.
387 TL/04 Bugzilla 1281 - Spec typo.
388 Bugzilla 1283 - Spec typo.
389 Bugzilla 1290 - Spec grammar fixes.
391 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
393 TL/06 Add Experimental DMARC support using libopendmarc libraries.
395 TL/07 Fix an out of order global option causing a segfault. Reported to dev
396 mailing list by by Dmitry Isaikin.
398 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
400 JH/02 Support "G" suffix to numbers in ${if comparisons.
402 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
404 NM/01 Bugzilla 1197 - Spec typo
405 Bugzilla 1196 - Spec examples corrections
407 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
409 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
410 gnutls_enable_pkcs11, but renamed to more accurately indicate its
413 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
414 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
416 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
417 "acl {{name}{arg}...}", and optional args on acl condition
420 JH/05 Permit multiple router/transport headers_add/remove lines.
422 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
424 JH/07 Avoid using a waiting database for a single-message-only transport.
425 Performance patch from Paul Fisher. Bugzilla 1262.
427 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
430 JH/09 Add $headers_added variable, with content from use of ACL modifier
431 add_header (but not yet added to the message). Bugzilla 199.
433 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
434 Pulled from Bugzilla 817 by Wolfgang Breyha.
436 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
438 (nb: this is the same fix as in Exim 4.80.1)
440 JH/11 Add A= logging on delivery lines, and a client_set_id option on
443 JH/12 Add optional authenticated_sender logging to A= and a log_selector
446 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
448 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
449 advertise SMTP AUTH mechanism to us, instead of a generic
450 protocol violation error. Also, make Exim more robust to bad
451 data from the Dovecot auth socket.
453 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
455 When a queue runner is handling a message, Exim first routes the
456 recipient addresses, during which it prunes them based on the retry
457 hints database. After that it attempts to deliver the message to
458 any remaining recipients. It then updates the hints database using
461 So if a recipient address works intermittently, it can get repeatedly
462 deferred at routing time. The retry hints record remains fresh so the
463 address never reaches the final cutoff time.
465 This is a fairly common occurrence when a user is bumping up against
466 their storage quota. Exim had some logic in its local delivery code
467 to deal with this. However it did not apply to per-recipient defers
468 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
470 This change adds a proper retry rule check during routing so that the
471 final cutoff time is checked against the message's age. We only do
472 this check if there is an address retry record and there is not a
473 domain retry record; this implies that previous attempts to handle
474 the address had the retry_use_local_parts option turned on. We use
475 this as an approximation for the destination being like a local
476 delivery, as in LMTP.
478 I suspect this new check makes the old local delivery cutoff check
479 redundant, but I have not verified this so I left the code in place.
481 TF/02 Correct gecos expansion when From: is a prefix of the username.
483 Test 0254 submits a message to Exim with the header
487 When I ran the test suite under the user fanf2, Exim expanded
488 the header to contain my full name, whereas it should have added
489 a Resent-Sender: header. It erroneously treats any prefix of the
490 username as equal to the username.
492 This change corrects that bug.
494 GF/01 DCC debug and logging tidyup
495 Error conditions log to paniclog rather than rejectlog.
496 Debug lines prefixed by "DCC: " to remove any ambiguity.
498 TF/03 Avoid unnecessary rebuilds of lookup-related code.
500 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
501 Bug spotted by Jeremy Harris; was flawed since initial commit.
502 Would have resulted in OCSP responses post-SNI triggering an Exim
503 NULL dereference and crash.
505 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
507 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
508 Bug detection, analysis and fix by Samuel Thibault.
509 Bugzilla 1331, Debian bug #698092.
511 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
513 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
514 Server implementation by Todd Lyons, client by JH.
515 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
516 config variable "prdr_enable" controls whether the server
517 advertises the facility. If the client requests PRDR a new
518 acl_data_smtp_prdr ACL is called once for each recipient, after
519 the body content is received and before the acl_smtp_data ACL.
520 The client is controlled by bolth of: a hosts_try_prdr option
521 on the smtp transport, and the server advertisement.
522 Default client logging of deliveries and rejections involving
523 PRDR are flagged with the string "PRDR".
525 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
526 fclose(). Diagnosis by Todd Lyons.
528 PP/17 Update configure.default to handle IPv6 localhost better.
529 Patch by Alain Williams (plus minor tweaks).
532 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
533 This is now consistent with GnuTLS, and is now documented: the
534 previous undocumented portable approach to treating the option as
535 unset was to force an expansion failure. That still works, and
536 an empty string is now equivalent.
538 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
539 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
540 not performing validation itself.
542 PP/20 Added force_command boolean option to pipe transport.
543 Patch from Nick Koston, of cPanel Inc.
545 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
548 TF/04 Added udpsend ACL modifer and hexquote expansion operator
550 PP/21 Fix eximon continuous updating with timestamped log-files.
551 Broken in a format-string cleanup in 4.80, missed when I repaired the
552 other false fix of the same issue.
553 Report and fix from Heiko Schlichting.
556 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
557 Report from Prashanth Katuri.
559 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
560 It's SecureTransport, so affects any MacOS clients which use the
561 system-integrated TLS libraries, including email clients.
563 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
564 using a MIME ACL for non-SMTP local injection.
565 Report and assistance in diagnosis by Warren Baker.
567 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
569 JH/16 Fix comparisons for 64b. Bugzilla 1385.
571 TL/09 Add expansion variable $authenticated_fail_id to keep track of
572 last id that failed so it may be referenced in subsequent ACL's.
574 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
577 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
578 ldap library initialization, allowing self-signed CA's to be
579 used. Also properly sets require_cert option later in code by
580 using NULL (global ldap config) instead of ldap handle (per
581 session). Bug diagnosis and testing by alxgomz.
583 TL/12 Enhanced documentation in the ratelimit.pl script provided in
584 the src/util/ subdirectory.
586 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
587 renamed to Transport Post Delivery Action by Jeremy Harris, as
590 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
591 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
592 redis_servers = needs to be configured which will be used by the redis
593 lookup. Patch from Warren Baker, of The Packet Hub.
595 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
597 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
598 hostname or reverse DNS when processing a host list. Used suggestions
599 from multiple comments on this bug.
601 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
603 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
604 Missed a few lines, added it to make the runtest require no keyboard
607 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
608 contains upper case chars. Make router use caseful_local_part.
610 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
611 support when GnuTLS has been built with p11-kit.
617 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
619 This, or similar/improved, will also be change PP/11 of 4.82.
625 PP/01 Handle short writes when writing local log-files.
626 In practice, only affects FreeBSD (8 onwards).
627 Bugzilla 1053, with thanks to Dmitry Isaikin.
629 NM/01 Bugzilla 949 - Documentation tweak
631 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
634 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
636 PP/02 Implemented gsasl authenticator.
638 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
640 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
641 `pkg-config foo` for cflags/libs.
643 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
644 with rest of GSASL and with heimdal_gssapi.
646 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
647 `pkg-config foo` for cflags/libs for the TLS implementation.
649 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
650 properties get this fed in as external SSF. A number of robustness
651 and debugging improvements to the cyrus_sasl authenticator.
653 PP/08 cyrus_sasl server now expands the server_realm option.
655 PP/09 Bugzilla 1214 - Log authentication information in reject log.
656 Patch by Jeremy Harris.
658 PP/10 Added dbmjz lookup type.
660 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
662 PP/12 MAIL args handles TAB as well as SP, for better interop with
663 non-compliant senders.
664 Analysis and variant patch by Todd Lyons.
666 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
667 Bug report from Lars Müller <lars@samba.org> (via SUSE),
668 Patch from Dirk Mueller <dmueller@suse.com>
670 PP/13 tls_peerdn now print-escaped for spool files.
671 Observed some $tls_peerdn in wild which contained \n, which resulted
672 in spool file corruption.
674 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
675 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
676 or write after TLS renegotiation, which otherwise led to messages
679 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
680 as a tracking header (ie: a signed header comes before the signature).
681 Patch from Wolfgang Breyha.
683 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
684 comma-sep list; embedded commas doubled.
686 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
688 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
690 Report and patch from Dmitry Banschikov.
692 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
693 Removed SSL_clear() after SSL_new() which led to protocol negotiation
694 failures. We appear to now support TLS1.1+ with Exim.
696 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
697 lets Exim select keys and certificates based upon TLS SNI from client.
698 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
699 before an outbound SMTP session. New log_selector, +tls_sni.
701 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
702 NULL dereference. Report and patch from Alun Jones.
704 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
705 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
706 Not seeing resolver debug output on NetBSD, but suspect this is a
707 resolver implementation change.
709 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
710 Left warnings. Added "eximon gdb" invocation mode.
712 PP/21 Defaulting "accept_8bitmime" to true, not false.
714 PP/22 Added -bw for inetd wait mode support.
716 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
717 locate the relevant includes and libraries. Made this the default.
719 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
720 Bugzilla 1246, report and most of solution from Tomasz Kusy.
722 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
723 This may cause build issues on older platforms.
725 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
726 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
727 gnutls_require_mac & gnutls_require_protocols (no longer supported).
728 Added SNI support via GnuTLS too.
729 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
731 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
733 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
734 Applied second patch from Janne, implementing suggestion to default
735 multiple-strings-in-record handling to match SPF spec.
737 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
739 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
740 read-only, out of scope).
741 Patch from Wolfgang Breyha, report from Stuart Northfield.
743 PP/29 Fix three issues highlighted by clang analyser static analysis.
744 Only crash-plausible issue would require the Cambridge-specific
745 iplookup router and a misconfiguration.
746 Report from Marcin Mirosław.
748 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
750 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
751 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
752 As part of this, removing so much warning spew let me fix some minor
753 real issues in debug logging.
755 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
756 assignment on my part. Fixed.
758 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
759 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
760 Janne Snabb (who went above and beyond: thank you).
762 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
763 string otherwise requires a connection and a bunch more work and it's
764 relatively easy to get wrong. Should also expose TLS library linkage
767 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
768 64-bit ${eval} (JH/03).
770 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
771 GNU libc to support some of the 64-bit stuff, should not lead to
772 conflicts. Defined before os.h is pulled in, so if a given platform
773 needs to override this, it can.
775 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
776 protection layer was required, which is not implemented.
777 Bugzilla 1254, patch from Wolfgang Breyha.
779 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
780 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
781 tls_dhparam take prime identifiers. Also unbreak combination of
782 OpenSSL+DH_params+TLSSNI.
784 PP/39 Disable SSLv2 by default in OpenSSL support.
790 PP/01 Solaris build fix for Oracle's LDAP libraries.
791 Bugzilla 1109, patch from Stephen Usher.
793 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
795 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
798 TF/02 Fix a couple more cases where we did not log the error message
799 when unlink() failed. See also change 4.74-TF/03.
801 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
802 lock up or crash if it happened to be inside a call to libc when it
803 got a SIGUSR1 from exiwhat.
805 The SIGUSR1 handler appends the current process status to the process
806 log which is later printed by exiwhat. It used to use the general
807 purpose logging code to do this, but several functions it calls are
808 not safe for signals.
810 The new output code in the SIGUSR1 handler is specific to the process
811 log, and simple enough that it's easy to inspect for signal safety.
812 Removing some special cases also simplifies the general logging code.
813 Removing the spurious timestamps from the process log simplifies
816 TF/04 Improved ratelimit ACL condition.
818 The /noupdate option has been deprecated in favour of /readonly which
819 has clearer semantics. The /leaky, /strict, and /readonly update modes
820 are mutually exclusive. The update mode is no longer included in the
821 database key; it just determines when the database is updated. (This
822 means that when you upgrde Exim will forget old rate measurements.)
824 Exim now checks that the per_* options are used with an update mode that
825 makes sense for the current ACL. For example, when Exim is processing a
826 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
827 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
828 must specify per_mail/readonly. If you omit the update mode it defaults to
829 /leaky where that makes sense (as before) or /readonly where required.
831 The /noupdate option is now undocumented but still supported for
832 backwards compatibility. It is equivalent to /readonly except that in
833 ACLs where /readonly is required you may specify /leaky/noupdate or
834 /strict/noupdate which are treated the same as /readonly.
836 A useful new feature is the /count= option. This is a generalization
837 of the per_byte option, so that you can measure the throughput of other
838 aggregate values. For example, the per_byte option is now equivalent
839 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
841 The per_rcpt option has been generalized using the /count= mechanism
842 (though it's more complicated than the per_byte equivalence). When it is
843 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
844 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
845 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
846 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
847 Note that using per_rcpt with a non-readonly update mode in more than
848 one ACL will cause the recipients to be double-counted. (The per_mail
849 and per_byte options don't have this problem.)
851 The handling of very low rates has changed slightly. If the computed rate
852 is less than the event's count (usually one) then this event is the first
853 after a long gap. In this case the rate is set to the same as this event's
854 count, so that the first message of a spam run is counted properly.
856 The major new feature is a mechanism for counting the rate of unique
857 events. The new per_addr option counts the number of different
858 recipients that someone has sent messages to in the last time period. It
859 behaves like per_rcpt if all the recipient addresses are different, but
860 duplicate recipient addresses do not increase the measured rate. Like
861 the /count= option this is a general mechanism, so the per_addr option
862 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
863 example, measure the rate that a client uses different sender addresses
864 with the options per_mail/unique=$sender_address. There are further
865 details in the main documentation.
867 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
869 TF/06 Removed a few PCRE remnants.
871 TF/07 Automatically extract Exim's version number from tags in the git
872 repository when doing development or release builds.
874 PP/02 Raise smtp_cmd_buffer_size to 16kB.
875 Bugzilla 879. Patch from Paul Fisher.
877 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
878 Heavily based on revision 40f9a89a from Simon Arlott's tree.
881 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
883 PP/05 Variable $av_failed, true if the AV scanner deferred.
884 Bugzilla 1078. Patch from John Horne.
886 PP/06 Stop make process more reliably on build failure.
887 Bugzilla 1087. Patch from Heiko Schlittermann.
889 PP/07 Make maildir_use_size_file an _expandable_ boolean.
890 Bugzilla 1089. Patch from Heiko Schlittermann.
892 PP/08 Handle ${run} returning more data than OS pipe buffer size.
893 Bugzilla 1131. Patch from Holger Weiß.
895 PP/09 Handle IPv6 addresses with SPF.
896 Bugzilla 860. Patch from Wolfgang Breyha.
898 PP/10 GnuTLS: support TLS 1.2 & 1.1.
900 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
903 PP/11 match_* no longer expand right-hand-side by default.
904 New compile-time build option, EXPAND_LISTMATCH_RHS.
905 New expansion conditions, "inlist", "inlisti".
907 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
909 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
911 PP/14 fix log_write() format string regression from TF/03.
912 Bugzilla 1152. Patch from Dmitry Isaikin.
918 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
920 PP/02 Harmonised TLS library version reporting; only show if debugging.
921 Layout now matches that introduced for other libraries in 4.74 PP/03.
923 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
925 PP/04 New "dns_use_edns0" global option.
927 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
930 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
931 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
933 TK/01 Updated PolarSSL code to 0.14.2.
934 Bugzilla 1097. Patch from Andreas Metzler.
936 PP/07 Catch divide-by-zero in ${eval:...}.
939 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
942 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
943 format-string attack -- SECURITY: remote arbitrary code execution.
945 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
946 time unintentionally subject to list matching rules, letting the header
947 cause arbitrary Exim lookups (of items which can occur in lists, *not*
948 arbitrary string expansion). This allowed for information disclosure.
950 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
951 INT_MIN/-1 -- value coerced to INT_MAX.
957 NM/01 Workround for PCRE version dependency in version reporting
960 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
961 This fixes portability to compilers other than gcc, notably
962 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
964 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
965 makefiles for portability to HP-UX and POSIX correctness.
967 PP/01 Permit LOOKUP_foo enabling on the make command-line.
968 Also via indented variable definition in the Makefile.
969 (Debugging by Oliver Heesakkers).
971 PP/02 Restore caching of spamd results with expanded spamd_address.
972 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
974 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
975 Improves build reliability. Fix from: Frank Elsner
977 NM/02 Fix wide character breakage in the rfc2047 coding
978 Fixes bug 1064. Patch from Andrey N. Oktyabrski
980 NM/03 Allow underscore in dnslist lookups
981 Fixes bug 1026. Patch from Graeme Fowler
983 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
984 Code patches from Adam Ciarcinski of NetBSD.
986 NM/04 Fixed exiqgrep to cope with mailq missing size issue
989 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
990 is logged, to avoid truncation. Patch from John Horne.
992 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
993 Patch from Jakob Hirsch.
995 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
996 SQL string expansion failure details.
997 Patch from Andrey Oktyabrski.
999 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1000 Patch from Simon Arlott.
1002 PP/09 New lookups functionality failed to compile on old gcc which rejects
1003 extern declarations in function scope.
1004 Patch from Oliver Fleischmann
1006 PP/10 Use sig_atomic_t for flags set from signal handlers.
1007 Check getgroups() return and improve debugging.
1008 Fixed developed for diagnosis in bug 927 (which turned out to be
1011 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1012 Patch from Mark Zealey.
1014 PP/12 Bugzilla 1056: Improved spamd server selection.
1015 Patch from Mark Zealey.
1017 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1018 Based on patch from Heiko Schlittermann.
1020 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1021 Patch from Uwe Doering, sign-off by Michael Haardt.
1023 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1024 variable declaration deep within a block. Bug and patch from
1027 PP/15 lookups-Makefile IRIX compatibilty coercion.
1029 PP/16 Make DISABLE_DKIM build knob functional.
1031 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1032 Patch by Simon Arlott
1034 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1035 variable argument macros. Our copy now differs from upstream.
1041 TF/01 Failure to get a lock on a hints database can have serious
1042 consequences so log it to the panic log.
1044 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1045 controlled using the smtp_confirmation log selector.
1047 TF/03 Include the error message when we fail to unlink a spool file.
1049 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1050 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1051 for maintaining out-of-tree patches for some time.
1053 PP/01 Bugzilla 139: Documentation and portability issues.
1054 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1055 Handle per-OS dynamic-module compilation flags.
1057 PP/02 Let /dev/null have normal permissions.
1058 The 4.73 fixes were a little too stringent and complained about the
1059 permissions on /dev/null. Exempt it from some checks.
1060 Reported by Andreas M. Kirchwitz.
1062 PP/03 Report version information for many libraries, including
1063 Exim version information for dynamically loaded libraries. Created
1064 version.h, now support a version extension string for distributors
1065 who patch heavily. Dynamic module ABI change.
1067 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1068 privilege escalation vulnerability whereby the Exim run-time user
1069 can cause root to append content of the attacker's choosing to
1072 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1075 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1076 If dropping privileges for untrusted macros, we disabled normal logging
1077 on the basis that it would fail; for the Exim run-time user, this is not
1078 the case, and it resulted in successful deliveries going unlogged.
1079 Fixed. Reported by Andreas Metzler.
1085 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1086 only prepend for the Resent-* case. Fixes regression introduced in
1087 Exim 4.70 by NM/22 for Bugzilla 607.
1089 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1090 increasing numbers of administrators be bitten by this.
1092 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1094 PP/03 Bugzilla 994: added openssl_options main configuration option.
1096 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1098 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1100 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1102 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1103 without a peer certificate, leading to a segfault because of an
1104 assumption that peers always have certificates. Be a little more
1105 paranoid. Problem reported by Martin Tscholak.
1107 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1108 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1109 NB: ClamAV planning to remove STREAM in "middle of 2010".
1110 CL also introduces -bmalware, various -d+acl logging additions and
1111 more caution in buffer sizes.
1113 PP/09 Implemented reverse_ip expansion operator.
1115 PP/10 Bugzilla 937: provide a "debug" ACL control.
1117 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1119 PP/12 Bugzilla 973: Implement --version.
1121 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1123 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1125 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1127 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1128 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1129 ignore trailing whitespace.
1131 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1133 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1136 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1137 Notification from Dr Andrew Aitchison.
1139 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1140 ExtendedDetectionInfo response format.
1141 Notification from John Horne.
1143 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1146 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1147 XSL and documented dependency on system catalogs, with examples of how
1150 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1153 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1154 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1155 configuration file which is writeable by the Exim user or group.
1157 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1158 of configuration files to cover files specified with the -C option if
1159 they are going to be used with root privileges, not just the default
1162 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1163 option (effectively making it always true).
1165 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1166 files to be used while preserving root privileges.
1168 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1169 that rogue child processes cannot use them.
1171 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1172 run-time user, instead of root.
1174 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1175 Exim run-time user without dropping privileges.
1177 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1178 result string, instead of calling string_vformat() twice with the same
1181 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1182 for other users. Others should always drop root privileges if they use
1183 -C on the command line, even for a whitelisted configure file.
1185 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1187 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1193 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1194 $data_path, and $header_path variables; fixed documentation bugs and
1197 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1198 exipick to access non-standard spools, including the "frozen" queue
1201 NM/01 Bugzilla 965: Support mysql stored procedures.
1202 Patch from Alain Williams
1204 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1206 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1207 Patch from Andreas Metzler
1209 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1210 Patch from Kirill Miazine
1212 NM/05 Bugzilla 671: Added umask to procmail example.
1214 JJ/03 installed exipick 20100323.0, fixing doc bug
1216 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1217 directory. Notification and patch from Dan Rosenberg.
1219 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1221 TK/02 Improve log output when DKIM signing operation fails.
1223 MH/01 Treat the transport option dkim_domain as a colon separated
1224 list, not as a single string, and sign the message with each element,
1225 omitting multiple occurences of the same signer.
1227 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1228 Bugzilla 985, 986. Patch by Simon Arlott
1230 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1231 Patch by Simon Arlott
1233 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1234 MBX locking. Notification from Dan Rosenberg.
1240 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1242 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1244 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1246 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1248 NM/04 Bugzilla 915: Flag broken perl installation during build.
1254 TK/01 Added patch by Johannes Berg that expands the main option
1255 "spamd_address" if it starts with a dollar sign.
1257 TK/02 Write list of recipients to X-Envelope-Sender header when building
1258 the mbox-format spool file for content scanning (suggested by Jakob
1261 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1262 (http://www.dcc-servers.net/) support via dccifd. Activated by
1263 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1265 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1266 by Mark Daniel Reidel <mr@df.eu>.
1268 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1269 When building exim an external PCRE library is now needed -
1270 PCRE is a system library on the majority of modern systems.
1271 See entry on PCRE_LIBS in EDITME file.
1273 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1274 conversation. Added nologin parameter to request.
1275 Patch contributed by Kirill Miazine.
1277 TF/01 Do not log submission mode rewrites if they do not change the address.
1279 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1281 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1282 log files in place. Contributed by Roberto Lima.
1284 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1286 TF/03 Bugzilla 615: When checking the local_parts router precondition
1287 after a local_part_suffix or local_part_prefix option, Exim now
1288 does not use the address's named list lookup cache, since this
1289 contains cached lookups for the whole local part.
1291 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1292 Robert Millan. Documentation is in experimental-spec.txt.
1294 TF/04 Bugzilla 668: Fix parallel build (make -j).
1296 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1298 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1299 Patch provided by Jan Srzednicki.
1301 TF/05 Leading white space used to be stripped from $spam_report which
1302 wrecked the formatting. Now it is preserved.
1304 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1305 that they are available at delivery time.
1307 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1309 TF/08 TLS error reporting now respects the incoming_interface and
1310 incoming_port log selectors.
1312 TF/09 Produce a more useful error message if an SMTP transport's hosts
1313 setting expands to an empty string.
1315 NM/06 Bugzilla 744: EXPN did not work under TLS.
1316 Patch provided by Phil Pennock.
1318 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1319 Patch provided by Richard Godbee.
1321 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1322 acl_smtp_notquit, added index entry.
1324 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1325 Patch provided by Eugene Bujak.
1327 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1328 accept(). Patch provided by Maxim Dounin.
1330 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1331 Patch provided by Phil Pennock.
1333 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1335 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1336 Patch provided by Brad "anomie" Jorsch.
1338 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1339 Patch provided by Dean Brooks.
1341 TK/05 Add native DKIM support (does not depend on external libraries).
1343 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1344 Patch provided by Graeme Fowler.
1346 NM/16 Bugzilla 851: Documentation example syntax fix.
1348 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1350 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1353 NM/19 Bugzilla 745: TLS version reporting.
1354 Patch provided by Phil Pennock.
1356 NM/20 Bugzilla 167: bool: condition support.
1357 Patch provided by Phil Pennock.
1359 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1360 clients. Patch provided by Phil Pennock.
1362 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1363 Patch provided by Brad "anomie" Jorsch.
1365 NM/23 Bugzilla 687: Fix misparses in eximstats.
1366 Patch provided by Heiko Schlittermann.
1368 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1369 Patch provided by Heiko Schlittermann.
1371 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1372 plus update to original patch.
1374 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1376 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1377 Patch provided by David Brownlee.
1379 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1381 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1383 NM/30 Bugzilla 888: TLS documentation bugfixes.
1385 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1387 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1388 Unlike the original bugzilla I have changed all shell scripts in src tree.
1390 NM/33 Bugzilla 898: Transport filter timeout fix.
1391 Patch by Todd Rinaldo.
1393 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1394 Patch by Serge Demonchaux.
1396 NM/35 Bugzilla 39: Base64 decode bug fixes.
1397 Patch by Jakob Hirsch.
1399 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1401 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1403 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1405 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1411 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1412 ALT-N's libdkim that I have put here:
1413 http://duncanthrax.net/exim-experimental/
1415 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1416 were called 'true' and it seems that C99 defines that as a reserved
1417 keyword to be used with 'bool' variable types. That means you could
1418 not include C99-style headers which use bools without triggering
1419 build errors in sieve.c.
1421 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1422 as mailq or other aliases. Changed the --help handling significantly
1423 to do whats expected. exim_usage() emits usage/help information.
1425 SC/01 Added the -bylocaldomain option to eximstats.
1427 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1429 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1431 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1437 PH/01 Another patch from the Sieve maintainer.
1439 PH/02 When an IPv6 address is converted to a string for single-key lookup
1440 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1441 dots are used instead of colons so that keys in lsearch files need not
1442 contain colons. This was done some time before quoting was made available
1443 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1444 (notated using the quote facility) so as to distinguish them from IPv4
1445 keys. This meant that lookups for IP addresses in host lists did not work
1446 for iplsearch lookups.
1448 This has been fixed by arranging for IPv6 addresses to be expressed with
1449 colons if the lookup type is iplsearch. This is not incompatible, because
1450 previously such lookups could never work.
1452 The situation is now rather anomolous, since one *can* have colons in
1453 ordinary lsearch keys. However, making the change in all cases is
1454 incompatible and would probably break a number of configurations.
1456 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1459 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1460 conversion specification without a maximum field width, thereby enabling
1461 a rogue spamd server to cause a buffer overflow. While nobody in their
1462 right mind would setup Exim to query an untrusted spamd server, an
1463 attacker that gains access to a server running spamd could potentially
1464 exploit this vulnerability to run arbitrary code as the Exim user.
1466 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1467 $primary_hostname instead of what libspf2 thinks the hosts name is.
1469 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1470 a directory entry by the name of the lookup key. Previously, if a
1471 symlink pointed to a non-existing file or a file in a directory that
1472 Exim lacked permissions to read, a lookup for a key matching that
1473 symlink would fail. Now it is enough that a matching directory entry
1474 exists, symlink or not. (Bugzilla 503.)
1476 PH/03 The body_linecount and body_zerocount variables are now exported in the
1479 PH/04 Added the $dnslist_matched variable.
1481 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1482 This means they are set thereafter only if the connection becomes
1485 PH/06 Added the client_condition to authenticators so that some can be skipped
1486 by clients under certain conditions.
1488 PH/07 The error message for a badly-placed control=no_multiline_responses left
1489 "_responses" off the end of the name.
1491 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1493 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1494 (without spaces) instead of just copying the configuration text.
1496 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1498 PH/11 Added $max_received_linelength.
1500 PH/12 Added +ignore_defer and +include_defer to host lists.
1502 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1503 way in which PCRE > 7.0 is built.
1505 PH/14 Implemented queue_only_load_latch.
1507 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1508 MAIL command. The effect was to mangle the value on 64-bit systems.
1510 PH/16 Another patch from the Sieve maintainer.
1512 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1514 PH/18 If a system quota error occurred while trying to create the file for
1515 a maildir delivery, the message "Mailbox is full" was not appended to the
1516 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1517 only to a quota excession during the actual writing of the file.
1519 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1520 characters?) which causes problems in log lines. The DN values are now
1521 passed through string_printing() before being added to log lines.
1523 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1524 and InterBase are left for another time.)
1526 PH/21 Added message_body_newlines option.
1528 PH/22 Guard against possible overflow in moan_check_errorcopy().
1530 PH/23 POSIX allows open() to be a macro; guard against that.
1532 PH/24 If the recipient of an error message contained an @ in the local part
1533 (suitably quoted, of course), incorrect values were put in $domain and
1534 $local_part during the evaluation of errors_copy.
1540 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1541 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1544 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1545 issue a MAIL command.
1547 PH/02 In an ACL statement such as
1549 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1551 if a client was not listed at all, or was listed with a value other than
1552 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1553 the condition was not true (as it should be), so access was not denied.
1554 The bug was that the ! inversion was incorrectly passed on to the second
1555 item. This has been fixed.
1557 PH/03 Added additional dnslists conditions == and =& which are different from
1558 = and & when the dns lookup returns more than one IP address.
1560 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1561 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1563 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1564 FSYNC, which compiles an option called disable_fsync that allows for
1565 bypassing fsync(). The documentation is heavily laced with warnings.
1567 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1569 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1570 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1571 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1572 including adding "make clean"; (3) Added -fPIC when compiling the test
1573 dynamically loaded module, to get rid of a warning.
1575 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1576 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1577 = 0s. The bug is otherwise harmless.
1579 PH/07 There was a bug in the dovecot authenticator such that the value of
1580 $auth1 could be overwritten, and so not correctly preserved, after a
1581 successful authentication. This usually meant that the value preserved by
1582 the server_setid option was incorrect.
1584 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1586 PH/09 Installed PCRE release 7.0.
1588 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1589 run for batched SMTP input. It is now run at the start of every message
1590 in the batch. While fixing this I discovered that the process information
1591 (output by running exiwhat) was not always getting set for -bs and -bS
1592 input. This is fixed, and it now also says "batched" for BSMTP.
1594 PH/11 Added control=no_pipelining.
1596 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1597 patch, slightly modified), and move the expansion of helo_data till after
1598 the connection is made in the smtp transport (so it can use these
1601 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1603 PH/14 Added log_selector = +pid.
1605 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1607 PH/16 Add ${if forany and ${if forall.
1609 PH/17 Added dsn_from option to vary the From: line in DSNs.
1611 PH/18 Flush SMTP output before performing a callout, unless control =
1612 no_callout_flush is set.
1614 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1615 was true (the default) a successful delivery failed to delete the retry
1616 item, thus causing premature timeout of the address. The bug is now
1619 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1621 PH/21 Long custom messages for fakedefer and fakereject are now split up
1622 into multiline reponses in the same way that messages for "deny" and
1623 other ACL rejections are.
1625 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1626 with slight modification.
1628 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1629 draft, changing the syntax and factoring some duplicate code".
1631 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1632 for deliveries of the second and subsequent messages over the same SMTP
1635 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1636 ${reduce, with only minor "tidies".
1638 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1640 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1641 expansion side effects.
1643 PH/27 When a message times out after an over-quota error from an Exim-imposed
1644 quota, the bounce message says "mailbox is full". This message was not
1645 being given when it was a system quota that was exceeded. It now should
1648 MH/03 Made $recipients available in local_scan(). local_scan() already has
1649 better access to the recipient list through recipients_list[], but
1650 $recipients can be useful in postmaster-provided expansion strings.
1652 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1653 in the case of a MAIL command with additional options following the
1654 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1655 were accidentally chopped off.
1657 PH/29 SMTP synchronization checks are implemented when a command is read -
1658 there is a check that no more input is waiting when there shouldn't be
1659 any. However, for some commands, a delay in an ACL can mean that it is
1660 some time before the response is written. In this time, more input might
1661 arrive, invalidly. So now there are extra checks after an ACL has run for
1662 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1663 pipelining has not been advertised.
1665 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1667 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1668 returned in the SMTP response when smtp_return_error_details was false.
1669 This has been fixed.
1671 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1672 instead of the C I/O that was originally supplied, because problems were
1673 reported on Solaris.
1675 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1676 Exim which did not show up earlier: it was assuming that a call to
1677 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1678 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1679 was a macro that became an assignment, so it seemed to work. This has
1680 changed to a proper function call with a void return, hence the compile
1681 error. Exim's code has been fixed.
1683 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1686 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1688 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1690 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1691 because it turns out to be too big or there is a timeout) there is no
1692 "Completed" line in the log. When some messages of this type were
1693 selected by exigrep, they were listed as "not completed". Others were
1694 picked up by some special patterns. I have improved the selection
1695 criteria to be more general.
1697 PH/38 The host_find_failed option in the manualroute router can now be set
1698 to "ignore", to completely ignore a host whose IP address cannot be
1699 found. If all hosts are ignored, the behaviour is controlled by the new
1700 host_all_ignored option.
1702 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1703 homing or because of multiple MX records with /mx) generated more than
1704 one IP address, and the following item turned out to be the local host,
1705 all the secondary addresses of the first item were incorrectly removed
1706 from the list, along with the local host and any following hosts (which
1707 is what is supposed to happen).
1709 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1710 whoever called Exim into the -H file. In the case of the daemon it was
1711 behaving confusingly. When first started, it used values for whoever
1712 started the daemon, but after a SIGHUP it used the Exim user (because it
1713 calls itself on a restart). I have changed the code so that it now always
1716 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1717 message are rejected with the same error (e.g. no authentication or bad
1718 sender address), and a DATA command is nevertheless sent (as can happen
1719 with PIPELINING or a stupid MUA), the error message that was given to the
1720 RCPT commands is included in the rejection of the DATA command. This is
1721 intended to be helpful for MUAs that show only the final error to their
1724 PH/42 Another patch from the Sieve maintainer.
1726 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1727 Eximstats - Fixed some broken HTML links and added missing column headers
1729 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1730 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1732 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1734 PH/43 Yet another patch from the Sieve maintainer.
1736 PH/44 I found a way to check for a TCP/IP connection going away before sending
1737 the response to the final '.' that terminates a message, but only in the
1738 case where the client has not sent further data following the '.'
1739 (unfortunately, this is allowed). However, in many cases there won't be
1740 any further data because there won't be any more messages to send. A call
1741 to select() can be used: if it shows that the input is "ready", there is
1742 either input waiting, or the socket has been closed. An attempt to read
1743 the next input character can distinguish the two cases. Previously, Exim
1744 would have sent an OK response which the client would never have see.
1745 This could lead to message repetition. This fix should cure that, at
1746 least in a lot of common cases.
1748 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1749 advertised in response to EHLO.
1755 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1756 fixed by 4.65/MH/01 (is this a record?) are fixed:
1758 (i) An empty string was always treated as zero by the numeric comparison
1759 operators. This behaviour has been restored.
1761 (ii) It is documented that the numeric comparison operators always treat
1762 their arguments as decimal numbers. This was broken in that numbers
1763 starting with 0 were being interpreted as octal.
1765 While fixing these problems I realized that there was another issue that
1766 hadn't been noticed. Values of message_size_limit (both the global option
1767 and the transport option) were treated as octal if they started with 0.
1768 The documentation was vague. These values are now always treated as
1769 decimal, and I will make that clear in the documentation.
1775 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1776 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1779 MH/01 Don't check that the operands of numeric comparison operators are
1780 integers when their expansion is in "skipping" mode (fixes bug
1781 introduced by 4.64-PH/07).
1783 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1784 child addresses, Exim now panics and dies. Previously, because the count
1785 is held in a short int, deliveries were likely to be lost. As such a
1786 large number of recipients for a single message is ridiculous
1787 (performance will be very, very poor), I have chosen to impose a limit
1788 rather than extend the field.
1794 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1795 leftover -K file (the existence of which was triggered by #402).
1796 While we were at it, introduced process PID as part of the -K
1797 filename. This should rule out race conditions when creating
1800 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1801 processing considerably. Previous code took too long for large mails,
1802 triggering a timeout which in turn triggers #401.
1804 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1805 in the DK code in transports.c. sendfile() is not really portable,
1806 hence the _LINUX specificness.
1808 TF/01 In the add_headers option to the mail command in an Exim filter,
1809 there was a bug that Exim would claim a syntax error in any
1810 header after the first one which had an odd number of characters
1813 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1814 callout verification, Exim cached a "reject" for the entire domain. This
1815 is correct for most verifications, but it is not correct for a recipient
1816 verification with use_sender or use_postmaster set, because in that case
1817 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1818 case of MAIL FROM:<> rejection from other early rejections (e.g.
1819 rejection of HELO). When verifying a recipient using a non-null MAIL
1820 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1821 Whatever the result of the callout, the value of the domain cache is
1822 left unchanged (for any other kind of callout, getting as far as trying
1823 RCPT means that the domain itself is ok).
1825 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1828 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1829 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1830 session, the daemon ignores ECONNECT errors and logs others; it now
1831 ignores EPIPE as well.
1833 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1834 (quoted-printable decoding).
1836 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1837 later the small subsequent patch to fix an introduced bug.
1839 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1841 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1843 PH/08 An error is now given if message_size_limit is specified negative.
1845 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1846 to be given (somewhat) arbitrary names.
1848 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1851 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1852 miscellaneous code fixes
1854 PH/10 Added the log_reject_target ACL modifier to specify where to log
1857 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1858 hostname. This is wrong, because it relates to the incoming message (and
1859 probably the interface on which it is arriving) and not to the outgoing
1860 callout (which could be using a different interface). This has been
1861 changed to use the value of the helo_data option from the smtp transport
1862 instead - this is what is used when a message is actually being sent. If
1863 there is no remote transport (possible with a router that sets up host
1864 addresses), $smtp_active_hostname is used.
1866 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1867 tweaks were necessary in order to get it to work (see also 21 below):
1868 (a) The code assumed that strncpy() returns a negative number on buffer
1869 overflow, which isn't the case. Replaced with Exim's string_format()
1871 (b) There were several signed/unsigned issues. I just did the minimum
1872 hacking in of casts. There is scope for a larger refactoring.
1873 (c) The code used strcasecmp() which is not a standard C function.
1874 Replaced with Exim's strcmpic() function.
1875 (d) The code set only $1; it now sets $auth1 as well.
1876 (e) A simple test gave the error "authentication client didn't specify
1877 service in request". It would seem that Dovecot has changed its
1878 interface. Fortunately there's a specification; I followed it and
1879 changed what the client sends and it appears to be working now.
1881 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1884 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1885 address A is aliased to B and C, where B exists and C does not. Without
1886 -v the output is "A verified" because verification stops after a
1887 successful redirection if more than one address is generated. However,
1888 with -v the child addresses are also verified. Exim was outputting "A
1889 failed to verify" and then showing the successful verification for C,
1890 with its parentage. It now outputs "B failed to verify", showing B's
1891 parentage before showing the successful verification of C.
1893 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1894 look up a TXT record in a specific list after matching in a combined
1897 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1898 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1899 they consult the DNS. I had assumed they would set it the way they
1900 wanted; and indeed my experiments on Linux seem to show that in some
1901 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1902 To be on the safe side, however, I have now made the interface to
1903 host_find_byname() similar to host_find_bydns(), with an argument
1904 containing the DNS resolver options. The host_find_byname() function now
1905 sets these options at its start, just as host_find_bydns() does. The smtp
1906 transport options dns_qualify_single and dns_search_parents are passed to
1907 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1908 of host_find_byname() use the default settings of RES_DEFNAMES
1909 (qualify_single) but not RES_DNSRCH (search_parents).
1911 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1912 spool_read_header() do less string testing, by means of a preliminary
1913 switch on the second character of optional "-foo" lines. (This is
1914 overdue, caused by the large number of possibilities that now exist.
1915 Originally there were few.) While I was there, I also converted the
1916 str(n)cmp tests so they don't re-test the leading "-" and the first
1917 character, in the hope this might squeeze out yet more improvement.
1919 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1920 flag allowing group syntax was set by the header_syntax check but not
1921 turned off, possible causing trouble later; (2) The flag was not being
1922 set at all for the header_verify test, causing "group"-style headers to
1923 be rejected. I have now set it in this case, and also caused header_
1924 verify to ignore an empty address taken from a group. While doing this, I
1925 came across some other cases where the code for allowing group syntax
1926 while scanning a header line wasn't quite right (mostly, not resetting
1927 the flag correctly in the right place). These bugs could have caused
1928 trouble for malformed header lines. I hope it is now all correct.
1930 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1931 with the "reply" argument non-NULL. The code, however (which originally
1932 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1933 but it didn't always do it. This confused somebody who was copying the
1934 code for some other use. I have removed all the tests.
1936 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1937 feature that was used to support insecure browsers during the U.S. crypto
1938 embargo. It requires special client support, and Exim is probably the
1939 only MTA that supported it -- and would never use it because real RSA is
1940 always available. This code has been removed, because it had the bad
1941 effect of slowing Exim down by computing (never used) parameters for the
1942 RSA_EXPORT functionality.
1944 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1945 authenticator to fail if there's a tab character in the incoming data
1946 (there should never be unless someone is messing about, as it's supposed
1947 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1948 if the connection is using TLS or if the remote IP is the same as the
1949 local IP, and the "valid-client-cert option" if a client certificate has
1952 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1953 authenticators. This can be used for authorization after authentication
1954 succeeds. (In the case of plaintext, it servers for both authentication
1957 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1958 if any retry times were supplied.
1960 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1961 connection, where there is no client IP address to check. In this
1962 situation, the verify now always succeeds.
1964 PH/25 Applied John Jetmore's -Mset patch.
1966 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1968 PH/27 In a string expansion for a processed (not raw) header when multiple
1969 headers of the same name were present, leading whitespace was being
1970 removed from all of them, but trailing whitespace was being removed only
1971 from the last one. Now trailing whitespace is removed from each header
1972 before concatenation. Completely empty headers in a concatenation (as
1973 before) are ignored.
1975 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1976 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1978 PH/29 [Removed. This was a change that I later backed out, and forgot to
1979 correct the ChangeLog entry (that I had efficiently created) before
1980 committing the later change.]
1982 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1983 address errors (4xx response to RCPT) over the same connection as other
1984 messages routed to the same hosts. Such deliveries are always "forced",
1985 so retry times are not inspected. This resulted in far too many retries
1986 for the affected addresses. The effect occurred only when there were more
1987 hosts than the hosts_max_try setting in the smtp transport when it had
1988 the 4xx errors. Those hosts that it had tried were not added to the list
1989 of hosts for which the message was waiting, so if all were tried, there
1990 was no problem. Two fixes have been applied:
1992 (i) If there are any address or message errors in an SMTP delivery, none
1993 of the hosts (tried or untried) are now added to the list of hosts
1994 for which the message is waiting, so the message should not be a
1995 candidate for sending over the same connection that was used for a
1996 successful delivery of some other message. This seems entirely
1997 reasonable: after all the message is NOT "waiting for some host".
1998 This is so "obvious" that I'm not sure why it wasn't done
1999 previously. Hope I haven't missed anything, but it can't do any
2000 harm, as the worst effect is to miss an optimization.
2002 (ii) If, despite (i), such a delivery is accidentally attempted, the
2003 routing retry time is respected, so at least it doesn't keep
2004 hammering the server.
2006 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2007 in ${readsocket because some servers need this prod.
2009 PH/32 Added some extra debug output when updating a wait-xxx database.
2011 PH/33 The hint "could be header name not terminated by colon", which has been
2012 given for certain expansion errors for a long time, was not being given
2013 for the ${if def:h_colon_omitted{... case.
2015 PH/34 The spec says: "With one important exception, whenever a domain list is
2016 being scanned, $domain contains the subject domain." There was at least
2017 one case where this was not true.
2019 PH/35 The error "getsockname() failed: connection reset by peer" was being
2020 written to the panic log as well as the main log, but it isn't really
2021 panic-worthy as it just means the connection died rather early on. I have
2022 removed the panic log writing for the ECONNRESET error when getsockname()
2025 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2026 runs only) independently of the message's sender address. This meant
2027 that, if the 4xx error was in fact related to the sender, a different
2028 message to the same recipient with a different sender could confuse
2029 things. In particualar, this can happen when sending to a greylisting
2030 server, but other circumstances could also provoke similar problems.
2031 I have changed the default so that the retry time for these errors is now
2032 based a combination of the sender and recipient addresses. This change
2033 can be overridden by setting address_retry_include_sender=false in the
2036 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2037 remote server are returned as part of bounce messages. This was not
2038 happening for LMTP over a pipe (the lmtp transport), but now it is the
2039 same for both kinds of LMTP.
2041 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2042 in header lines that were in fact CNAMEs. This is no longer the case.
2044 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2045 and queue runs started by the daemon processed all messages. This has
2046 been fixed so that -R and -S can now usefully be given with -q<time>.
2048 PH/40 Import PCRE release 6.7 (fixes some bugs).
2050 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2052 PH/42 Give an error if -q is specified more than once.
2054 PH/43 Renamed the variables $interface_address and $interface_port as
2055 $received_ip_address and $received_port, to make it clear that these
2056 values apply to message reception, and not to the outgoing interface when
2057 a message is delivered. (The old names remain recognized, of course.)
2059 PH/44 There was no timeout on the connect() call when using a Unix domain
2060 socket in the ${readsocket expansion. There now is.
2062 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2063 be meaningful with "accept".
2065 SC/01 Eximstats V1.43
2066 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2068 SC/02 Eximstats V1.44
2069 Use a glob alias rather than an array ref in the generated
2070 parser. This improves both readability and performance.
2072 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2073 Collect SpamAssassin and rejection statistics.
2074 Don't display local sender or destination tables unless
2075 there is data to show.
2076 Added average volumes into the top table text output.
2078 SC/04 Eximstats V1.46
2079 Collect data on the number of addresses (recipients)
2080 as well as the number of messages.
2082 SC/05 Eximstats V1.47
2083 Added 'Message too big' to the list of mail rejection
2084 reasons (thanks to Marco Gaiarin).
2086 SC/06 Eximstats V1.48
2087 Mainlog lines which have GMT offsets and are too short to
2088 have a flag are now skipped.
2090 SC/07 Eximstats V1.49 (Alain Williams)
2091 Added the -emptyok flag.
2093 SC/08 Eximstats V1.50
2094 Fixes for obtaining the IP address from reject messages.
2096 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2097 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2098 whitesspace changes from 4.64-PH/27
2100 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2103 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2104 are found, allow negative numbers in numeric criteria)
2106 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2108 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2109 to match changes made in 4.64-PH/43
2111 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2113 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2114 "domain" is actually an IP address, and force a failure. This locks out
2115 those revolvers/nameservers that support "A-for-A" lookups, in
2116 contravention of the specifications.
2118 PH/48 When a host name was looked up from an IP address, and the subsequent
2119 forward lookup of the name timed out, the host name was left in
2120 $sender_host_name, contrary to the specification.
2122 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2123 restricted to single-key lookups, Exim was not diagnosing an error if
2124 * or *@ was used with a query-style lookup.
2126 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2128 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2129 long ago, but noone interested enough thought of it. Let's just say that
2130 the "1.1" means that there are some new functions that weren't there at
2131 some point in the past.
2133 PH/51 Error processing for expansion failure of helo_data from an smtp
2134 transport during callout processing was broken.
2136 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2137 tested/used via the -bh/-bhc/-bs options.
2139 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2140 bug, fixed in subsequent PCRE releases).
2142 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2143 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2145 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2151 SC/01 Use a glob alias rather than an array ref in eximstats generated
2152 parser. This improves both readability and performance.
2154 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2155 Don't display local sender or destination tables in eximstats unless
2156 there is data to show.
2157 Added average volumes into the eximstats top table text output.
2159 SC/03 Collect data on the number of addresses (recipients) as well
2160 as the number of messages in eximstats.
2162 TF/01 Correct an error in the documentation for the redirect router. Exim
2163 does (usually) call initgroups() when daemonizing.
2165 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2166 with consistent privilege compared to when running as a daemon.
2168 TF/03 Note in the spec that $authenticated_id is not set for local
2169 submissions from trusted users.
2171 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2172 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2174 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2175 by adding some example configuration directives to the default
2176 configuration file. A little bit of work is required to uncomment the
2177 directives and define how usernames and passwords are checked, but
2178 there is now a framework to start from.
2180 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2181 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2182 without this. I don't know how relevant this is to other LDAP libraries.
2184 PH/02 Add the verb name to the "unknown ACL verb" error.
2186 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2188 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2190 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2191 directory not expanded when it should be if an expanded home directory
2192 was set for the address (which is overridden by the transport).
2194 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2197 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2198 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2199 because it is too late at that time, and has no effect.
2201 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2202 security issue with \' (bugzilla #107). I could not use the
2203 PQescapeStringConn() function, because it needs a PGconn value as one of
2206 PH/08 When testing addresses using -bt, indicate those final addresses that
2207 are duplicates that would not cause an additional delivery. At least one
2208 person was confused, thinking that -bt output corresponded to deliveries.
2209 (Suppressing duplicates isn't a good idea as you lose the information
2210 about possibly different redirections that led to the duplicates.)
2212 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2213 systems where poll() doesn't work, in particular OS X.
2215 PH/10 Added more information to debugging output for retry time not reached.
2217 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2218 operations in malware.c.
2220 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2223 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2224 syslog_duplication set false, log lines that would normally be written
2225 both the the main log and to the reject log were not written to syslog at
2228 PH/14 In the default configuration, change the use of "message" in ACL warn
2229 statements to "add_header".
2231 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2232 not followed by a command (e.g. "seen endif").
2234 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2235 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2238 PH/17 Added extra conditions to the default value of delay_warning_condition
2242 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2243 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2244 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2247 The Auto-Submitted: and various List- headers are standardised, whereas I
2248 don't think Precedence: ever was.
2250 PH/18 Refactored debugging code in route_finduser() to show more information,
2251 in particular, the error code if getpwnam() issues one.
2253 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2254 This is apparently needed in addition to the PH/07 change above to avoid
2255 any possible encoding problems.
2257 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2258 but not after initializing Perl.
2260 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2261 output them only if debugging. By default they are written stderr,
2262 apparently, which is not desirable.
2264 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2267 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2270 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2272 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2273 authenticated or an ident call has been made. Suppress the default
2274 values for $authenticated_id and $authenticated_sender (but permit -oMai
2275 and -oMas) when testing with -bh.
2277 PH/24 Re-jigged the order of the tests in the default configuration so that the
2278 tests for valid domains and recipients precede the DNS black list and CSA
2279 tests, on the grounds that those ones are more expensive.
2281 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2282 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2283 This bug exists in every version of Exim that I still have, right back to
2286 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2287 However, an attempt to turn on case-sensitivity in a regex key by
2288 including (?-i) didn't work because the subject string was already
2289 lowercased, and the effects were non-intuitive. It turns out that a
2290 one-line patch can be used to allow (?-i) to work as expected.
2296 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2297 other effects) broke the use of negated acl sub-conditions.
2299 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2302 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2303 "Deny" causes Exim to reject the incoming connection with a 554 error.
2304 Unfortunately, if there is a major crisis, such as a disk failure,
2305 tcp-wrappers gives "deny", whereas what one would like would be some
2306 kind of temporary error. A kludge has been added to help with this.
2307 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2308 554 error is used if errno is still zero or contains ENOENT (which occurs
2309 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2312 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2314 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2315 errors. Otherwise a message that provokes a temporary error (when other
2316 messages do not) can cause a whole host to time out.
2318 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2319 addresses were routed directly to files or pipes from a redirect router.
2320 File deliveries just didn't batch; pipe deliveries might have suffered
2323 PH/06 A failure to get a lock for a hints database would erroneously always say
2324 "Failed to get write lock", even when it was really a read lock.
2326 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2327 of 0600. This has been changed to use the value of the lockfile_mode
2328 option (which defaults to 0600).
2330 PH/08 Applied small patch from the Sieve maintainer.
2332 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2333 folder from quota calculations, a direct delivery into this folder messed
2334 up the contents of the maildirsize file. This was because the regex was
2335 used only to exclude .Trash (or whatever) when the size of the mailbox
2336 was calculated. There was no check that a delivery was happening into an
2337 excluded directory. This bug has been fixed by ignoring all quota
2338 processing for deliveries into excluded directories.
2340 PH/10 Added the maildirfolder_create_regex option to appendfile.
2346 PH/01 The code for finding all the local interface addresses on a FreeBSD
2347 system running IPv6 was broken. This may well have applied to all BSD
2348 systems, as well as to others that have similar system calls. The broken
2349 code found IPv4 interfaces correctly, but gave incorrect values for the
2350 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2351 that it would not match correctly against @[] and not recognize the IPv6
2354 PH/02 The ipliteral router was not recognizing addresses of the form user@
2355 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2357 PH/03 Added disable_ipv6.
2359 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2360 decoded form, because it is most often used to construct To: headers
2361 lines in autoreplies, and the decoded form may well be syntactically
2362 invalid. However, $reply_address has leading white space removed, and all
2363 newlines turned into spaces so that the autoreply transport does not
2366 PH/05 If group was specified without a user on a router, and no group or user
2367 was specified on a transport, the group from the router was ignored.
2369 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2370 for visible compile-time settings that can be used to change these
2371 numbers, for those that want even more. Backwards compatibility with old
2372 spool files has been maintained. However, going back to a previous Exim
2373 release will lost any variables that are in spool files.
2375 PH/07 Two small changes when running in the test harness: increase delay when
2376 passing a TCP/IP connection to a new process, in case the original
2377 process has to generate a bounce, and remove special handling of
2378 127.0.0.2 (sic), which is no longer necessary.
2380 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2381 be the same on different OS.
2383 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2386 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2387 whether --show-vars was specified or not
2389 JJ/02 exipick: Added support for new ACL variable spool format introduced
2392 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2393 syntactically invalid From: or Reply-to: line, and a filter used this to
2394 generate an autoreply, and therefore failed to obtain an address for the
2395 autoreply, Exim could try to deliver to a non-existent relative file
2396 name, causing unrelated and misleading errors. What now happens is that
2397 it logs this as a hard delivery error, but does not attempt to create a
2400 PH/11 The exinext utility has a -C option for testing purposes, but although
2401 the given file was scanned by exinext itself; it wasn't being passed on
2402 when Exim was called.
2404 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2405 an end-of-file indication when reading a command response.
2407 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2408 compiled. In many other places in Exim, IPv6 addresses are always
2409 recognized, so I have changed this. It also means that IPv4 domain
2410 literals of the form [IPV4:n.n.n.n] are now always recognized.
2412 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2413 used if the router is not running as root, for example, when verifying at
2414 ACL time, or when using -bh. The debugging output from this situation was
2415 non-existent - all you got was a failure to exec. I have made two
2418 (a) Failures to set uid/gid, the current directory, or a process leader
2419 in a subprocess such as that created by queryprogram now generate
2420 suitable debugging ouput when -d is set.
2422 (b) The queryprogram router detects when it is not running as root,
2423 outputs suitable debugging information if -d is set, and then runs
2424 the subprocess without attempting to change uid/gid.
2426 PH/15 Minor change to Makefile for building test_host (undocumented testing
2429 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2430 additional section of a DNS packet that returns MX or SRV records.
2431 Instead, it always explicitly searches for A/AAAA records. This avoids
2432 major problems that occur when a DNS server includes only records of one
2433 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2434 fixed another bug: if SRV records were looked up and the corresponding
2435 address records were *not* found in the additional section, the port
2436 values from the SRV records were lost.
2438 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2439 using the correct key (the original address) when searching the retry
2440 rules in order to find which one to use for generating the retry hint.
2442 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2443 adding the default one. Similarly, if it contains a Reply-To: header, the
2444 errors_reply_to option, if set, is not used.
2446 PH/19 When calculating a retry time, Exim used to measure the "time since
2447 failure" by looking at the "first failed" field in the retry record. Now
2448 it does not use this if it is later than than the arrival time of the
2449 message. Instead it uses the arrival time. This makes for better
2450 behaviour in cases where some deliveries succeed, thus re-setting the
2451 "first failed" field. An example is a quota failure for a huge message
2452 when small messages continue to be delivered. Without this change, the
2453 "time since failure" will always be short, possible causing more frequent
2454 delivery attempts for the huge message than are intended.
2455 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2457 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2458 $1, $2, $3) because the numerical variables can be reset during some
2459 expansion items (e.g. "match"), thereby losing the authentication data.
2461 PH/21 Make -bV show the size of off_t variables so that the test suite can
2462 decide whether to run tests for quotas > 2G.
2464 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2465 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2466 filecount value is greater than 2G or if a quota value is greater than 2G
2467 on a system where the size of off_t is not greater than 4, a panic error
2470 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2471 never match. The debug and -bh output now contains an explicit error
2472 message indicating a malformed IPv4 address or mask.
2474 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2475 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2476 PH/23 above applies.
2478 PH/25 Do not write to syslog when running in the test harness. The only
2479 occasion when this arises is a failure to open the main or panic logs
2480 (for which there is an explicit test).
2482 PH/26 Added the /no_tell option to "control=freeze".
2484 PH/27 If a host name lookup failed very early in a connection, for example, if
2485 the IP address matched host_lookup and the reverse lookup yielded a name
2486 that did not have a forward lookup, an error message of the form "no IP
2487 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2488 could be logged. Now it outputs the IP address instead of "NULL".
2490 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2491 allows the sender and the authenticated sender to be set when
2492 submitting a message from within Exim. Since child_open_exim() is
2493 documented for local_scan(), the new function should be too.
2495 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2496 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2497 results in an empty string is now treated as unset.
2499 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2501 PH/31 Added sender_verify_fail logging option.
2503 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2504 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2505 tidied the source and removed it altogether.
2507 PH/33 When a queue run was abandoned because the load average was too high, a
2508 log line was always written; now it is written only if the queue_run log
2509 selector is set. In addition, the log line for abandonment now contains
2510 information about the queue run such as the pid. This is always present
2511 in "start" and "stop" lines but was omitted from the "abandon" line.
2513 PH/34 Omit spaces between a header name and the colon in the error message that
2514 is given when verify = headers_syntax fails (if there are lots of them,
2515 the message gets confusing).
2517 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2518 names, as there are now some PTR records that contain slashes. This check
2519 is only to protect against broken name servers that fall over on strange
2520 characters, so the fact that it applies to all lookups doesn't matter.
2522 PH/36 Now that the new test suite is complete, we can remove some of the
2523 special code in Exim that was needed for the old test suite. For example,
2524 sorting DNS records because real resolvers return them in an arbitrary
2525 order. The new test suite's fake resolver always returns records in the
2528 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2529 bounces) except when queue_only is set, to avoid logging races between
2530 the different processes.
2532 PH/38 Panic-die if .include specifies a non-absolute path.
2534 PH/39 A tweak to the "H" retry rule from its user.
2536 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2537 a label. They prevented compilation on older perls.
2539 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2540 a warning to be raised on newish perls.
2542 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2543 on queue. Changes to match documented behaviour of showing count of
2544 messages matching specified criteria.
2546 PH/40 Changed the default ident timeout from 30s to 5s.
2548 PH/41 Added support for the use of login_cap features, on those BSD systems
2549 that have them, for controlling the resources used by pipe deliveries.
2551 PH/42 The content-scanning code uses fopen() to create files in which to put
2552 message data. Previously it was not paying any attention to the mode of
2553 the files. Exim runs with umask(0) because the rest of the code creates
2554 files with open(), and sets the required mode explicitly. Thus, these
2555 files were ending up world-writeable. This was not a big issue, because,
2556 being within the spool directory, they were not world-accessible. I have
2557 created a function called modefopen, which takes an additional mode
2558 argument. It sets umask(777), creates the file, chmods it to the required
2559 mode, then resets the umask. All the relevant calls to fopen() in the
2560 content scanning code have been changed to use this function.
2562 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2563 to 24 hours. This avoids potential overflow problems when processing G
2564 and H retry rules. I suspect nobody ever tinkers with this value.
2566 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2568 PH/45 When the plaintext authenticator is running as a client, the server's
2569 challenges are checked to ensure they are valid base64 strings. By
2570 default, the authentication attempt is cancelled if an invalid string is
2571 received. Setting client_ignore_invalid_base64 true ignores these errors.
2572 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2573 they are received. Thus, the responses can be made to depend on the
2574 challenges. If an invalid string is ignored, an empty string is placed in
2577 PH/46 Messages that are created by the autoreply transport now contains a
2578 References: header, in accordance with RFCs 2822 and 3834.
2580 PH/47 Added authenticated_sender_force to the smtp transport.
2582 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2584 PH/49 Installed latest patch from the Sieve maintainer.
2586 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2587 was also set, the appendfile transport was unnecessarily scanning a
2588 directory of message files (e.g. for maildir delivery) to find the count
2589 of files (along with the size), even though it did not need this
2590 information. It now does the scan only if it needs to find either the
2591 size of the count of files.
2593 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2595 PH/52 Two bugs concerned with error handling when the smtp transport is
2598 (i) Exim was not creating retry information for temporary errors given
2599 for individual recipients after the DATA command when the smtp transport
2600 was used in LMTP mode. This meant that they could be retried too
2601 frequently, and not timed out correctly.
2603 (ii) Exim was setting the flag that allows error details to be returned
2604 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2605 recipients that were returned after the DATA command.
2607 PH/53 This is related to PH/52, but is more general: for any failing address,
2608 when detailed error information was permitted to be returned to the
2609 sender, but the error was temporary, then after the final timeout, only
2610 "retry timeout exceeded" was returned. Now it returns the full error as
2611 well as "retry timeout exceeded".
2613 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2614 do this, and (what is worse) MTAs that accept it.
2616 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2617 will now be deprecated.
2619 PH/56 New os.c-cygwin from the Cygwin maintainer.
2621 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2622 formats (previously only available in exim formats via -bpr, -bpru,
2623 and -bpra. Now also available in native and exiqgrep formats)
2625 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2626 with very large, slow to parse queues
2628 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2630 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2632 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2633 responses to authentication challenges, though it was showing the
2634 challenges; (ii) I've removed the CR characters from the debug output for
2637 PH/58 Allow for the insertion of a newline as well as a space when a string
2638 is turned into more than one encoded-word during RFC 2047 encoding. The
2639 Sieve code now uses this.
2641 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2642 data_4xx, lost_connection, tls_required.
2644 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2645 message was being sent as an SMTP response.
2647 PH/61 Add -l and -k options to exicyclog.
2649 PH/62 When verifying, if an address was redirected to one new address, so that
2650 verification continued, and the new address failed or deferred after
2651 having set something in $address_data, the value of $address_data was not
2652 passed back to the ACL. This was different to the case when no
2653 redirection occurred. The value is now passed back in both cases.
2655 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2656 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2657 use login_cap.h, so on its own it isn't the distinguishing feature. The
2658 new name refers directly to the setclassresources() function.
2660 PH/65 Added configuration files for NetBSD3.
2662 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2664 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2665 is preferred over IPv4.
2667 PH/68 The bounce_return_message and bounce_return_body options were not being
2668 honoured for bounces generated during the reception of non-SMTP messages.
2669 In particular, this applied to messages rejected by the ACL. This bug has
2670 been fixed. However, if bounce_return_message is true and bounce_return_
2671 body is false, the headers that are returned for a non-SMTP message
2672 include only those that have been read before the error was detected.
2673 (In the case of an ACL rejection, they have all been read.)
2675 PH/69 The HTML version of the specification is now built in a directory called
2676 spec_html instead of spec.html, because the latter looks like a path with
2677 a MIME-type, and this confuses some software.
2679 PH/70 Catch two compiler warnings in sieve.c.
2681 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2682 function verify_get_ident() calls ip_connect() to connect a socket, but
2683 if the "connect()" function timed out, ip_connect() used to close the
2684 socket. However, verify_get_ident() also closes the socket later, and in
2685 between Exim writes to the log, which may get opened at this point. When
2686 the socket was closed in ip_connect(), the log could get the same file
2687 descriptor number as the socket. This naturally causes chaos. The fix is
2688 not to close the socket in ip_connect(); the socket should be closed by
2689 the function that creates it. There was only one place in the code where
2690 this was missing, in the iplookup router, which I don't think anybody now
2691 uses, but I've fixed it anyway.
2693 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2694 well as to direct DNS lookups. Otherwise the handling of names in host
2695 lists is inconsistent and therefore confusing.
2701 PH/01 Two changes to the default runtime configuration:
2703 (1) Move the checks for relay_from_hosts and authenticated clients from
2704 after to before the (commented out) DNS black list checks.
2706 (2) Add control=submission to the relay_from_hosts and authenticated
2707 clients checks, on the grounds that messages accepted by these
2708 statements are most likely to be submissions.
2710 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2712 (1) Generate an error if the third argument for the ${prvs expansion is
2715 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2718 (3) Reset the variables that are obtained from the first argument of
2719 ${prvscheck and used in the second argument before leaving the code,
2720 because their memory is reclaimed, so using them afterwards may do
2723 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2724 one (it's much easier than Tom thought :-).
2726 (5) Because of (4), we can now allow for the use of $prvscheck_result
2727 inside the third argument.
2729 PH/03 For some reason, the default setting of PATH when running a command from
2730 a pipe transport was just "/usr/bin". I have changed it to
2733 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2734 anything to be listed in the output from -bV.
2736 PH/05 When a filter generated an autoreply, the entire To: header line was
2737 quoted in the delivery log line, like this:
2739 => >A.N.Other <ano@some.domain> <original@ddress> ...
2741 This has been changed so that it extracts the operative address. There
2742 may be more than one such address. If so, they are comma-separated, like
2745 => >ano@some.domain,ona@other.domain <original@ddress> ...
2747 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2748 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2749 not being looked up in the rDNS to get a host name, Exim was showing the
2750 IP address twice in Received: lines, even though the IP addresses were
2751 identical. For example:
2753 Received: from [1.2.3.4] (helo=[1.2.3.4])
2755 However, if the real host name was known, it was omitting the HELO data
2756 if it matched the actual IP address. This has been tidied up so that it
2757 doesn't show the same IP address twice.
2759 PH/07 When both +timestamp and +memory debugging was on, the value given by
2760 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2761 called by the debug printing, thereby overwriting the timestamp buffer.
2762 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2764 PH/08 When the original message was included in an autoreply transport, it
2765 always said "this is a copy of the message, including all the headers",
2766 even if body_only or headers_only was set. It now gives an appropriate
2769 PH/09 Applied a patch from the Sieve maintainer which:
2771 o fixes some comments
2772 o adds the (disabled) notify extension core
2773 o adds some debug output for the result of if/elsif tests
2774 o points to the current vacation draft in the documentation
2775 and documents the missing references header update
2779 o fixes a bug in processing the envelope test (when testing
2780 multiple envelope elements, the last element determinted the
2783 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2784 Electronic Mail") by including:
2786 Auto-submitted: auto-generated
2788 in the messages that it generates (bounce messages and others, such as
2789 warnings). In the case of bounce messages for non-SMTP mesages, there was
2790 also a typo: it was using "Auto_submitted" (underscore instead of
2791 hyphen). Since every message generated by Exim is necessarily in response
2792 to another message, thes have all been changed to:
2794 Auto-Submitted: auto-replied
2796 in accordance with these statements in the RFC:
2798 The auto-replied keyword:
2800 - SHOULD be used on messages sent in direct response to another
2801 message by an automatic process,
2803 - MUST NOT be used on manually-generated messages,
2805 - MAY be used on Delivery Status Notifications (DSNs) and Message
2806 Disposition Notifications (MDNs),
2808 - MUST NOT be used on messages generated by automatic or periodic
2809 processes, except for messages which are automatic responses to
2812 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2813 to the default Received: header definition.
2815 PH/12 Added log selector acl_warn_skipped (default on).
2817 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2818 variables because (a) they are in the wrong storage pool and (b) even if
2819 they were copied, it wouldn't work properly because of the caching.
2821 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2822 checking when decoding. Apparently there are clients that generate
2823 overlong encoded strings. Why am I not surprised?
2825 PH/15 If the first argument of "${if match_address" was not empty, but did not
2826 contain an "@" character, Exim crashed. Now it writes a panic log message
2827 and treats the condition as false.
2829 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2831 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2832 extension "envelope-auth". The code is finished and in agreement with
2833 other implementations, but there is no documentation so far and in fact,
2834 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2835 not changing the active code.
2837 Print executed "if" and "elsif" statements when debugging is used. This
2838 helps a great deal to understand what a filter does.
2840 Document more things not specified clearly in RFC3028. I had all this
2841 sorted out, when out of a sudden new issues came to my mind. Oops."
2843 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2846 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2847 canonical form (as documented). However, after a host name lookup from
2848 the IP address, check_host() was doing a simple string comparison with
2849 addresses acquired from the DNS when checking that the found name did
2850 have the original IP as one of its addresses. Since any found IPv6
2851 addresses are likely to be in abbreviated form, the comparison could
2852 fail. Luckily, there already exists a function for doing the comparison
2853 by converting both addresses to binary, so now that is used instead of
2854 the text comparison.
2856 PH/20 There was another similar case to PH/19, when a complete host name was
2857 given in a host list; looking up its IP address could give an abbreviated
2858 form, whereas the current host's name might or might not be abbreviated.
2859 The same fix has been applied.
2865 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2866 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2869 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2870 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2872 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2874 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2875 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2876 Also, for "Auto-Submitted", treat anything other than "no" as
2877 non-personal, in accordance with RFC 3834. (Previously it treated
2878 anything starting "auto-" as non-personal.)
2880 TF/01 The control=submission/name=... option had a problem with syntax
2881 errors if the name included a slash character. The /name= option
2882 now slurps the rest of the string, so it can include any characters
2883 but it must come last in the list of options (after /sender_retain
2886 PH/05 Some modifications to the interface to the fake nameserver for the new
2894 TK/01 Added the "success_on_redirect" address verification option. See
2895 NewStuff for rationale and an example.
2897 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2899 PH/02 Patch to exigrep to allow it to work on syslog lines.
2901 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2902 fread() to skip over the body file's header line, because in Cygwin the
2903 header line is locked and is inaccessible.
2905 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2906 co-exist for some time) to make it clear that it is the Exim ID that is
2907 referenced, not the Message-ID: header line.
2909 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2910 string_format() function, because snprintf() does not exist on all
2913 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2914 ${stat: expansion item.
2916 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2917 protocol synchronization error", to keep the pedants happy.
2919 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2920 well as for IRIX systems, when gcc is being used. See the host.c source
2923 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2925 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2928 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2929 transport and to the smtp transport in LMTP mode.
2931 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2933 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2934 run a filter in a subprocess. This could lead to confusion in subsequent
2935 lookups in the parent process. There should also be a search_tidyup() at
2936 the end of the subprocess.
2938 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2939 only if the host matched helo_try_verify_hosts, which caused the
2940 verification to occur when the EHLO/HELO command was issued. The ACL just
2941 tested the remembered result. Now, if a previous verification attempt has
2942 not happened, "verify = helo" does it there and then.
2944 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2946 TK/03 Fix log output including CR from clamd.
2948 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2949 exist provoked a memory error which could cause a segfault.
2951 PH/15 Installed PCRE 6.2
2953 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2955 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2956 of the problem. Specifically, suggested +O2 rather than +O1 for the
2959 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2961 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2962 record, the port was not correctly logged when the outgoing_port log
2963 selector was set (it logged the transort's default port).
2965 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2966 fallback_hosts, and "hosts" in the smtp transport.
2968 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2969 host errors such as "Connection refused".
2971 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2972 authentication with radiusclient 0.4.9:
2974 - Error returned from rc_read_config was caught wrongly
2975 - Username/password not passed on to radius server due to wrong length.
2977 The presumption is that some radiusclient API changes for 4.51/PH/17
2978 were not taken care of correctly. The code is still untested by me (my
2979 Linux distribution still has 0.3.2 of radiusclient), but it was
2980 contributed by a Radius user.
2982 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2983 expanding the "port" option of the smtp transport.
2985 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2986 while reading a MIME header. Thanks to Tom Hughes for a patch.
2988 PH/24 Include config.h inside local_scan.h so that configuration settings are
2991 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2992 that in an ACL for RCPT (for example), you can examine exactly what was
2995 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2996 commands, but it was not correctly comparing the address with the actual
2997 client host address. Thus, it would show the EHLO address in Received:
2998 header lines when this was not necessary.
3000 PH/27 Added the % operator to ${eval:}.
3002 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3003 it should be ignoring failures (because with -C, for example, it has lost
3004 privilege). It wasn't ignoring creation failures other than "already
3007 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3010 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3011 because an input line was too long, either on its own, or by virtue of
3012 too many continuations, the temporary file was not being removed, and the
3013 return code was incorrect.
3015 PH/31 Missing "BOOL" in function definition in filtertest.c.
3017 PH/32 Applied Sieve patches from the maintainer.
3019 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3021 PH/33 Added "verify = not_blind".
3023 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3024 Local/Makefile (with some defaults set). These are used in built scripts
3025 such as exicyclog, but they have never been used in the exim_install
3026 script (though there are many overriding facilities there). I have
3027 arranged that the exim_install script now takes note of these two
3030 PH/35 Installed configuration files for Dragonfly.
3032 PH/36 When a locally submitted message by a trusted user did not contain a
3033 From: header, and the sender address was obtained from -f or from an SMTP
3034 MAIL command, and the trusted user did not use -F to supply a sender
3035 name, $originator_name was incorrectly used when constructing a From:
3036 header. Furthermore, $originator_name was used for submission mode
3037 messages from external hosts without From: headers in a similar way,
3038 which is clearly wrong.
3040 PH/37 Added control=suppress_local_fixups.
3042 PH/38 When log_selector = +received_sender was set, and the addition of the
3043 sender made the log line's construction buffer exactly full, or one byte
3044 less than full, an overflow happened when the terminating "\n" was
3047 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3048 when the result of a list match is failure because a DNS lookup failed.
3050 PH/40 RM_COMMAND is now used in the building process.
3052 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3053 the "build-* directories that it finds.
3055 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3056 address in a domain literal was a prefix of an interface address.
3058 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3059 when verifying a sender address, unless rewrite_headers is false.
3061 PH/44 Wrote a long comment about why errors_to addresses are verified as
3062 recipients, not senders.
3064 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3065 the ratelimit ACL was added.
3067 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3069 PH/46 Added extra information about PostgreSQL errors to the error string.
3071 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3072 avoiding the need to install special zones in a real server. This is
3073 backwards compatible; if it can't find the fake resolver, it drops back.
3074 Thus, both old and new test suites can be run.
3076 TF/02 Added util/ratelimit.pl
3078 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3079 clock is set back in time.
3081 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3082 Candler <B.Candler@pobox.com>.
3084 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3085 OK for addresses that are the result of redirections.
3087 PH/48 A number of further additions for the benefit of the new test suite,
3088 including a fake gethostbyname() that interfaces to the fake DNS resolver
3091 TF/06 The fix for widen_domains has also been applied to qualify_single and
3092 search_parents which are the other dnslookup options that can cause
3095 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3098 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3100 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3101 service" if it encountered an MX record with an empty target hostname.
3102 The message is now "an MX or SRV record indicated no SMTP service".
3104 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3105 if the DNS of the sending site is misconfigured. This is quite a
3106 common situation. This change restores the behaviour of treating a
3107 helo verification defer as a failure.
3109 PH/51 If self=fail was set on a router, the bounce message did not include the
3110 actual error message.
3116 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3118 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3119 command itself ended in error, the underlying message about the transport
3120 filter timeout was being overwritten with the pipe command error. Now the
3121 underlying error message should be appended to the second error message.
3123 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3125 PH/02 Reduce the amount of output that "make" produces by default. Full output
3126 can still be requested.
3128 PH/03 The warning log line about a condition test deferring for a "warn" verb
3129 was being output only once per connection, rather than after each
3130 occurrence (because it was using the same function as for successful
3131 "warn" verbs). This seems wrong, so I have changed it.
3133 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3134 it should not have, which might have caused a crash in the right
3135 circumstances, but probably never did.
3137 PH/04 Installed a modified version of Tony Finch's patch to make submission
3138 mode fix the return path as well as the Sender: header line, and to
3139 add a /name= option so that you can make the user's friendly name appear
3142 TF/03 Added the control = fakedefer ACL modifier.
3144 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3145 Mark Lowes for thorough testing.
3147 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3149 TK/03 Merged latest SRS patch from Miles Wilton.
3151 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3152 with the definition in sysexits.h (which is #included earlier).
3153 Fortunately, Exim does not actually use EX_OK. The code used to try to
3154 preserve the sysexits.h value, by assumimg that macro definitions were
3155 scanned for macro replacements. I have been disabused of this notion,
3156 so now the code just undefines EX_OK before #including unistd.h.
3158 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3159 in the smtp transport. When a block could not be written in a single
3160 write() function, the timeout was being re-applied to each part-write.
3161 This seems wrong - if the receiver was accepting one byte at a time it
3162 would take for ever. The timeout is now adjusted when this happens. It
3163 doesn't have to be particularly precise.
3165 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3166 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3168 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3169 without a domain if the check to <postmaster@domain> fails.
3171 SC/01 Eximstats: added -xls and the ability to specify output files
3172 (patch written by Frank Heydlauf).
3174 SC/02 Eximstats: use FileHandles for outputing results.
3176 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3178 SC/04 Eximstats: fixed display of large numbers with -nvr option
3180 SC/05 Eximstats: fixed merging of reports with empty tables.
3182 SC/06 Eximstats: added the -include_original_destination flag
3184 SC/07 Eximstats: removed tabs and trailing whitespace.
3186 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3188 TK/06 MBOX spool code: Add real "From " MBOX separator line
3189 so the .eml file is really in mbox format (even though
3190 most programs do not really care). Patch from Alex Miller.
3192 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3193 The latter is generated from $received_to and is only set if the
3194 message has one envelope recipient. SA can use these headers,
3195 obviously out-of-the-box. Patch from Alex Miller.
3197 PH/08 The ${def test on a variable was returning false if the variable's
3198 value was "0", contrary to what the specification has always said!
3199 The result should be true unless the variable is empty.
3201 PH/09 The syntax error of a character other than { following "${if
3202 def:variable_name" (after optional whitespace) was not being diagnosed.
3203 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3204 accidental colon was present, for example, could give incorrect results.
3206 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3207 result is used (not including appendfile, where other changes are about
3210 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3211 This involved changing a lot of size variables from int to off_t. It
3212 should work with maildirs and everything.
3214 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3215 spamd dying while we are connected to it.
3217 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3220 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3221 function for BDB 4.3.
3223 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3225 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3226 the sources for PCRE have been re-arranged and more files are now
3229 PH/15 The code I had for printing potentially long long variables in PH/11
3230 above was not the best (it lost precision). The length of off_t variables
3231 is now inspected at build time, and an appropriate printing format (%ld
3232 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3233 to be "long long int" or "long int". This is needed for the internal
3234 formatting function string_vformat().
3236 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3237 the configuration file to be ":syslog", then the script "guesses" where
3238 the logs files are, rather than using the compiled in default. In our
3239 case the guess is not the same as the compiled default, so the script
3240 suddenly stopped working when I started to use syslog. The patch checks
3241 to see if log_file_path is "". If so, it attempts to read it from exim
3242 with no configuration file to get the compiled in version, before it
3243 falls back to the previous guessing code."
3245 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3246 implementing BATV in an Exim configuration. See NewStuff for the gory
3249 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3250 Makefile that are specific to HP-UX.
3252 PH/18 If the "use_postmaster" option was set for a recipient callout together
3253 with the "random" option, the postmaster address was used as the MAIL
3254 FROM address for the random test, but not for the subsequent recipient
3255 test. It is now used for both.
3257 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3258 patch removes a few documentation additions to RFC 3028, because the
3259 latest draft now contains them. It adds the new en;ascii-case comparator
3260 and a new error check for 8bit text in MIME parts. Comparator and
3261 require names are now matched exactly. I enabled the subaddress
3262 extension, but it is not well tested yet (read: it works for me)."
3264 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3265 rework some of the code of TK/09 above to avoid the hardwired use of
3266 "%lld" and "long long". Replaced the call to snprintf() with a call to
3269 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3270 records point to non-existent hosts", "retry timeout exceeded", and
3271 "retry time not reached for any host after a long failure period".
3273 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3274 experimental DomainKeys support:
3276 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3277 (2) On an error such as an illegally used "control", the wrong name for
3278 the control was given.
3280 These problems did NOT occur unless DomainKeys support was compiled.
3282 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3284 PH/24 Added ${if match_ip condition.
3286 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3287 databases so that it will be absolutely obvious if a crash occurs in the
3288 DB library. This is a regular occurrence (often caused by mis-matched
3291 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3292 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3293 picked up on a user's system that detects such things. There doesn't seem
3294 to be a gcc warning option for this - only an attribute that has to be
3295 put on the function's prototype. It seems that in Fedora Core 4 they have
3296 set this on a number of new functions. No doubt there will be more in due
3299 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3300 specify a transport. However, if an address that was verified by such a
3301 router was the subject of a callout, Exim crashed because it tried to
3302 read the rcpt_include_affixes from the non-existent transport. Now it
3303 just assumes that the setting of that option is false. This bug was
3304 introduced by 4.51/PH/31.
3306 PH/28 Changed -d+all to exclude +memory, because that information is very
3307 rarely of interest, but it makes the output a lot bigger. People tend to
3308 do -d+all out of habit.
3310 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3311 code in os-type was giving problems when libc.so lives in lib64, like on
3314 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3315 aren't the modern standard, and it seems that some systems' include files
3316 don't always have them. Exim was already checking for some of the newer
3317 ones like T_AAAA, and defining it itself. I've added checks for all the
3318 record types that Exim uses.
3320 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3321 not automatically generating a new one, as it is supposed to. This
3322 prevented TLS from working. If the file did exist, but contained invalid
3323 data, a new version was generated, as expected. It was only the case of a
3324 non-existent file that was broken.
3326 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3327 with a change in libdomainkeys > 0.64.
3329 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3330 from DNS. If the selector record carries the flag, it now has
3331 precedence over the domain-wide flag.
3333 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3335 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3336 the use of a port name that isn't defined in /etc/services) occurred, the
3337 message was deferred as in a normal delivery, and thus remained on the
3338 spool, instead of being failed because of the mua_wrapper setting. This
3339 is now fixed, and I tidied up some of the mua_wrapper messages at the
3342 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3343 the messages in a hash of arrays rather than using individual hashes.
3344 This is a bit cleaner and results in dramatic memory savings, albeit
3345 at a slight CPU cost.
3347 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3348 as requested by Marc Sherman.
3350 SC/10 Eximstats: added histograms for user specified patterns as requested
3353 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3355 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3356 fopen() in the content-scanning modules that did not already have it.
3362 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3363 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3365 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3367 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3369 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3370 to test Sieve filters that use "vacation".
3372 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3373 that changes the way the GnuTLS parameters are stored in the cache file.
3374 The new format can be generated externally. For backward compatibility,
3375 if the data in the cache doesn't make sense, Exim assumes it has read an
3376 old-format file, and it generates new data and writes a new file. This
3377 means that you can't go back to an older release without removing the
3380 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3381 work if there are any delivery delays because "one_time" forces the
3382 parent to be marked "delivered", so its unseen clone is never tried
3383 again. For this reason, Exim now forbids the simultaneous setting of
3386 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3387 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3388 entry if you want to know the details. The fix, however, neglected to
3389 consider the case when local delivery batching is involved. The test for
3390 "previously delivered" was not happening when checking to see if an
3391 address could be batched with a previous (undelivered) one; under
3392 certain circumstances this could lead to multiple deliveries to the same
3395 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3396 in its include files, and this causes problems building Exim.
3398 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3399 header_syntax) but Exim was just ignoring anything given after a slash.
3400 In particular, this caused confusion with an attempt to use "verify =
3401 reverse_host_lookup/defer_ok". An error is now given when options are
3402 supplied for verify items that do not have them. (Maybe reverse_host_
3403 lookup should have a defer_ok option, but that's a different point.)
3405 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3406 defined by RFC 821) to 2048, because there were problems with some AUTH
3407 commands, and RFC 1869 says the size should be increased for extended
3408 SMTP commands that take arguments.
3410 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3413 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3414 "unknown" error; now it says that the functionality isn't in the binary.
3416 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3417 an address' error message when a string expansion fails (syntax or
3418 whatever). Otherwise the password may appear in the log. Following change
3419 PH/42 below, there is no longer a chance of it appearing in a bounce
3422 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3424 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3425 of its addresses was ever tried. (Bugzilla bug #2.)
3427 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3428 the result incorrectly in the debug output. (It correctly added a newline
3429 to what was transported.)
3431 TF/01 Added $received_time.
3433 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3434 commented out examples of how to interface to a virus scanner and to
3435 SpamAssassin. Also added commented examples of av_scanner and
3436 spamd_address settings.
3438 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3439 and controls are allowed in which ACLs. There were a couple of minor
3440 errors. Some of the entries in the conditions table (which is a table of
3441 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3442 negation of where the condition IS allowed.
3444 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3446 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3447 header file does not have a version number, so I've had to invent a new
3448 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3449 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3450 radiusclient), but it was contributed by a Radius user.
3452 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3453 files or directories, for OpenSSL.
3455 PH/19 When an Exim process that is running as root has to create an Exim log
3456 file, it does so in a subprocess that runs as exim:exim so as to get the
3457 ownership right at creation (otherwise, other Exim processes might see
3458 the file with the wrong ownership). There was no test for failure of this
3459 fork() call, which would lead to the process getting stuck as it waited
3460 for a non-existent subprocess. Forks do occasionally fail when resources
3461 run out. I reviewed all the other calls to fork(); they all seem to check
3464 PH/20 When checking for unexpected SMTP input at connect time (before writing
3465 the banner), Exim was not dealing correctly with a non-positive return
3466 from the read() function. If the client had disconnected by this time,
3467 the result was a log entry for a synchronization error with an empty
3468 string after "input=" when read() returned zero. If read() returned -1
3469 (an event I could not check), uninitialized data bytes were printed.
3470 There were reports of junk text (parts of files, etc) appearing after
3473 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3475 PH/22 Added support for macro redefinition, and (re)definition in between
3476 driver and ACL definitions.
3478 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3479 forgetting to use the resulting value; it was using the unexpanded value.
3481 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3482 hadn't been configured. The fix is from Juergen Kreileder, who
3483 understands it better than I do:
3485 "Here's what I see happening with three configured cyrus_sasl
3486 authenticators configured (plain, login, cram-md5):
3488 On startup auth_cyrus_sasl_init() gets called for each of these.
3489 This means three calls to sasl_listmech() without a specified mech_list.
3490 => SASL tests which mechs of all available mechs actually work
3491 => three warnings about OTP not working
3492 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3494 With the patch, sasl_listmech() also gets called three times. But now
3495 SASL's mech_list option is set to the server_mech specified in the the
3496 authenticator. Or in other words, the answer from sasl_listmech()
3497 gets limited to just the mech you're testing for (which is different
3499 => the return list contains just 'plain' or 'login', 'cram-md5' or
3500 nothing depending on the value of ob->server_mech.
3502 I've just tested the patch: Authentication still works fine,
3503 unavailable mechs specified in the exim configuration are still
3504 caught, and the auth.log warnings about OTP are gone."
3506 PH/25 When debugging is enabled, the contents of the command line are added
3507 to the debugging output, even when log_selector=+arguments is not
3510 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3511 answer is "GNU", and only if the return is "GNU/something" is the answer
3514 PH/27 $acl_verify_message is now set immediately after the failure of a
3515 verification in an ACL, and so is available in subsequent modifiers. In
3516 particular, the message can be preserved by coding like this:
3518 warn !verify = sender
3519 set acl_m0 = $acl_verify_message
3521 Previously, $acl_verify_message was set only while expanding "message"
3522 and "log_message" when a very denied access.
3524 PH/28 Modified OS/os.c-Linux with
3526 -#ifndef OS_LOAD_AVERAGE
3527 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3529 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3530 nomenclature these days.)
3532 PH/29 Installed patch from the Sieve maintainer that adds the options
3533 sieve_useraddress and sieve_subaddress to the redirect router.
3535 PH/30 In these circumstances:
3536 . Two addresses routed to the same list of hosts;
3537 . First host does not offer TLS;
3538 . First host accepts first address;
3539 . First host gives temporary error to second address;
3540 . Second host offers TLS and a TLS session is established;
3541 . Second host accepts second address.
3542 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3543 and peerdn, if requested) that were in fact used only for the second
3546 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3547 attention to any local part prefix or suffix that was matched by the
3548 router that accepted the address. It now behaves in the same way as it
3549 does for delivery: the affixes are removed from the local part unless
3550 rcpt_include_affixes is set on the transport.
3552 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3553 timeout during the DATA phase of an incoming message.
3555 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3556 applied a patch sanctioned by the Sieve maintainer.
3558 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3559 the uid or gid is negative. A case of a negative gid caused this to be
3560 noticed. The fix allows for either to be negative.
3562 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3563 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3566 PH/36 The change PH/12 above was broken. Fixed it.
3568 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3569 the grounds that routing the same address twice would always produce the
3570 same answer. This might have been true once, but it is certainly no
3571 longer true now. Routing a child address may depend on the previous
3572 routing that produced that child. Some complicated redirection strategies
3573 went wrong when messages had multiple recipients, and made Exim's
3574 behaviour dependent on the order in which the addresses were given.
3576 I have moved the duplicate checking until after the routing is complete.
3577 Exim scans the addresses that are assigned to local and remote
3578 transports, and removes any duplicates. This means that more work will be
3579 done, as duplicates will always all be routed, but duplicates are
3580 presumably rare, so I don't expect this is of any significance.
3582 For deliveries to pipes, files, and autoreplies, the duplicate checking
3583 still happens during the routing process, since they are not going to be
3586 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3587 It corrects a timeout issue with spamd. This is Ian's comment: "The
3588 background is that sometimes spamd either never reads data from a
3589 connection it has accepted, or it never writes response data. The exiscan
3590 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3591 blindly assumes that writes won't block so it may never time out."
3593 PH/39 Allow G after quota size as well as K and M.
3595 PH/40 The value set for $authenticated_id in an authenticator may not contain
3596 binary zeroes or newlines because the value is written to log lines and
3597 to spool files. There was no check on this. Now the value is run through
3598 the string_printing() function so that such characters are converted to
3599 printable escape sequences.
3601 PH/41 $message_linecount is a new variable that contains the total number of
3602 lines in the message. Compare $body_linecount, which is the count for the
3605 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3606 bounce and delay warning messages, except in certain special cases, which
3609 (a) An SMTP error message from a remote host;
3610 (b) A message specified in a :fail: redirection;
3611 (c) A message specified in a "fail" command in a system filter;
3612 (d) A message specified in a FAIL return from the queryprogram router;
3613 (e) A message specified by the cannot_route_message router option.
3615 In these cases only, Exim does include the error details in bounce and
3616 warning messages. There are also a few cases where bland messages such
3617 as "unrouteable address" or "local delivery error" are given.
3619 PH/43 $value is now also set for the "else" part of a ${run expansion.
3621 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3622 being worked on, but at least Exim now implements the latest version to
3625 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3626 process to complete was treated as a delivery failure, a timeout while
3627 writing the message to the pipe was logged, but erroneously treated as a
3628 successful delivery. Such timeouts include transport filter timeouts. For
3629 consistency with the overall process timeout, these timeouts are now
3630 treated as errors, giving rise to delivery failures by default. However,
3631 there is now a new Boolean option for the pipe transport called
3632 timeout_defer, which, if set TRUE, converts the failures into defers for
3633 both kinds of timeout. A transport filter timeout is now identified in
3636 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3637 systems where "make" and "gmake" are different, calling "gmake" at top
3638 level broke things. I've arranged for the value of $(MAKE) to be passed
3639 from the Makefile to this script so that it can call the same version of
3643 A note about Exim versions 4.44 and 4.50
3644 ----------------------------------------
3646 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3647 changes of various kinds. As a consequence, a big documentation update was
3648 needed. This delayed the release for rather longer than seemed good, especially
3649 in the light of a couple of (minor) security issues. Therefore, the changes
3650 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3651 release. So 4.44 and 4.50 are in effect two different branches that both start
3654 I have left the 4.50 change log unchanged; it contains all the changes since
3655 4.43. The change log for 4.44 is below; many of its items are identical to
3656 those for 4.50. This seems to be the most sensible way to preserve the
3657 historical information.
3663 1. Minor wording change to the doc/README.SIEVE file.
3665 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3666 computation of the current number of files was incorrect.
3668 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3669 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3672 4. Give more explanation in the error message when the command for a transport
3673 filter fails to execute.
3675 5. There are several places where Exim runs a non-Exim command in a
3676 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3677 was being done only for the command run by the queryprogram router. It is
3678 now done for all such subprocesses. The other cases are: ${run, transport
3679 filters, and the commands run by the lmtp and pipe transports.
3681 6. Added CONFIGURE_GROUP build-time option.
3683 7. Some older OS have a limit of 256 on the maximum number of file
3684 descriptors. Exim was using setrlimit() to set 1000 as a large value
3685 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3686 systems. I've change it so that if it can't get 1000, it tries for 256.
3688 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3689 was an oversight, and furthermore, ever since the addition of extra
3690 controls (e.g. 4.43/32), the checks on when to allow different forms of
3691 "control" were broken. There should now be diagnostics for all cases when a
3692 control that does not make sense is encountered.
3694 9. Added the /retain_sender option to "control=submission".
3696 10. $recipients is now available in the predata ACL (oversight).
3698 11. Tidy the search cache before the fork to do a delivery from a message
3699 received from the command line. Otherwise the child will trigger a lookup
3700 failure and thereby defer the delivery if it tries to use (for example) a
3701 cached ldap connection that the parent has called unbind on.
3703 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3704 of $address_data from the recipient verification was clobbered by the
3705 sender verification.
3707 13. The value of address_data from a sender verification is now available in
3708 $sender_address_data in subsequent conditions in the ACL statement.
3710 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3712 15. Added a new option "connect=<time>" to callout options, to set a different
3715 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3716 was its contents. (It was OK if the option was not defined at all.)
3718 17. A "Completed" log line is now written for messages that are removed from
3719 the spool by the -Mrm option.
3721 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3722 information about exactly what failed.
3724 19. Added -dd to debug only the daemon process.
3726 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3727 handles timeouts, both on the server side and network timeouts. Renamed the
3728 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3730 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3731 It is now set to "smtps".
3733 22. $host_address is now set to the target address during the checking of
3734 ignore_target_hosts.
3736 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3737 being passed; this would have caused $sender_host_name to have been used if
3738 matching the list had actually called for a host name (not very likely,
3739 since this list is usually IP addresses). A host name is now passed as
3742 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3743 code that specifies a non-restarting handler (typically sigaction() in
3744 modern systems) in an attempt to fix a rare and obscure crash bug.
3746 25. Narrowed the window for a race in the daemon that could cause it to ignore
3747 SIGCHLD signals. This is not a major problem, because they are used only to
3748 wake it up if nothing else does.
3750 26. A malformed maildirsize file could cause Exim to calculate negative values
3751 for the mailbox size or file count. Odd effects could occur as a result.
3752 The maildirsize information is now recalculated if the size or filecount
3755 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3756 support for a long time. Removed HAVE_SYS_VFS_H.
3758 28. Installed the latest version of exipick from John Jetmore.
3760 29. In an address list, if the pattern was not a regular expression, an empty
3761 subject address (from a bounce message) matched only if the pattern was an
3762 empty string. Non-empty patterns were not even tested. This was the wrong
3763 because it is perfectly reasonable to use an empty address as part of a
3764 database query. An empty address is now tested by patterns that are
3765 lookups. However, all the other forms of pattern expect the subject to
3766 contain a local part and a domain, and therefore, for them, an empty
3767 address still always fails if the pattern is not itself empty.
3769 30. Exim went into a mad DNS loop when attempting to do a callout where the
3770 host was specified on an smtp transport, and looking it up yielded more
3771 than one IP address.
3773 31. Re-factored the code for checking spool and log partition space into a
3774 function that finds that data and another that does the check. The former
3775 is then used to implement four new variables: $spool_space, $log_space,
3776 $spool_inodes, and $log_inodes.
3778 32. The RFC2047 encoding function was originally intended for short strings
3779 such as real names; it was not keeping to the 75-character limit for
3780 encoded words that the RFC imposes. It now respects the limit, and
3781 generates multiple encoded words if necessary. To be on the safe side, I
3782 have increased the buffer size for the ${rfc2047: expansion operator from
3785 33. It is now permitted to omit both strings after an "if" condition; if the
3786 condition is true, the result is "true". As before, when the second string
3787 is omitted, a false condition yields an empty string. This makes it less
3788 cumbersome to write custom ACL and router conditions.
3790 34. Failure to deliver a bounce message always caused it to be frozen, even if
3791 there was an errors_to setting on the router. The errors_to setting is now
3794 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3795 canonical form (fully expanded) before being placed in
3796 $sender_host_address.
3798 36. The table in the code that translates DNS record types into text (T_A to
3799 "A" for instance) was missing entries for NS and CNAME. It is just possible
3800 that this could have caused confusion if both these types were looked up
3801 for the same domain, because the text type is used as part of Exim's
3802 per-process caching. But the chance of anyone hitting this buglet seems
3805 37. The dnsdb lookup has been extended in a number of ways.
3807 (1) There is a new type, "zns", which walks up the domain tree until it
3808 finds some nameserver records. It should be used with care.
3810 (2) There is a new type, "mxh", which is like "mx" except that it returns
3811 just the host names, not the priorities.
3813 (3) It is now possible to give a list of domains (or IP addresses) to be
3814 looked up. The behaviour when one of the lookups defers can be
3815 controlled by a keyword.
3817 (4) It is now possible to specify the separator character for use when
3818 multiple records are returned.
3820 38. The dnslists ACL condition has been extended: it is now possible to supply
3821 a list of IP addresses and/or domains to be looked up in a particular DNS
3824 39. Added log_selector=+queue_time_overall.
3826 40. When running the queue in the test harness, wait just a tad after forking a
3827 delivery process, to get repeatability of debugging output.
3829 41. Include certificate and key file names in error message when GnuTLS fails
3830 to set them up, because the GnuTLS error message doesn't include the name
3831 of the failing file when there is a problem reading it.
3833 42. Allow both -bf and -bF in the same test run.
3835 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3837 44. The "Exiscan patch" is now merged into the mainline Exim source.
3839 45. Sometimes the final signoff response after QUIT could fail to get
3840 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3841 < 0 before doing a fflush(). This bug looks as though it goes back to the
3842 introduction of TLS in release 3.20, but "sometimes" must have been rare
3843 because the tests only now provoked it.
3845 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3846 (this can affect the format of dates).
3848 47. exim_tidydb, when checking for the continued existence of a message for
3849 which it has found a message-specific retry record, was not finding
3850 messages that were in split spool directories. Consequently, it was
3851 deleting retry records that should have stayed in existence.
3853 48. Steve fixed some bugs in eximstats.
3855 49. The SPA authentication driver was not abandoning authentication and moving
3856 on to the next authenticator when an expansion was forced to fail,
3857 contradicting the general specification for all authenticators. Instead it
3858 was generating a temporary error. It now behaves as specified.
3860 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3861 (the order specifies the preference for clients). The order is now AES256,
3862 AES128, 3DES, ARCFOUR128.
3864 51. Small patch to Sieve code - explicitly set From: when generating an
3867 52. Exim crashed if a remote delivery caused a very long error message to be
3868 recorded - for instance if somebody sent an entire SpamAssassin report back
3869 as a large number of 550 error lines. This bug was coincidentally fixed by
3870 increasing the size of one of Exim's internal buffers (big_buffer) that
3871 happened as part of the Exiscan merge. However, to be on the safe side, I
3872 have made the code more robust (and fixed the comments that describe what
3875 53. Now that there can be additional text after "Completed" in log lines (if
3876 the queue_time_overall log selector is set), a one-byte patch to exigrep
3877 was needed to allow it to recognize "Completed" as not the last thing in
3880 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3881 patch that reportedly fixes this has been added. I am not expert enough to
3882 create a test for it. This is what the patch creator wrote:
3884 "I found a little strange behaviour of ldap code when working with
3885 Windows 2003 AD Domain, where users was placed in more than one
3886 Organization Units. When I tried to give exim partial DN, the exit code
3887 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3888 But simultaneously result of request was absolutely normal ldap result,
3889 so I produce this patch..."
3891 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3892 REFERENCE, so I have modified the code to exclude the patch when that macro
3895 55. Some experimental protocols are using DNS PTR records for new purposes. The
3896 keys for these records are domain names, not reversed IP addresses. The
3897 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3898 leaves it alone. Component reversal etc. now happens only for IP addresses.
3901 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3903 57. Double the size of the debug message buffer (to 2048) so that more of very
3904 long debug lines gets shown.
3906 58. The exicyclog utility now does better if the number of log files to keep
3907 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3909 59. Two changes related to the smtp_active_hostname option:
3911 (1) $smtp_active_hostname is now available as a variable.
3912 (2) The default for smtp_banner uses $smtp_active_hostname instead
3913 of $primary_hostname.
3915 60. The host_aton() function is supposed to be passed a string that is known
3916 to be a valid IP address. However, in the case of IPv6 addresses, it was
3917 not checking this. This is a hostage to fortune. Exim now panics and dies
3918 if the condition is not met. A case was found where this could be provoked
3919 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3920 components; fortuitously, this particular loophole had already been fixed
3921 by change 4.50/55 above.
3923 If there are any other similar loopholes, the new check in host_aton()
3924 itself should stop them being exploited. The report I received stated that
3925 data on the command line could provoke the exploit when Exim was running as
3926 exim, but did not say which command line option was involved. All I could
3927 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3928 running as the user.
3931 61. There was a buffer overflow vulnerability in the SPA authentication code
3932 (which came originally from the Samba project). I have added a test to the
3933 spa_base64_to_bits() function which I hope fixes it.
3936 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3937 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3939 63. The daemon start-up calls getloadavg() while still root for those OS that
3940 need the first call to be done as root, but it missed one case: when
3941 deliver_queue_load_max is set with deliver_drop_privilege. This is
3942 necessary for the benefit of the queue runner, because there is no re-exec
3943 when deliver_drop_privilege is set.
3945 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3946 This has been fixed.
3948 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3949 was in use, was not putting the data itself into the right store pool;
3950 consequently, it could be overwritten for a subsequent message in the same
3951 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3954 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3956 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3957 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3958 the function were treating the return as a boolean value, which happened to
3959 work because 0=false and not-0=true, but is not correct code.
3961 68. The host_aton() function was not handling scoped IPv6 addresses (those
3962 with, for example, "%eth0" on the end) correctly.
3964 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3965 negated items (that is, ~something) in unsigned ints. Some compilers
3966 apparently mutter when there is no cast.
3968 70. If an address verification called from an ACL failed, and did not produce a
3969 user-specific message (i.e. there was only a "system" message), nothing was
3970 put in $acl_verify_message. In this situation, it now puts the system
3973 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3974 session; change 4.31/43 added the unwanted input to the log line - except
3975 that it did not do this in the start of session case. It now does.
3977 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3978 This is wrong and can cause the other end to generate a synchronization
3979 error if it is another Exim or anything else that does the synchronization
3980 check. A QUIT command is no longer sent after a timeout.
3982 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3983 during host lookups.
3985 74. The defer_ok option of callout verification was not working if it was used
3986 when verifying addresses in header lines, that is, for this case:
3988 verify = header_sender/callout=defer_ok
3990 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3991 those file descriptors could be used for SMTP connections. If anything
3992 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3993 could be sent to the SMTP client, causing chaos. The daemon now opens
3994 stdin, stdout, and stderr to /dev/null when it puts itself into the
3997 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3998 log by default. The user can override this with suitable Perl magic.
4000 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4001 the log message when discard triggers, was not working for the DATA ACL or
4002 for the non-SMTP ACL.
4004 78. Error message wording change in sieve.c.
4006 79. If smtp_accept_max_per_host was set, the number of connections could be
4007 restricted to fewer than expected, because the daemon was trying to set up
4008 a new connection before checking whether the processes handling previous
4009 connections had finished. The check for completed processes is now done
4010 earlier. On busy systems, this bug wouldn't be noticed because something
4011 else would have woken the daemon, and it would have reaped the completed
4014 80. If a message was submitted locally by a user whose login name contained one
4015 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4016 It caused a spool format error. I have fixed the spool reading code. A
4017 related problem was that the "from" clause in the Received: line became
4018 illegal because of the space(s). It is now covered by ${quote_local_part.
4020 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4023 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4025 83. Patch from Sieve maintainer to fix unterminated string problem in
4026 "vacation" handling.
4028 84. Some minor changes to the Linux configuration files to help with other
4029 OS variants using glibc.
4031 85. One more patch for Sieve to update vacation handling to latest spec.
4034 ----------------------------------------------------
4035 See the note above about the 4.44 and 4.50 releases.
4036 ----------------------------------------------------
4042 1. Change 4.43/35 introduced a bug that caused file counts to be
4043 incorrectly computed when quota_filecount was set in an appendfile
4046 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4047 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4050 3. Give more explanation in the error message when the command for a transport
4051 filter fails to execute.
4053 4. There are several places where Exim runs a non-Exim command in a
4054 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4055 was being done only for the command run by the queryprogram router. It is
4056 now done for all such subprocesses. The other cases are: ${run, transport
4057 filters, and the commands run by the lmtp and pipe transports.
4059 5. Some older OS have a limit of 256 on the maximum number of file
4060 descriptors. Exim was using setrlimit() to set 1000 as a large value
4061 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4062 systems. I've change it so that if it can't get 1000, it tries for 256.
4064 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4065 was an oversight, and furthermore, ever since the addition of extra
4066 controls (e.g. 4.43/32), the checks on when to allow different forms of
4067 "control" were broken. There should now be diagnostics for all cases when a
4068 control that does not make sense is encountered.
4070 7. $recipients is now available in the predata ACL (oversight).
4072 8. Tidy the search cache before the fork to do a delivery from a message
4073 received from the command line. Otherwise the child will trigger a lookup
4074 failure and thereby defer the delivery if it tries to use (for example) a
4075 cached ldap connection that the parent has called unbind on.
4077 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4078 of $address_data from the recipient verification was clobbered by the
4079 sender verification.
4081 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4082 was its contents. (It was OK if the option was not defined at all.)
4084 11. A "Completed" log line is now written for messages that are removed from
4085 the spool by the -Mrm option.
4087 12. $host_address is now set to the target address during the checking of
4088 ignore_target_hosts.
4090 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4091 being passed; this would have caused $sender_host_name to have been used if
4092 matching the list had actually called for a host name (not very likely,
4093 since this list is usually IP addresses). A host name is now passed as
4096 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4097 code that specifies a non-restarting handler (typically sigaction() in
4098 modern systems) in an attempt to fix a rare and obscure crash bug.
4100 15. Narrowed the window for a race in the daemon that could cause it to ignore
4101 SIGCHLD signals. This is not a major problem, because they are used only to
4102 wake it up if nothing else does.
4104 16. A malformed maildirsize file could cause Exim to calculate negative values
4105 for the mailbox size or file count. Odd effects could occur as a result.
4106 The maildirsize information is now recalculated if the size or filecount
4109 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4110 support for a long time. Removed HAVE_SYS_VFS_H.
4112 18. Updated exipick to current release from John Jetmore.
4114 19. Allow an empty sender to be matched against a lookup in an address list.
4115 Previously the only cases considered were a regular expression, or an
4118 20. Exim went into a mad DNS lookup loop when doing a callout where the
4119 host was specified on the transport, if the DNS lookup yielded more than
4122 21. The RFC2047 encoding function was originally intended for short strings
4123 such as real names; it was not keeping to the 75-character limit for
4124 encoded words that the RFC imposes. It now respects the limit, and
4125 generates multiple encoded words if necessary. To be on the safe side, I
4126 have increased the buffer size for the ${rfc2047: expansion operator from
4129 22. Failure to deliver a bounce message always caused it to be frozen, even if
4130 there was an errors_to setting on the router. The errors_to setting is now
4133 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4134 canonical form (fully expanded) before being placed in
4135 $sender_host_address.
4137 24. Updated eximstats to version 1.33
4139 25. Include certificate and key file names in error message when GnuTLS fails
4140 to set them up, because the GnuTLS error message doesn't include the name
4141 of the failing file when there is a problem reading it.
4143 26. Expand error message when OpenSSL has problems setting up cert/key files.
4146 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4147 (this can affect the format of dates).
4149 28. exim_tidydb, when checking for the continued existence of a message for
4150 which it has found a message-specific retry record, was not finding
4151 messages that were in split spool directories. Consequently, it was
4152 deleting retry records that should have stayed in existence.
4154 29. eximstats updated to version 1.35
4155 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4156 1.35 - bugfix such that pie charts by volume are generated correctly
4158 30. The SPA authentication driver was not abandoning authentication and moving
4159 on to the next authenticator when an expansion was forced to fail,
4160 contradicting the general specification for all authenticators. Instead it
4161 was generating a temporary error. It now behaves as specified.
4163 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4164 (the order specifies the preference for clients). The order is now AES256,
4165 AES128, 3DES, ARCFOUR128.
4167 31. Small patch to Sieve code - explicitly set From: when generating an
4170 32. Exim crashed if a remote delivery caused a very long error message to be
4171 recorded - for instance if somebody sent an entire SpamAssassin report back
4172 as a large number of 550 error lines. This bug was coincidentally fixed by
4173 increasing the size of one of Exim's internal buffers (big_buffer) that
4174 happened as part of the Exiscan merge. However, to be on the safe side, I
4175 have made the code more robust (and fixed the comments that describe what
4178 33. Some experimental protocols are using DNS PTR records for new purposes. The
4179 keys for these records are domain names, not reversed IP addresses. The
4180 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4181 leaves it alone. Component reversal etc. now happens only for IP addresses.
4184 34. The host_aton() function is supposed to be passed a string that is known
4185 to be a valid IP address. However, in the case of IPv6 addresses, it was
4186 not checking this. This is a hostage to fortune. Exim now panics and dies
4187 if the condition is not met. A case was found where this could be provoked
4188 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4189 components; fortuitously, this particular loophole had already been fixed
4190 by change 4.50/55 or 4.44/33 above.
4192 If there are any other similar loopholes, the new check in host_aton()
4193 itself should stop them being exploited. The report I received stated that
4194 data on the command line could provoke the exploit when Exim was running as
4195 exim, but did not say which command line option was involved. All I could
4196 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4197 running as the user.
4200 35. There was a buffer overflow vulnerability in the SPA authentication code
4201 (which came originally from the Samba project). I have added a test to the
4202 spa_base64_to_bits() function which I hope fixes it.
4205 36. The daemon start-up calls getloadavg() while still root for those OS that
4206 need the first call to be done as root, but it missed one case: when
4207 deliver_queue_load_max is set with deliver_drop_privilege. This is
4208 necessary for the benefit of the queue runner, because there is no re-exec
4209 when deliver_drop_privilege is set.
4211 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4212 was in use, was not putting the data itself into the right store pool;
4213 consequently, it could be overwritten for a subsequent message in the same
4214 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4217 38. Sometimes the final signoff response after QUIT could fail to get
4218 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4219 < 0 before doing a fflush(). This bug looks as though it goes back to the
4220 introduction of TLS in release 3.20, but "sometimes" must have been rare
4221 because the tests only now provoked it.
4227 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4228 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4229 Now it returns an integer. A number of calls were still expecting a T/F
4230 return. Fortuitously, in all cases, the tests worked in OK situations,
4231 which is the norm. However, things would have gone wrong on any write
4232 failures on the smtp file descriptor. This function is used when sending
4233 messages over SMTP and also when doing verify callouts.
4235 2. When Exim is called to do synchronous delivery of a locally submitted
4236 message (the -odf or -odi options), it no longer closes stderr before doing
4239 3. Implemented the mua_wrapper option.
4241 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4243 5. Implemented the functions header_remove(), header_testname(),
4244 header_add_at_position(), and receive_remove_recipient(), and exported them
4247 6. If an ACL "warn" statement specified the addition of headers, Exim already
4248 inserted X-ACL-Warn: at the start if there was no header name. However, it
4249 was not making this test for the second and subsequent header lines if
4250 there were newlines in the string. This meant that an invalid header could
4251 be inserted if Exim was badly configured.
4253 7. Allow an ACL "warn" statement to add header lines at the start or after all
4254 the Received: headers, as well as at the end.
4256 8. Added the rcpt_4xx retry error code.
4258 9. Added postmaster_mailfrom=xxx to callout verification option.
4260 10. Added mailfrom=xxxx to the callout verification option, for verify=
4263 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4264 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4266 12. Inserted some casts to stop certain compilers warning when using pointer
4267 differences as field lengths or precisions in printf-type calls (mostly
4268 affecting debugging statements).
4270 13. Added optional readline() support for -be (dynamically loaded).
4272 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4273 same clock tick as a message's arrival, so that its received time was the
4274 same as the "first fail" time on the retry record, and that message
4275 remained on the queue past the ultimate address timeout, every queue runner
4276 would try a delivery (because it was past the ultimate address timeout) but
4277 after another failure, the ultimate address timeout, which should have then
4278 bounced the address, did not kick in. This was a "< instead of <=" error;
4279 in most cases the first failure would have been in the next clock tick
4280 after the received time, and all would be well.
4282 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4283 being recognized when the domain list was tested by the match_domain
4284 condition in an expansion string.
4286 16. Added the ${str2b64: operator.
4288 17. Exim was always calling setrlimit() to set a large limit for the number of
4289 processes, without checking whether the existing limit was already
4290 adequate. (It did check for the limit on file descriptors.) Furthermore,
4291 errors from getrlimit() and setrlimit() were being ignored. Now they are
4292 logged to the main and panic logs, but Exim does carry on, to try to do its
4293 job under whatever limits there are.
4295 18. Imported PCRE 5.0.
4297 19. Trivial typo in log message " temporarily refused connection" (the leading
4300 20. If the log selector return_path_on_delivery was set and an address was
4301 redirected to /dev/null, the delivery process crashed because it assumed
4302 that a return path would always be set for a "successful" delivery. In this
4303 case, the whole delivery is bypassed as an optimization, and therefore no
4306 21. Internal re-arrangement: the function for sending a challenge and reading
4307 a response while authentication was assuming a zero-terminated challenge
4308 string. It's now changed to take a pointer and a length, to allow for
4309 binary data in such strings.
4311 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4313 23. Exim was not respecting finduser_retries when seeking the login of the
4314 uid under which it was called; it was always trying 10 times. (The default
4315 setting of finduser_retries is zero.) Also, it was sleeping after the final
4316 failure, which is pointless.
4318 24. Implemented tls_on_connect_ports.
4320 25. Implemented acl_smtp_predata.
4322 26. If the domain in control=submission is set empty, Exim assumes that the
4323 authenticated id is a complete email address when it generates From: or
4324 Sender: header lines.
4326 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4327 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4328 chown and chgrp in /bin and hostname in /usr/bin.
4330 28. Exim was keeping the "process log" file open after each use, just as it
4331 does for the main log. This opens the possibility of it remaining open for
4332 long periods when the USR1 signal hits a daemon. Occasional processlog
4333 errors were reported, that could have been caused by this. Anyway, it seems
4334 much more sensible not to leave this file open at all, so that is what now
4337 29. The long-running daemon process does not normally write to the log once it
4338 has entered its main loop, and it closes the log before doing so. This is
4339 so that log files can straightforwardly be renamed and moved. However,
4340 there are a couple of unusual error situations where the daemon does write
4341 log entries, and I had neglected to close the log afterwards.
4343 30. The text of an SMTP error response that was received during a remote
4344 delivery was being truncated at 512 bytes. This is too short for some of
4345 the long messages that one sometimes sees. I've increased the limit to
4348 31. It is now possible to make retry rules that apply only when a message has a
4349 specific sender, in particular, an empty sender.
4351 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4352 it possible to be selective about when SMTP synchronization is enforced.
4354 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4356 32. Implemented hosts_connection_nolog.
4358 33. Added an ACL for QUIT.
4360 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4363 35. Added mailbox_size and mailbox_filecount to appendfile.
4365 36. Added control = no_multiline_responses to ACLs.
4367 37. There was a bug in the logic of the code that waits for the clock to tick
4368 in the case where the clock went backwards by a substantial amount such
4369 that the microsecond fraction of "now" was more than the microsecond
4370 fraction of "then" (but the whole seconds number was less).
4372 38. Added support for the libradius Radius client library this is found on
4373 FreeBSD (previously only the radiusclient library was supported).
4379 1. When certain lookups returned multiple values in the form name=value, the
4380 quoting of the values was not always being done properly. Specifically:
4381 (a) If the value started with a double quote, but contained no whitespace,
4383 (b) If the value contained whitespace other than a space character (i.e.
4384 tabs or newlines or carriage returns) it was not quoted.
4385 This fix has been applied to the mysql and pgsql lookups by writing a
4386 separate quoting function and calling it from the lookup code. The fix
4387 should probably also be applied to nisplus, ibase and oracle lookups, but
4388 since I cannot test any of those, I have not disturbed their existing code.
4390 2. A hit in the callout cache for a specific address caused a log line with no
4391 reason for rejecting RCPT. Now it says "Previous (cached) callout
4392 verification failure".
4394 3. There was an off-by-one bug in the queryprogram router. An over-long
4395 return line was truncated at 256 instead of 255 characters, thereby
4396 overflowing its buffer with the terminating zero. As well as fixing this, I
4397 have increased the buffer size to 1024 (and made a note to document this).
4399 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4400 when Exim is waiting for an SMTP response from a remote server, Exim
4401 restarts its select() call on the socket, thereby resetting its timeout.
4402 This is not a problem when such interrupts are rare. Somebody set up a cron
4403 job to run exiwhat every 2 minutes, which is less than the normal select()
4404 timeout (5 or 10 minutes). This meant that the select() timeout never
4405 kicked in because it was always reset. I have fixed this by comparing the
4406 time when an interrupt arrives with the time at the start of the first call
4407 to select(). If more time than the timeout has elapsed, the interrupt is
4408 treated as a timeout.
4410 5. Some internal re-factoring in preparation for the addition of Sieve
4411 extensions (by MH). In particular, the "personal" test is moved to a
4412 separate function, and given an option for scanning Cc: and Bcc: (which is
4413 not set for Exim filters).
4415 6. When Exim created an email address using the login of the caller as the
4416 local part (e.g. when creating a From: or Sender: header line), it was not
4417 quoting the local part when it contained special characters such as @.
4419 7. Installed new OpenBSD configuration files.
4421 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4422 try to make them clearer.
4424 9. Callout options, other than the timeout value, were being ignored when
4425 verifying sender addresses in header lines. For example, when using
4427 verify = header_sender/callout=no_cache
4429 the cache was (incorrectly) being used.
4431 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4432 only the Cygwin environment.
4434 11. When return_path_on_delivery was set as a log selector, if different remote
4435 addresses in the same message used different return paths and parallel
4436 remote delivery occurred, the wrong values would sometimes be logged.
4437 (Whenever a remote delivery process finished, the return path value from
4438 the most recently started remote delivery process was logged.)
4440 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4441 lines when AUTH and/or TLS are in use. This is the "received protocol"
4442 field. Exim used to use "asmtp" for authenticated SMTP, without any
4443 indication (in the protocol name) for TLS use. Now it follows the RFC and
4444 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4445 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4446 names appear in log lines as well as in Received: header lines.
4448 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4449 extensions, and comparison tests, and to fix some bugs.
4451 14. Changes to the "personal" filter test:
4453 (1) The test was buggy in that it was just doing the equivalent of
4454 "contains" tests on header lines. For example, if a user's address was
4455 anne@some.where, the "personal" test would incorrectly be true for
4457 To: susanne@some.where
4459 This test is now done by extracting each address from the header in turn,
4460 and checking the entire address. Other tests that are part of "personal"
4461 are now done using regular expressions (for example, to check local parts
4462 of addresses in From: header lines).
4464 (2) The list of non-personal local parts in From: addresses has been
4465 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4466 taken from the Sieve specification recommendations.
4468 (3) If the message contains any header line starting with "List-" it is
4469 treated as non-personal.
4471 (4) The test for "circular" in the Subject: header line has been removed
4472 because it now seems ill-conceived.
4474 15. Minor typos in src/EDITME comments corrected.
4476 16. Installed latest exipick from John Jetmore.
4478 17. If headers_add on a router specified a text string that was too long for
4479 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4480 of string_sprintf() is now avoided.
4482 18. $message_body_size was not set (it was always zero) when running the DATA
4483 ACL and the local_scan() function.
4485 19. For the "mail" command in an Exim filter, no default was being set for
4486 the once_repeat time, causing a random time value to be used if "once" was
4487 specified. (If the value happened to be <= 0, no repeat happened.) The
4488 default is now 0s, meaning "never repeat". The "vacation" command was OK
4489 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4490 (I found it when inspecting the code).
4492 20. There is now an overall timeout for performing a callout verification. It
4493 defaults to 4 times the callout timeout, which applies to individual SMTP
4494 commands during the callout. The overall timeout applies when there is more
4495 than one host that can be tried. The timeout is checked before trying the
4496 next host. This prevents very long delays if there are a large number of
4497 hosts and all are timing out (e.g. when the network connections are timing
4498 out). The value of the overall timeout can be changed by specifying an
4499 additional sub-option for "callout", called "maxwait". For example:
4501 verify = sender/callout=5s,maxwait=20s
4503 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4504 to the end before writing, but this should make it even safer).
4506 22. Exim was forgetting that it had advertised PIPELINING for the second and
4507 subsequent messages on an SMTP connection. It was also not resetting its
4508 memory on STARTTLS and an internal HELO.
4510 23. When Exim logs an SMTP synchronization error within a session, it now
4511 records whether PIPELINING has been advertised or not.
4513 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4514 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4515 rather than long int.
4517 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4519 26. Added the never_mail option to autoreply.
4525 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4526 crash if the getsockname() call failed; this can happen if a connection is
4527 closed very soon after it is established. The problem was simply in the
4528 order in which certain operations were done, causing Exim to try to write
4529 to the SMTP stream before it had set up the file descriptor. The bug has
4530 been fixed by making things happen in the correct order.
4536 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4537 before the connection was closed, thus losing the rejection response.
4539 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4540 some early Solaris releases, but causes trouble in current releases where
4541 socklen_t is defined.
4543 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4546 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4549 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4550 that releasing the top of it at the end releases what was used for sub-
4551 expansions (unless the block got too big). However, discard this block if
4552 the first thing is a variable or header, so that we can use its block when
4553 it is dynamic (useful for very large $message_headers, for example).
4555 6. Lookups now cache *every* query, not just the most recent. A new, separate
4556 store pool is used for this. It can be recovered when all lookup caches are
4557 flushed. Lookups now release memory at the end of their result strings.
4558 This has involved some general refactoring of the lookup sources.
4560 7. Some code has been added to the store_xxx() functions to reduce the amount
4561 of flapping under certain conditions.
4563 8. log_incoming_interface used to affect only the <= reception log lines. Now
4564 it causes the local interface and port to be added to several more SMTP log
4565 lines, for example "SMTP connection from", and rejection lines.
4567 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4569 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4571 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4572 could be overwritten at the end of the current message (or the start of a
4573 new message if it was set in a HELO ACL). The value is now preserved for
4574 the duration of the SMTP connection.
4576 12. If a transport had a headers_rewrite setting, and a matching header line
4577 contained an unqualified address, that address was qualified, even if it
4578 did not match any rewriting rules. The underlying bug was that the values
4579 of the flags that permit the existence of unqualified sender and recipient
4580 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4581 non-local messages, and by -bnq for local messages) were not being
4582 preserved with the message after it was received.
4584 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4585 "next input=" as part of the text comprising the host identity instead of
4586 the correct text. The code was using the same buffer for two different
4587 strings. However, depending on which order the printing function evaluated
4588 its arguments, the bug did not always show up. Under Linux, for example, my
4589 test suite worked just fine.
4591 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4592 doesn't work with some older versions of Perl. It has been changed to "my",
4593 which in any case is probably the better facility to use.
4595 15. A really picky compiler found some instances of statements for creating
4596 error messages that either had too many or two few arguments for the format
4599 16. The size of the buffer for calls to the DNS resolver has been increased
4600 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4601 for addresses that have a lot of PTR records. This alleviates a problem; it
4602 does not fully solve it.
4604 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4605 buffer now truncates the list and logs the incident, which is the same
4606 action as happens when Exim is looking up a host name and its aliases.
4607 Previously in this situation something unpredictable would happen;
4608 sometimes it was "internal error: store_reset failed".
4610 18. If a server dropped the connection unexpectedly when an Exim client was
4611 using GnuTLS and trying to read a response, the client delivery process
4612 crashed while trying to generate an error log message.
4614 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4615 string, for example:
4617 warn message = H1: something\nH2: something
4619 the text was added as a single header line from Exim's point of view
4620 though it ended up OK in the delivered message. However, searching for the
4621 second and subsequent header lines using $h_h2: did not work. This has been
4622 fixed. Similarly, if a system filter added multiple headers in this way,
4623 the routers could not see them.
4625 20. Expanded the error message when iplsearch is called with an invalid key to
4626 suggest using net-iplsearch in a host list.
4628 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4629 ACLs are no longer actually imposed (and a message to that effect is
4632 22. If a "gecos" field in a passwd entry contained escaped characters, in
4633 particular, if it contained a \" sequence, Exim got it wrong when building
4634 a From: or a Sender: header from that name. A second bug also caused
4635 incorrect handling when an unquoted " was present following a character
4636 that needed quoting.
4638 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4639 was not being matched caselessly.
4641 24. Arranged for all hyphens in the exim.8 source to be escaped with
4644 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4645 properly. Recipient callout cache records were still being keyed to include
4646 the sender, even when use_sender was set false. This led to far more
4647 callouts that were necessary. The sender is no longer included in the key
4648 when use_sender is false.
4650 26. Added "control = submission" modifier to ACLs.
4652 27. Added the ${base62d: operator to decode base 62 numbers.
4654 28. dnsdb lookups can now access SRV records.
4656 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4657 the configuration file.
4659 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4660 (-v) mode. This makes the output for a verbose queue run more intelligible.
4662 31. Added a use_postmaster feature to recipient callouts.
4664 32. Added the $body_zerocount variable, containing the number of binary zero
4665 bytes in the message body.
4667 33. The time of last modification of the "new" subdirectory is now used as the
4668 "mailbox time last read" when there is a quota error for a maildir
4671 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4673 35. Added +ignore_unknown as a special item in host lists.
4675 36. Code for decoding IPv6 addresses in host lists is now included, even if
4676 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4677 address was recognized as an IP address, but was then not correctly decoded
4678 into binary, causing unexpected and incorrect effects when compared with
4685 1. Very minor rewording of debugging text in manualroute to say "list of
4686 hosts" instead of "hostlist".
4688 2. If verify=header_syntax was set, and a header line with an unqualified
4689 address (no domain) and a large number of spaces between the end of the
4690 name and the colon was received, the reception process suffered a buffer
4691 overflow, and (when I tested it) crashed. This was caused by some obsolete
4692 code that should have been removed. The fix is to remove it!
4694 3. When running in the test harness, delay a bit after writing a bounce
4695 message to get a bit more predictability in the log output.
4697 4. Added a call to search_tidyup() just before forking a reception process. In
4698 theory, someone could use a lookup in the expansion of smtp_accept_max_
4699 per_host which, without the tidyup, could leave open a database connection.
4701 5. Added the variables $recipient_data and $sender_data which get set from a
4702 lookup success in an ACL "recipients" or "senders" condition, or a router
4703 "senders" option, similar to $domain_data and $local_part_data.
4705 6. Moved the writing of debug_print from before to after the "senders" test
4708 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4709 problems for message scanning, either using a data ACL, or using
4710 local_scan() because the Received: header was not generated till after they
4711 were called (in order to set the time as the time of reception completion).
4712 I have revised the way this works. The header is now generated after the
4713 body is received, but before the ACL or local_scan() are called. After they
4714 are run, the timestamp in the header is updated.
4720 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4721 before starting a queue runner without re-exec. This happened only when
4722 deliver_drop_privilege was set or when the Exim user was set to root. The
4723 effect of the bug was that timeouts during subsequent deliveries caused
4724 crashes instead of being properly handled. The handler is now left at its
4725 default (and expected) setting.
4727 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4728 message, again when deliver_drop_privilege is set or Exim is run as root.
4729 The bug described in (1) was not present in this case, but the tidying up
4730 of the other signals was missing. I have made the two cases consistent.
4732 3. The ignore_target_hosts setting on a manualroute router was being ignored
4733 for hosts that were looked up using the /MX notation.
4735 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4738 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4739 operated on the sender address. After changing the $sender_address to <>
4740 for the sender address verify, Exim was re-instated it as the original
4741 (before rewriting) address, but remembering that it had rewritten it, so it
4742 wasn't rewriting it again. This bug also had the effect of breaking the
4743 sender address verification caching when the sender address was rewritten.
4745 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4746 This has been changed so that if the ip literal address matches
4747 ignore_target_hosts, the router declines.
4749 7. Added expansion conditions match_domain, match_address, and match_local_
4750 part (NOT match_host).
4752 8. The placeholder for the Received: header didn't have a length field set.
4754 9. Added code to Exim itself and to exim_lock to test for a specific race
4755 condition that could lead to file corruption when using MBX delivery. The
4756 issue is with the lockfile that is created in /tmp. If this file is removed
4757 after a process has opened it but before that process has acquired a lock,
4758 there is the potential for a second process to recreate the file and also
4759 acquire a lock. This could lead to two Exim processes writing to the file
4760 at the same time. The added code performs the same test as UW imapd; it
4761 checks after acquiring the lock that its file descriptor still refers to
4762 the same named file.
4764 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4765 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4768 11. Added the smtp_active_hostname option. If used, this will typically be made
4769 to depend on the incoming interface address. Because $interface_address is
4770 not set up until the daemon has forked a reception process, error responses
4771 that can happen earlier (such as "too many connections") no longer contain
4774 12. If an expansion in a condition on a "warn" statement fails because a lookup
4775 defers, the "warn" statement is abandoned, and the next ACL statement is
4776 processed. Previously this caused the whole ACL to be aborted.
4778 13. Added the iplsearch lookup type.
4780 14. Added ident_timeout as a log selector.
4782 15. Added tls_certificate_verified as a log selector.
4784 16. Added a global option tls_require_ciphers (compare the smtp transport
4785 option of the same name). This controls incoming TLS connections.
4787 17. I finally figured out how to make tls_require_ciphers do a similar thing
4788 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4789 before starting the TLS session.
4791 18. Tabs are now shown as \t in -bP output.
4793 19. If the log selector return_path_on_delivery was set, Exim crashed when
4794 bouncing a message because it had too many Received: header lines.
4796 20. If two routers both had headers_remove settings, and the first one included
4797 a superfluous trailing colon, the final name in the first list and the
4798 first name in the second list were incorrectly joined into one item (with a
4799 colon in the middle).
4805 1. Added -C and -D options to the exinext utility, mainly to make it easier
4806 to include in the automated testing, but these could be helpful when
4807 multiple configurations are in use.
4809 2. The exinext utility was not formatting the output nicely when there was
4810 an alternate port involved in the retry record key, nor when there was a
4811 message id as well (for retries that were specific to a specific message
4812 and a specific host). It was also confused by IPv6 addresses, because of
4813 the additional colons they contain. I have fixed the IPv4 problem, and
4814 patched it up to do a reasonable job for IPv6.
4816 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4817 delivery, the log line now contains "pipelined" if PIPELINING was used.
4819 4. An SMTP transport process used to panic and die if the bind() call to set
4820 an explicit outgoing interface failed. This has been changed; it is now
4821 treated in the same way as a connect() failure.
4823 5. A reference to $sender_host_name in the part of a conditional expansion
4824 that was being skipped was still causing a DNS lookup. This no longer
4827 6. The def: expansion condition was not recognizing references to header lines
4828 that used bh_ and bheader_.
4830 7. Added the _cache feature to named lists.
4832 8. The code for checking quota_filecount in the appendfile transport was
4833 allowing one more file than it should have been.
4835 9. For compatibility with Sendmail, the command line option
4843 and sets the incoming protocol and host name (for trusted callers). The
4844 host name and its colon can be omitted when only the protocol is to be set.
4845 Note the Exim already has two private options, -pd and -ps, that refer to
4846 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4847 "s", but I don't think that's a major issue.
4849 10. A number of refactoring changes to the code, none of which should affect
4852 (a) The number of logging options was getting close to filling up the
4853 32-bit word that was used as a bit map. I have split them into two classes:
4854 those that are passed in the argument to log_write(), and those that are
4855 only ever tested independently outside of that function. These are now in
4856 separate 32-bit words, so there is plenty of room for expansion again.
4857 There is no change in the user interface or the logging behaviour.
4859 (b) When building, for example, log lines, the code previously used a
4860 macro that called string_cat() twice, in order to add two strings. This is
4861 not really sufficiently general. Furthermore, there was one instance where
4862 it was actually wrong because one of the argument was used twice, and in
4863 one call a function was used. (As it happened, calling the function twice
4864 did not affect the overall behaviour.) The macro has been replaced by a
4865 function that can join an arbitrary number of extra strings onto a growing
4868 (c) The code for expansion conditions now uses a table and a binary chop
4869 instead of a serial search (which was left over from when there were very
4870 few conditions). Also, it now recognizes conditions like "pam" even when
4871 the relevant support is not compiled in: a suitably worded error message is
4872 given if an attempt is made to use such a condition.
4874 11. Added ${time_interval:xxxxx}.
4876 12. A bug was causing one of the ddress fields not to be passed back correctly
4877 from remote delivery subprocesses. The field in question was not being
4878 subsequently used, so this caused to problems in practice.
4880 13. Added new log selectors queue_time and deliver_time.
4882 14. Might have fixed a bug in maildirsizefile handling that threw up
4883 "unexpected character" debug warnings, and recalculated the data
4884 unnecessarily. In any case, I expanded the warning message to give more
4887 15. Added the message "Restricted characters in address" to the statements in
4888 the default ACL that block characters like @ and % in local parts.
4890 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4891 Three changes have been made:
4893 (a) There was a serious bug; a negative response to MAIL caused the whole
4894 recipient domain to be cached as invalid, thereby blocking all messages
4895 to all local parts at the same domain, from all senders. This bug has
4896 been fixed. The domain is no longer cached after a negative response to
4897 MAIL if the sender used is not empty.
4899 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4902 (c) A new callout option, "use_sender" has been added for people who want
4903 the modified behaviour.
4909 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4912 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4913 indeed breaks things for older releases.
4915 3. Added additional logging to the case where there is a problem reading data
4916 from a filter that is running in a subprocess using a pipe, in order to
4917 try to track down a specific problem.
4919 4. Testing facility fudge: when running in the test harness and attempting
4920 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4921 getting "No route to host". Convert this to a timeout.
4923 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4926 6. Some OS don't have socklen_t but use size_t instead. This affects the
4927 fifth argument of getsockopt() amongst other things. This is now
4928 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4929 can be set for individual OS. I have set it for SunOS5, OSF1, and
4930 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4931 some earlier ones do not.
4933 7. Change 4.30/15 was not doing the test caselessly.
4935 8. The standard form for an IPv6 address literal was being rejected by address
4936 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4937 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4938 this, as well as the form without the "IPv6" on the front (but only when
4939 address literals are enabled, of course).
4941 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4943 10. Exim crashed if a message with an empty sender address specified by -f
4944 encountered a router with an errors_to setting. This could be provoked only
4945 by a command such as
4949 where an empty string was supplied; "<>" did not hit this bug.
4951 11. Installed PCRE release 4.5.
4953 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4954 remained set. It is now erased.
4956 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4957 times from message ids (which are base 36 rather than the normal 62).
4959 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4960 were being counted as actual protocol errors, and logged if the log
4961 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4962 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4963 and DATA following a set of rejected RCPTs do not count as protocol errors.
4964 In other words, Exim assumes they were pipelined, though this may not
4965 actually be the case. Of course, in all cases the client gets an
4966 appropriate error code.
4968 15. If a lookup fails in an ACL condition, a message about the failure may
4969 be available; it is used if testing the ACL cannot continue, because most
4970 such messages specify what the cause of the deferral is. However, some
4971 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4972 that caused an old message to be retained and used if a later statement
4973 caused a defer, replacing the real cause of the deferral.
4975 16. If an IP address had so many PTR records that the DNS lookup buffer
4976 was not large enough to hold them, Exim could crash while trying to process
4977 the truncated data. It now detects and logs this case.
4979 17. Further to 4.21/58, another change has been made: if (and only if) the
4980 first line of a message (the first header line) ends with CRLF, a bare LF
4981 in a subsequent header line has a space inserted after it, so as not to
4982 terminate the header.
4984 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4985 unnecessarily, used atoi() instead of strtol(), and didn't check the
4986 termination when getting file sizes from file names by regex.
4988 19. Completely re-implemented the support for maildirsize files, in the light
4989 of a number of problems with the previous contributed implementation
4990 (4.30/29). In particular:
4992 . If the quota is zero, the maildirsize file is maintained, but no quota is
4995 . If the maildir directory does not exist, it is created before any attempt
4996 to write a maildirsize file.
4998 . The quota value in the file is just a cache; if the quota is changed in
4999 the transport, the new value overrides.
5001 . A regular expression is available for excluding directories from the
5004 20. The autoreply transport checks the characters in options that define the
5005 message's headers; it allows continued headers, but it was checking with
5006 isspace() after an embedded newline instead of explicitly looking for a
5009 21. If all the "regular" hosts to which an address was routed had passed their
5010 expiry times, and had not reached their retry times, the address was
5011 bounced, even if fallback hosts were defined. Now Exim should go on to try
5014 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5015 equivalent code in the SMTP transport. Some hosts send humungous responses
5016 to HELO/EHLO, more than 1024 it seems.
5018 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5019 gives compiler warnings in some environments. I've now done it "properly",
5022 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5023 (because of problems with the built-in one) was declared to return uschar *
5024 instead of char *, causing compiler failure.
5026 25. Fixed a file descriptor leak when processing alias/forward files.
5028 26. Fixed a minor format string issue in dbfn.c.
5030 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5032 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5033 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5036 29. When a sender address is verified, it is cached, to save repeating the test
5037 when there is more than one recipient in a message. However, when the
5038 verification involves a callout, it is possible for different callout
5039 options to be set for different recipients. It is too complicated to keep
5040 track of this in the cache, so now Exim always runs a verification when a
5041 callout is required, relying on the callout cache for the optimization.
5042 The overhead is duplication of the address routing, but this should not be
5045 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5046 to be verified with callout=postmaster, and the main callout worked but the
5047 postmaster check failed, the verification correctly failed. However, if a
5048 subsequent RCPT command asked for sender verification *without* the
5049 postmaster check, incorrect caching caused this verification also to fail,
5052 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5053 it was not caching the DNS options (qualify_single, search_parents) that
5054 were used when the lookup failed. A subsequent lookup with different
5055 options therefore always gave the same answer, though there were cases
5056 where it should not have. (Example: a "domains = !$mx_any" option on a
5057 dnslookup router: the "domains" option is always processed without any
5058 widening, but the router might have qualify_single set.) Now Exim uses the
5059 cached value only when the same options are set.
5061 32. Added John Jetmore's "exipick" utility to the distribution.
5063 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5064 than a timeout (e.g. a certificate is required, and is not provided), an
5065 Exim server now closes the connection immediately. Previously it waited for
5066 the client to close - but if the client is SSL, it seems that they each
5067 wait for each other, leading to a delay before one of them times out.
5069 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5070 maintained 0.8.x compatibility because I don't think many are using it, and
5071 it is clearly obsolete.
5073 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5076 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5077 client certificate was expired. A simple patch fixes this, though I don't
5078 understand the full logic of why the verify callback is called multiple
5081 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5082 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5083 which causes problems with some clients (such as the Certicom SSL Plus
5084 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5085 disables the coutermeasure allowing Eudora to connect."
5087 38. Exim was not checking that a write() to a log file succeeded. This could
5088 lead to Bad Things if a log got too big, in particular if it hit a file
5089 size limit. Exim now panics and dies if it cannot write to a log file, just
5090 as it does if it cannot open a log file.
5092 39. Modified OS/Makefile-Linux so that it now contains
5094 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5096 The two -D definitions ensure that Exim is compiled with large file
5097 support, which makes it possible to handle log files that are bigger than
5100 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5101 instance) a domain was checked against a named list that involved a lookup,
5102 causing $domain_data to be set, then another domain was checked against the
5103 same list, then the first domain was re-checked, the value of $domain_data
5104 after the final check could be wrong. In particular, if the second check
5105 failed, it could be set empty. This bug probably also applied to
5108 41. The strip_trailing_dot option was not being applied to the address given
5109 with the -f command-line option.
5111 42. The code for reading a message's header from the spool was incrementing
5112 $received_count, but never initializing it. This meant that the value was
5113 incorrect (doubled) while delivering a message in the same process in which
5114 it was received. In the most common configuration of Exim, this never
5115 happens - a fresh exec is done - but it can happen when
5116 deliver_drop_privilege is set.
5118 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5119 it now includes up to 150 characters of the unexpected data in the log
5122 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5123 and building data strings. The size of both of these buffers was 10 000
5124 bytes - far larger than anybody would *ever* want, thought I. Needless to
5125 say, somebody hit the limit. I have increased the maximum line length to
5126 20 000 and the maximum data length of concatenated lines to 100 000. I have
5127 also fixed two bugs, because there was no checking on these buffers. Tsk,
5128 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5129 buffer is too small.
5131 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5132 lsearch lookups. Now it does.
5134 46. When parsing a route_list item in a manualroute router, a fixed-length
5135 buffer was used for the list of hosts. I made this 1024 bytes long,
5136 thinking that nobody would ever have a list of hosts that long. Wrong.
5137 Somebody had a whole pile of complicated expansion conditions, and the
5138 string was silently truncated, leading to an expansion error. It turns out
5139 that it is easier to change to an unlimited length (owing to other changes
5140 that have happened since this code was originally written) than to build
5141 structure for giving a limitation error. The length of the item that
5142 expands into the list of hosts is now unlimited.
5144 47. The lsearch lookup could not handle data where the length of text line was
5145 more than 4095 characters. Such lines were truncated, leading to shortened
5146 data being returned. It should now handle lines of any length.
5148 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5149 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5152 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5153 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5154 no longer happens in comment lines. A list of replacements is now placed
5155 at the head of all of the source files, except those whose only change is
5156 to replace PERL_COMMAND in the very first #! line.
5158 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5159 messages on the queue, with a bottom-up merge sort, using code contributed
5160 by Michael Haardt. This should make operations like -bp somewhat faster on
5161 large queues. It won't affect queue runners, except when queue_run_in_order
5164 51. Installed eximstats 1.31 in the distribution.
5166 52. Added support for SRV lookups to the dnslookup router.
5168 53. If an ACL referred to $message_body or $message_body_end, the value was not
5169 reset for any messages that followed in the same SMTP session.
5171 54. The store-handling optimization for building very long strings was not
5172 differentiating between the different store pools. I don't think this
5173 actually made any difference in practice, but I've tidied it.
5175 55. While running the routers to verify a sender address, $sender_address
5176 was still set to the sender address. This is wrong, because when routing to
5177 send a bounce to the sender, it would be empty. Therefore, I have changed
5178 it so that, while verifying a sender address, $sender_address is set to <>.
5179 (There is no change to what happens when verifying a recipient address.)
5181 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5182 target A or AAAA records (if not already returned) without resetting the
5183 qualify_single or search_parents options of the DNS resolver. These are
5184 inappropriate in this case because the targets of MX and SRV records must
5185 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5186 target that, when qualified, matched something in the local domain. These
5187 two options are now turned off when doing these lookups.
5189 57. It seems that at least some releases of Reiserfs (which does not have the
5190 concept of a fixed number of inodes) returns zero and not -1 for the
5191 number of available inodes. This interacted badly with check_spool_inodes,
5192 which assumed that -1 was the "no such thing" setting. What I have done is
5193 to check that the total number of inodes is greater than zero before doing
5194 the test of how many are available.
5196 58. When a "warn" ACL statement has a log_message modifier, the message is
5197 remembered, and not repeated. This is to avoid a lot of repetition when a
5198 message has many recipients that cause the same warning to be written.
5199 Howewer, Exim was preserving the list of already written lines for an
5200 entire SMTP session, which doesn't seem right. The memory is now reset if a
5201 new message is started.
5203 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5204 debugging output unless log_rewrite was also set.
5206 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5207 of (int)(handle) when we know that handle contains (void *)(-1).
5209 61. The Exim daemon panic-logs an error return when it closes the incoming
5210 connection. However "connection reset by peer" seems to be common, and
5211 isn't really an error worthy of noting specially, so that particular error
5214 62. When Exim is trying to find all the local interfaces, it used to panic and
5215 die if the ioctl to get the interface flags failed. However, it seems that
5216 on at least one OS (Solaris 9) it is possible to have an interface that is
5217 included in the list of interfaces, but for which you get a failure error
5218 for this call. This happens when the interface is not "plumbed" into a
5219 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5220 failure of the "get flags" call assumes that the interface is down.
5222 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5223 makes life easier for people who are doing arithmetic on fields extracted
5224 from dates, where you often get leading zeros that should not be
5225 interpreted as octal.
5227 64. Added qualify_domain to the redirect router, to override the global
5230 65. If a pathologically long header line contained very many addresses (the
5231 report of this problem mentioned 10 000) and each of them was rewritten,
5232 Exim could use up a very large amount of memory. (It kept on making new
5233 copies of the header line as it rewrote, and never released the old ones.)
5234 At the expense of a bit more processing, the header rewriting function has
5235 been changed so that it no longer eats memory in this way.
5237 66. The generation of the Received: header has been moved from the time that a
5238 message starts to be received, to the time that it finishes. The timestamp
5239 in the Received: header should now be very close to that of the <= log
5240 line. There are two side-effects of this change:
5242 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5243 logged header lines no longer include the local Received: line, because
5244 it has not yet been created. The same applies to a copy of the message
5245 that is returned to a non-SMTP sender when a message is rejected.
5247 (b) When a filter file is tested using -bf, no additional Received: header
5248 is added to the test message. After some thought, I decided that this
5251 This change does not affect the value of $received_for. It is still set
5252 after address rewriting, but before local_scan() is called.
5254 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5256 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5257 gave an unhelpful panic error message, and a defer error. I have managed to
5258 change this behaviour so that it now rejects any supplied certificate,
5259 which seems right, as the list of acceptable certificates is empty.
5261 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5262 gave an unhelpful defer error. I have not managed to make this reject any
5263 supplied certificates, but the error message it gives is "no certificate
5264 supplied", which is not helpful.
5266 70. exigrep's output now also includes lines that are not associated with any
5267 message, but which match the given pattern. Implemented by a patch from
5268 Martin Sluka, which also tidied up the Perl a bit.
5270 71. Recipient callout verification, like sender verification, was using <> in
5271 the MAIL FROM command. This isn't really the right thing, since the actual
5272 sender may affect whether the remote host accepts the recipient or not. I
5273 have changed it to use the actual sender in the callout; this means that
5274 the cache record is now keyed on a recipient/sender pair, not just the
5275 recipient address. There doesn't seem to be a real danger of callout loops,
5276 since a callout by the remote host to check the sender would use <>.
5277 [SEE ABOVE: changed after hitting problems.]
5279 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5280 temporary errors. However, in the case of such a code being given after
5281 the end of a data transmission (i.e. after ".") Exim was failing to write
5282 a retry record for the message. (Yes, there was some broken host that was
5283 actually sending 8xx at this point.)
5285 73. An unknown lookup type in a host list could cause Exim to panic-die when
5286 the list was checked. (An example that provoked this was putting <; in the
5287 middle of a list instead of at the start.) If this happened during a DATA
5288 ACL check, a -D file could be left lying around. This kind of configuration
5289 error no longer causes Exim to die; instead it causes a defer errror. The
5290 incident is still logged to the main and panic logs.
5292 74. Buglet left over from Exim 3 conversion. The message "too many messages
5293 in one connection" was written to the rejectlog but not the mainlog, except
5294 when address rewriting (yes!) was being logged.
5296 75. Added write_rejectlog option.
5298 76. When a system filter was run not as root (that is, when system_filter_user
5299 was set), the values of the $n variables were not being returned to the
5300 main process; thus, they were not subsequently available in the $sn
5303 77. Added +return_path_on_delivery log selector.
5305 78. A connection timeout was being treated differently from recipients deferred
5306 when testing hosts_max_try with a message that was older than the host's
5307 retry timeout. (The host should not be counted, thus allowing all hosts to
5308 be tried at least once before bouncing.) This may have been the cause of an
5309 occasionally reported bug whereby a message would remain on the queue
5310 longer than the retry timeout, but would be bounced if a delivery was
5311 forced. I say "may" because I never totally pinned down the problem;
5312 setting up timeout/retry tests is difficult. See also the next item.
5314 79. The ultimate address timeout was not being applied to errors that involved
5315 a combination of host plus message (for example, a timeout on a MAIL
5316 command). When an address resolved to a number of possible hosts, and they
5317 were not all tried for each delivery (e.g. because of hosts_max_try), a
5318 message could remain on the queue longer than the retry timeout.
5320 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5323 81. Fixed an obscure SMTP outgoing bug which required at least the following
5324 conditions: (a) there was another message waiting for the same server;
5325 (b) the server returned 5xx to all RCPT commands in the first message so
5326 that the message was not completed; (c) the server dropped the connection
5327 or gave a negative response to the RSET that Exim sends to abort the
5328 transaction. The observed case was a dropped connection after DATA that had
5329 been sent in pipelining mode. That is, the server had advertised PIPELINING
5330 but was not implementing it correctly. The effect of the bug was incorrect
5331 behaviour, such as trying another host, and this could lead to a crash.
5337 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5338 and daemon.c were passed as pointers to ints; they should have been
5339 pointers to socklen_t variables (which are typically unsigned ints).
5341 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5344 3. Fixed a really odd bug that affected only the testing scheme; patching a
5345 certain fixed string in the binary changed the value of another string that
5346 happened to be identical to the end of the original first string.
5348 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5349 name", it returns that address as the IP address. On some operating
5350 systems (e.g. Solaris), it also passes back the IP address string as the
5351 "host name". However, on others (e.g. Linux), it passes back an empty
5352 string. Exim wasn't checking for this, and was changing the host name to an
5353 empty string, assuming it had been canonicized.
5355 5. Although rare, it is permitted to have more than one PTR record for a given
5356 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5357 all the names associated with an address, because they do in Solaris.
5358 However, it seems that they do not in Linux for data that comes from the
5359 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5360 I found this out when I moved to a new Linux workstation and tried to run
5361 the Exim test suite.
5363 To get round this problem I have changed the code so that it now does its
5364 own call to the DNS to look up PTR records when searching for a host name.
5365 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5366 addresses that are only in /etc/hosts are still found.
5368 This behaviour is, however, controlled by an option called host_lookup_
5369 order, which defaults to "bydns:byaddr". If people want to use the other
5370 order, or indeed, just use one or the other means of lookup, they can
5371 specify it in this variable.
5373 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5374 some operating systems, this comes back from gethostbyaddr() as an empty
5375 string, and this is what Exim used to test for. However, it seems that in
5376 other systems, "." is yielded. Exim now tests for this case too.
5378 7. The values of check_spool_space and check_log_space are now held internally
5379 as a number of kilobytes instead of an absolute number of bytes. If a
5380 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5381 kilobyte. This means that much larger values can be stored.
5383 8. Exim monitor: an attempt to get the action menu when not actually pointing
5384 at a message produces an empty menu entitled "No message selected". This
5385 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5386 no entries in it ("Shell widget menu has zero width and/or height"). So I
5387 have added a single, blank menu entry in this case.
5389 9. Added ${quote_local_part.
5391 10. MIME decoding is now applied to the contents of Subject: header lines when
5394 11. Now that a reference to $sender_host_address automatically causes a reverse
5395 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5396 host lookup before query-style lookups in lists that might use this
5397 variable. This has therefore been abolished, and the "net-" prefix is no
5398 longer necessary for query-style lookups.
5400 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5401 have been a typo for LFLAGS, so it has been changed.
5403 13. The install script calls Exim with "-C /dev/null" in order to find the
5404 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5405 to be output. Howeve, since Exim outputs its version number before the
5406 error, it didn't break the script. It just looked ugly. I fixed this by
5407 always allowing "-C /dev/null" if the caller is root.
5409 14. Ignore overlarge ACL variable number when reading spool file - insurance
5410 against a later release with more variables having written the file.
5412 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5413 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5414 this, as well as the form without the "IPv6" on the front.
5416 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5417 OS/Makefile-Darwin file.
5419 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5420 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5422 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5423 variable. However, some OS use ints for inodes. I've added cast to long int
5424 to get rid of the compiler warning.
5426 19. I had forgotten to lock out "/../" in configuration file names when
5427 ALT_CONFIG_PREFIX was set.
5429 20. Routers used for verification do not need to specify transports. However,
5430 if such a router generated a host list, and callout was configured, Exim
5431 crashed, because it could not find a port number from the (non-existent)
5432 transport. It now assumes port 25 in this circumstance.
5434 21. Added the -t option to exigrep.
5436 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5437 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5438 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5439 them is set, LOOKUP_LSEARCH is forced.
5441 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5443 24. Added sender and host information to the "rejected by local_scan()" log
5444 line; previously there was no indication of these.
5446 25. Added .include_if_exists.
5448 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5449 Linux. It turns out that not all file systems support this. Apparently some
5450 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5451 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5452 not supported on the file descriptor, is now ignored when Exim is trying to
5453 sync a directory. This applies only to Linux.
5455 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5457 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5458 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5461 29. Added support for maildirsize files from supplied patch (modified a bit).
5463 30. The use of :fail: followed by an empty string could lead Exim to respond to
5464 sender verification failures with (e.g.):
5466 550 Verification failed for <xxx>
5467 550 Sender verify failed
5469 where the first response line was missing the '-' that indicates it is not
5470 the final line of the response.
5472 31. The loop for finding the name of the user that called Exim had a hardwired
5473 limit of 10; it now uses the value of finduser_retries, which is used for
5474 all other user lookups.
5476 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5479 33. Exim was neglecting to zero errno before one call of strtol() when
5480 expanding a string and expecting an integer value. On some systems this
5481 resulted in spurious "integer overflow" errors. Also, it was casting the
5482 result into an int without checking.
5484 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5485 did not work. The code looks as if it has *never* worked, though it appears
5486 to have been documented since at least releast 1.62. I have made it work.
5488 35. The "timeout_DNS" error in retry rules, also documented since at least
5489 1.62, also never worked. As it isn't clear exactly what this means, and
5490 clearly it isn't a major issue, I have abolished the feature by treating it
5491 as "timeout", and writing a warning to the main and panic logs.
5493 36. The display of retry rules for -brt wasn't always showing the error code
5496 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5497 timeout_connect_A, timeout_connect_MX.
5499 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5500 to the empty sender.
5502 39. The daemon was not analysing the content of -oX till after it had closed
5503 stderr and disconnected from the controlling terminal. This meant that any
5504 syntax errors were only noted on the panic log, and the return code from
5505 the command was 0. By re-arranging the code a little, I've made the
5506 decoding happen first, so such errors now appear on stderr, and the return
5507 code is 1. However, the actual setting up of the sockets still happens in
5508 the disconnected process, so errors there are still only recorded on the
5511 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5512 connections (as happens on some IP stacks) was logged at start up time as
5513 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5514 it from "IPv6 and IPv4", which means that two separate sockets are being
5517 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5518 says whether AF_INET or AF_INET6 was passed as an argument.
5520 42. Exiwhat output was messed up when time zones were included in log
5523 43. Exiwhat now gives more information about the daemon's listening ports,
5524 and whether -tls-on-connect was used.
5526 44. The "port" option of the smtp transport is now expanded.
5528 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5529 silently ignored. Now an error message is written to the main and panic
5532 46. There's a new ACL modifier called "logwrite" which writes to a log file
5533 as soon as it is encountered.
5535 47. Added $local_user_uid and $local_user_gid at routing time.
5537 48. Exim crashed when trying to verify a sender address that was being
5540 49. Exim was recognizing only a space character after ".include". It now also
5541 recognizes a tab character.
5543 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5544 extracting the relevant information from the specification. The man page no
5545 longer contains scrambled data for the -d option, and I've added a section
5546 at the front about calling Exim under different names.
5548 51. Added "extra_headers" argument to the "mail" command in filter files.
5550 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5553 53. Installed eximstats 1.29.
5555 54. Added transport_filter_timeout as a generic transport option.
5557 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5558 Cc: header lines. This was required by RFC 822, but it not required by RFC
5561 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5562 incoming messages that did not have them. Now it does so only if the
5563 message originates locally, that is, if there is no associated remote host
5564 address. When Resent- header lines are present, this applies to the Resent-
5565 lines rather than the non-Resent- lines.
5567 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5568 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5570 58. Messages for configuration errors now include the name of the main
5571 configuration file - useful now that there may be more than one file in a
5572 list (.included file names were always shown).
5574 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5575 for those rare installations that do not start the daemon as root or run it
5576 setuid root. I've cut out the call to initgroups() if the daemon is not
5579 60. The Exim user and group can now be bound into the binary as text strings
5580 that are looked up at the start of Exim's processing.
5582 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5584 62. Added $mailstore_basename variable.
5586 63. Installed patch to sieve.c from Michael Haardt.
5588 64. When Exim failed to open the panic log after failing to open the main log,
5589 the original message it was trying to log was written to stderr and debug
5590 output, but if they were not available (the usual case in production), it
5591 was lost. Now it is written to syslog before the two lines that record the
5592 failures to open the logs.
5594 65. Users' Exim filters run in subprocesses under the user's uid. It is
5595 possible for a "deliver" command or an alias in a "personal" command to
5596 provoke an address rewrite. If logging of address rewriting is configured,
5597 this fails because the process is not running as root or exim. There may be
5598 a better way of dealing with this, but for the moment (because 4.30 needs
5599 to be released), I have disabled address rewrite logging when running a
5600 filter in a non-root, non-exim process.
5606 1. The buildconfig auxiliary program wasn't quoting the value set for
5607 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5608 not defined. This bug was masked in 4.22 by the effect that was fixed in
5611 2. Some messages that were rejected after a message id was allocated were
5612 shown as "incomplete" by exigrep. It no longer does this for messages that
5613 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5615 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5616 have allow_domain_literals set, the ID did not get logged in the <= line.
5617 Domain literals are now always recognized in Message-ID: header lines.
5619 4. The first argument for a ${extract expansion item is the key name or field
5620 number. Leading and trailing spaces in this item were not being ignored,
5621 causing some misleading effects.
5623 5. When deliver_drop_privilege was set, single queue runner processes started
5624 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5625 same command in the process it spins off) were not dropping privilege.
5627 6. When the daemon running as "exim" started a queue runner, it always
5628 re-executed Exim in the spun-off process. This is a waste of effort when
5629 deliver_drop_privilege is set. The new process now just calls the
5630 queue-runner function directly.
5636 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5639 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5640 ignored. Though the use of -oP was forcing the writing of a pid file, it
5641 was always written to the default place.
5643 3. If the message "no IP address found for host xxxx" is generated during
5644 incoming verification, it is now followed by identification of the incoming
5645 connection (so you can more easily find what provoked it).
5647 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5649 5. Added some features to "harden" Exim a bit more against certain attacks:
5651 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5652 be put in Local/Makefile. This is like the never_users runtime option,
5653 but it cannot be overridden. The default setting is "root".
5655 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5656 prefix string with which any file named in a -C command line option
5659 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5660 is retained for -C and -D only if the caller of Exim is root. Without
5661 it, the exim user may also use -C and -D and retain privilege.
5663 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5664 command line option is disabled.
5666 6. Macro names set by the -D option must start with an upper case letter, just
5667 like macro names defined in the configuration file.
5669 7. Added "dereference=" facility to LDAP.
5671 8. Two instances of the typo "uknown" in the source files are fixed.
5673 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5674 the Configure-Makefile script screwed up while processing it.
5676 10. Incorporated PCRE 4.4.
5678 11. The SMTP synchronization check was not operating right at the start of an
5679 SMTP session. For example, it could not catch a HELO sent before the client
5680 waited for the greeting. There is now a check for outstanding input at the
5681 point when the greeting is written. Because of the duplex, asynchronous
5682 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5683 way, but not yet received, when the check is performed.
5685 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5686 on TCP/IP sockets, because this apparently causes some broken clients to
5689 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5690 unchanged) from the Cygwin maintainer.
5692 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5693 was supported instead of testing for mbx. Effectively a typo.
5695 15. The spa authenticator server code was not checking that the input it
5696 received was valid base64.
5698 16. The debug output line for the "set" modifier in ACLs was not showing the
5699 name of the variable that was being set.
5701 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5703 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5704 lookup on its own. Something else was needed to trigger the lookup. For
5705 example, a match in host_lookup or the need for a host name in a host list.
5706 Now, if $sender_host_name is referenced and the host name has not yet been
5707 looked up, a lookup is performed. If the lookup fails, the variable remains
5708 empty, and $host_lookup_failed is set to "1".
5710 19. Added "eqi" as a case-independent comparison operator.
5712 20. The saslauthd authentication condition could segfault if neither service
5713 nor realm was specified.
5715 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5716 error message that was logged was misleading, and incoming SMTP
5717 connections were dropped. The message is now more accurate, and temporary
5718 errors are given to SMTP connections.
5720 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5721 (except RSET) with a 421 error, until QUIT is received. However, it was
5722 failing to send a response to QUIT.
5724 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5725 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5726 ACL is now run after the helo_try_verify_hosts code.
5728 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5729 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5730 case-independent, so other case variants are also recognized). Apparently
5731 some systems use these upper case variants.
5733 25. If more than two messages were waiting for the same host, and a transport
5734 filter was specified for the transport, Exim sent two messages over the
5735 same TCP/IP connection, and then failed with "socket operation on non-
5736 socket" when it tried to send the third.
5738 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5740 27. The extern definition of crypt16() in expand.c was not being excluded when
5741 the OS had its own crypt16() function.
5743 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5744 as a preferred synonym for return_size_limit, both as an option and as an
5747 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5749 30. Changed the default configuration ACL to relax the local part checking rule
5750 for addresses that are not in any local domains. For these addresses,
5751 slashes and pipe symbols are allowed within local parts, but the sequence
5752 /../ is explicitly forbidden.
5754 31. SPA server authentication was not clearing the challenge buffer before
5757 32. log_message in a "warn" ACL statement was writing to the reject log as
5758 well as to the main log, which contradicts the documentation and doesn't
5759 seem right (because no rejection is happening). So I have stopped it.
5761 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5762 However, I am unable to do any testing of this.
5764 34. Fixed an infelicity in the appendfile transport. When checking directories
5765 for a mailbox, to see if any needed to be created, it was accidentally
5766 using path names with one or more superfluous leading slashes; tracing
5767 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5769 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5770 contents are added to the log line that is written for every discarded
5771 recipient. (Previously a log_message setting was ignored.)
5773 36. The ${quote: operator now quotes the string if it is empty.
5775 37. The install script runs exim in order to find its version number. If for
5776 some reason other than non-existence or emptiness, which it checks, it
5777 could not run './exim', it was installing it with an empty version number,
5778 i.e. as "exim-". This error state is now caught, and the installation is
5781 38. An argument was missing from the function that creates an error message
5782 when Exim fails to connect to the socket for saslauthd authentication.
5783 This could cause Exim to crash, or give a corrupted message.
5785 39. Added isip, isip4, and isip6 to ${if conditions.
5787 40. The ACL variables $acl_xx are now saved with the message, and can be
5788 accessed later in routers, transports, and filters.
5790 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5791 strings in the file are not string-expanded.
5793 42. If a MAIL command specified a SIZE value that was too large to fit into an
5794 int variable, the check against message_size_limit failed. Such values are
5795 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5796 day this will have to be increased, but I don't think I want to be around
5797 when emails are that large.
5804 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5805 iconv() is not standard in FreeBSD.
5807 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5808 IPv6 enabled. The observed symptom was a segmentation fault on return from
5809 the function os_common_find_running_interfaces() in src/os.c.
5811 3. In the check_special_case() function in daemon.c I had used "errno" as an
5812 argument name, which causes warnings on some systems. This was basically a
5813 typo, since it was named "eno" in the comments!
5815 4. The code that waits for the clock to tick (at a resolution of some fraction
5816 of a second) so as to ensure message-id uniqueness was always waiting for
5817 at least one whole tick, when it could have waited for less. [This is
5818 almost certainly not relevant at current processor speeds, where it is
5819 unlikely to ever wait at all. But we try to future-proof.]
5821 5. The function that sleeps for a time interval that includes fractions of a
5822 second contained a race. It did not block SIGALRM between setting the
5823 timer, and suspending (a couple of lines later). If the interval was short
5824 and the sigsuspend() was delayed until after it had expired, the suspension
5825 never ended. On busy systems this could lead to processes getting stuck for
5828 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5829 process, before it forks any delivery processes. The open lookup caching
5830 mechanism meant that the open file or database connection was passed into
5831 the delivery process. The problem was that delivery processes always tidy
5832 up cached lookup data. This could cause a problem for the next delivery
5833 process started by the queue runner, because the external queue runner
5834 process does not know about the closure. So the next delivery process
5835 still has data in the lookup cache. In the case of a file lookup, there was
5836 no problem because closing a file descriptor in a subprocess doesn't affect
5837 the parent. However, if the lookup was caching a connection to a database,
5838 the connection was closed, and the second delivery process was likely to
5839 see errors such as "PGSQL: query failed: server closed the connection
5840 unexpectedly". The problem has been fixed by closing all cached lookups
5841 in a queue runner before running a delivery process.
5843 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5844 seem to have the "const" qualifier which it has on other OS. I've
5847 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5848 *of the same type* (client or server) with the same public name that an
5849 error should be diagnosed.
5851 9. When Exim looked up a host name for an IP address, but failed to find the
5852 original IP address when looking up the host name (a safety check), it
5853 output the message "<ip address> does not match any IP for NULL", which was
5854 confusing, to say the least. The bug was that the host name should have
5855 appeared instead of "NULL".
5857 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5858 user that is built into the binary, and the -C or -D options is used, root
5859 privilege is dropped before the configuration file is read. In addition,
5860 logging is switched to stderr instead of the normal log files. If the
5861 configuration then re-defines the Exim user, the unprivileged environment
5862 is probably not what is expected, so Exim logs a panic warning message (but
5865 However, if deliver_drop_privilege is set, the unprivileged state may well
5866 be exactly what is intended, so the warning has been cut out in that case,
5867 and Exim is allowed to try to write to its normal log files.
5873 1. smtp_return_error_details was not giving details for temporary sender
5874 or receiver verification errors.
5876 2. Diagnose a configuration error if two authenticators have the same public
5879 3. Exim used not to create the message log file for a message until the first
5880 delivery attempt. This could be confusing when incoming messages were held
5881 for policy or load reasons. The message log file is now created at the time
5882 the message is received, and an initial "Received" line is written to it.
5884 4. The automatically generated man page for command line options had a minor
5885 bug that caused no ill effects; however, a more serious problem was that
5886 the procedure for building the man page automatically didn't always
5887 operate. Consequently, release 4.20 contains an out-of-date version. This
5888 shouldn't happen again.
5890 5. When building Exim with embedded Perl support, the script that builds the
5891 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5892 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5894 6. The freeze_tell option was not being used for messages that were frozen on
5895 arrival, either by an ACL or by local_scan().
5897 7. Added the smtp_incomplete_transaction log selector.
5899 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5900 was accepting AUTH without a new EHLO.
5902 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5903 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5906 10. Logging of TCP/IP connections (when configured) now happens in the main
5907 daemon process instead of the child process, so that the TCP/IP connection
5908 count is more accurate (but it can never be perfect).
5910 11. The use of "drop" in a nested ACL was not being handled correctly in the
5911 outer ACL. Now, if condition failure induced by the nested "drop" causes
5912 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5913 or "require"), the connection is dropped.
5915 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5916 that yield "discard" can now be used with an "accept" or a "discard" verb,
5917 but an error is generated for any others (because I can't see a useful way
5918 to define what should happen).
5920 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5921 are now processed in the same way as lines in the Exim configuration file.
5922 In particular, continuation lines are supported.
5924 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5926 15. Added -ti meaning -t -i.
5928 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5929 domains, and warn by logging if others are found.
5931 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5932 returned from GIFCONF when Exim is trying to find the list of interfaces on
5933 a host. The code in os.c has been modified to copy each ifreq to an aligned
5934 structure in all cases.
5936 Also, in some cases, the returned ifreq's were being copied to a 'struct
5937 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5938 means the last couple of bytes of an IPv6 address could be chopped if the
5939 ifreq contained only a normal sockaddr (14 bytes storage).
5941 18. Named domain lists were not supported in the hosts_treat_as_local option.
5942 An entry such as +xxxx was not recognized, and was treated as a literal
5945 19. Ensure that header lines added by a DATA ACL are included in the reject log
5946 if the ACL subsequently rejects the message.
5948 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5949 MD5 (which is deprecated).
5951 21. When testing a filter file using -bf, Exim was writing a message when it
5952 took the sender from a "From " line in the message, but it was not doing so
5953 when it took $return_path from a Return-Path: header line. It now does.
5955 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5956 with a valid header line field name (a series of printing characters
5957 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5959 23. Changed "disc" in the source to "disk" to conform to the documentation and
5960 the book and for uniformity.
5962 24. Ignore Sendmail's -Ooption=value command line item.
5964 25. When execve() failed while trying to run a command in a pipe transport,
5965 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5966 could be confused with a return value of 69 from the command itself. This
5967 has been changed to 127, the value the shell returns if it is asked to run
5968 a non-existent command. The wording for the related log line suggests a
5969 non-existent command as the problem.
5971 26. If received_header_text expands to an empty string, do not add a Received:
5972 header line to the message. (Well, it adds a token one on the spool, but
5973 marks it "old" so that it doesn't get used or transmitted.)
5975 27. Installed eximstats 1.28 (addition of -nt option).
5977 28. There was no check for failure on the call to getsockname() in the daemon
5978 code. This can fail if there is a shortage of resources on the system, with
5979 ENOMEM, for example. A temporary error is now given on failure.
5981 29. Contrary to the C standard, it seems that in some environments, the
5982 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5983 program. Exim now does this explicitly; it affects the formatting of
5984 timestamps using strftime().
5986 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5987 complaints. I've now initialized all the variables, to avoid this.
5989 32. Header lines added by a system filter were not being "seen" during
5990 transport-time rewrites.
5992 33. The info_callback() function passed to OpenSSL is set up with type void
5993 (*)(SSL *, int, int), as described somewhere. However, when calling the
5994 function (actually a macro) that sets it up, the type void(*)() is
5995 expected. I've put in a cast to prevent warnings from picky compilers.
5997 34. If a DNS black list lookup found a CNAME record, but there were no A
5998 records associated with the domain it pointed at, Exim crashed.
6000 35. If a DNS black list lookup returned more than one A record, Exim ignored
6001 all but the first. It now scans all returned addresses if a particular IP
6002 value is being sought. In this situation, the contents of the
6003 $dnslist_value variable are a list of all the addresses, separated by a
6006 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6007 to accept a host name and all its aliases if the forward lookup for any of
6008 them yielded the IP address of the incoming connection. Now it accepts only
6009 those names whose forward lookup yields the correct IP address. Any other
6010 names are discarded. This closes a loophole whereby a rogue DNS
6011 administrator could create reverse DNS records to break through a
6012 wildcarded host restriction in an ACL.
6014 37. If a user filter or a system filter that ran in a subprocess used any of
6015 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6016 the wrong values were passed to the pipe command ($thisaddress had the
6017 value of $0, $0 had the value of $1, etc). This bug was introduced by
6018 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6020 38. Improved the line breaking for long SMTP error messages from ACLs.
6021 Previously, if there was no break point between 40 and 75 characters, Exim
6022 left the rest of the message alone. Two changes have been made: (a) I've
6023 reduced the minimum length to 35 characters; (b) if it can't find a break
6024 point between 35 and 75 characters, it looks ahead and uses the first one
6025 that it finds. This may give the occasional overlong line, but at least the
6026 remaining text gets split now.
6028 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6029 file descriptors might be low, and that setting 1000 would always raise it.
6030 It turns out that in some environments, the limit is already over 1000 and
6031 that lowering it causes trouble. So now Exim takes care not to decrease it.
6033 40. When delivering a message, the value of $return_path is set to $sender_
6034 address at the start of routing (routers may change the value). By an
6035 oversight, this default was not being set up when an address was tested by
6036 -bt or -bv, which affected the outcome if any router or filter referred to
6039 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6040 log only when "message" or "log_message" are set. However, if one of the
6041 conditions was an address verification, or a call to a nested ACL, the
6042 messages generated by the underlying test were being passed through. This
6043 no longer happens. The underlying message is available in $acl_verify_
6044 message for both "message" and "log_message" expansions, so it can be
6045 passed through if needed.
6047 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6048 new expansion $bh_ to give the encoded byte string without charset
6049 translation. Translation happens only if iconv() is available; HAVE_ICONV
6050 indicates this at build time. HEADERS_CHARSET gives the charset to
6051 translate to; headers_charset can change it in the configuration, and
6052 "headers charset" can change it in an individual filter file.
6054 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6055 that creates RFC 2047 encoded "words" labels them as that charset instead
6056 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6057 expansion operator; (ii) when Exim creates a From: line for a local
6058 message; (iii) when a header line is rewritten to include a "phrase" part.
6060 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6061 buggy, causing it to skip the first lines of messages whose message ID
6062 ended in 'D'. This would not have bitten before Exim release 4.14, because
6063 message IDs were unlikely to end in 'D' before then. The effect was to have
6064 incorrect size information for certain domains.
6066 45. #include "config.h" was missing at the start of the crypt16.c module. This
6067 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6070 46. If there was a timeout during a "random" callout check, Exim treated it as
6071 a failure of the random address, and carried on sending RSET and the real
6072 address. If the delay was just some slowness somewhere, the response to the
6073 original RCPT would be taken as a response to RSET and so on, causing
6074 mayhem of various kinds.
6076 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6077 when I implemented it. It didn't allow for the fact that some option values
6078 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6079 the right test for positive values.
6081 48. Domain names in DNS records are case-independent. Exim always looks them up
6082 in lower case. Some resolvers return domain names in exactly the case they
6083 appear in the zone file, that is, they may contain uppercase letters. Not
6084 all resolvers do this - some return always lower case. Exim was treating a
6085 change of case by a resolver as a change of domain, similar to a widening
6086 of a domain abbreviation. This triggered its re-routing code and so it was
6087 trying to route what was effectively the same domain again. This normally
6088 caused routing to fail (because the router wouldn't handle the domain
6089 twice). Now Exim checks for this case specially, and just changes the
6090 casing of the domain that it ultimately uses when it transmits the message
6093 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6096 50. If a filter generated a file delivery with a non-absolute name (possible if
6097 no home directory exists for the router), the forbid_file option was not
6100 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6101 the existing equality matching.
6103 52. Exim was using ints instead of ino_t variables in some places where it was
6104 dealing with inode numbers.
6106 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6107 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6108 called TMPDIR, and if it finds it is different, it changes its value.
6110 54. The smtp_printf() function is now made available to local_scan() so
6111 additional output lines can be written before returning. There is also an
6112 smtp_fflush() function to enable the detection of a dropped connection.
6113 The variables smtp_input and smtp_batched_input are exported to
6116 55. Changed the default runtime configuration: the message "Unknown user"
6117 has been removed from the ACL, and instead placed on the localuser router,
6118 using the cannot_route_message feature. This means that any verification
6119 failures that generate their own messages won't get overridden. Similarly,
6120 the "Unrouteable address" message that was in the ACL for unverifiable
6121 relay addresses has also been removed.
6123 56. Added hosts_avoid_esmtp to the smtp transport.
6125 57. The exicyclog script was not checking for the esoteric option
6126 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6127 will work only if exicyclog is run under the appropriate euid.
6129 58. Following a discussion on the list, the rules by which Exim recognises line
6130 endings on incoming messages have been changed. The -dropcr and drop_cr
6131 options are now no-ops, retained only for backwards compatibility. The
6132 following line terminators are recognized: LF CRLF CR. However, special
6133 processing applies to CR:
6135 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6136 nor a local message in the state where . is a terminator.
6138 (ii) If a bare CR is encountered in a header line, an extra space is added
6139 after the line terminator so as not to end the header. The reasoning
6140 behind this is that bare CRs in header lines are most likely either
6141 to be mistakes, or people trying to play silly games.
6143 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6144 was being incorrectly given as 18 bytes larger than it should have been.
6145 This is a VOB (very old bug).
6147 60. This may never have affected anything current, but just in case it has:
6148 When the local host is found other than at the start of a list of hosts,
6149 the local host, those with the same MX, and any that follow, are discarded.
6150 When the list in question was part of a longer list of hosts, the following
6151 hosts (not currently being processed) were also being discarded. This no
6152 longer happens. I'm not sure if this situation could ever has previously
6155 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6158 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6159 header. This is something that is recommended in a new Internet Draft, and
6160 is something that is documented as being done by Sendmail. There are two
6161 possible values. For messages generated by the autoreply transport, Exim
6164 Auto-Submitted: auto-replied
6166 whereas for all other generated messages (e.g. bounces) it adds
6168 Auto-Submitted: auto-generated
6170 63. The "personal" condition in filters now includes a test for the
6171 Auto-Submitted: header. If it contains the string "auto-" the message it
6172 not considered personal.
6174 64. Added rcpt_include_affixes as a generic transport option.
6176 65. Added queue_only_override (default true).
6178 66. Added the syslog_duplication option.
6180 67. If what should have been the first header line of a message consisted of
6181 a space followed by a colon, Exim was mis-interpreting it as a header line.
6182 It isn't of course - it is syntactically invalid and should therefore be
6183 treated as the start of the message body. The misbehaviour could have
6184 caused a number of strange effects, including loss of data in subsequent
6185 header lines, and spool format errors.
6187 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6188 client host had authenticated. This control can now be exercised by an ACL
6189 for more flexibility.
6191 69. By default, callouts do not happen when testing with -bh. There is now a
6192 variant, -bhc, which does actually run the callout code, including
6193 consulting and updating the callout cache.
6195 70. Added support for saslauthd authentication, courtesy of Alexander
6198 71. If statvfs() failed on the spool or log directories while checking their
6199 size for availability, Exim confusingly gave the error "space shortage".
6200 Furthermore, in debugging mode it crashed with a floating point exception.
6201 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6202 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6203 serious problem, Exim now writes to the main and panic logs when this
6204 happens, with details of the failure. It then refuses to accept the
6205 incoming message, giving the message "spool directory problem" or "log
6206 directory problem" with a 421 code for SMTP messages.
6208 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6209 0, 1, and 2 exist, because some OS complain for execs without them (see
6210 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6211 descriptors. However, the code omitted to check that the open succeeded,
6212 causing mysterious errors if for some reason the permissions on /dev/null
6213 got screwed. Now Exim writes a message to the main and panic logs, and
6214 bombs out if it can't open /dev/null.
6216 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6217 interact so that it is all more flexible. It is supposed to remain
6218 backwards compatible. Also added extra_local_interfaces.
6220 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6221 to bomb out with an assertion failure - to the client this appears as a
6222 connection drop. This problem occurs in the part of the code that was taken
6223 from the Samba project. Fortunately, the assertion is in a very simple
6224 function, so I have fixed this by reproducing the function inline in the
6225 one place where it is called, and arranging for authentication to fail
6226 instead of killing the process with assert().
6228 75. The SPA client code was not working when the server requested OEM rather
6229 than Unicode encoding.
6231 76. Added code to make require_files with a specific uid setting more usable in
6232 the case where statting the file as root fails - usually a non-root-mounted
6233 NFS file system. When this happens and the failure is EACCES, Exim now
6234 forks a subprocess and does the per-uid checking as the relevant uid.
6236 77. Added process_log_path.
6238 78. If log_file_path was not explicitly set, a setting of check_log_space or
6239 check_log_inodes was ignored.
6241 79. If a space check for the spool or log partitions fails, the incident is now
6242 logged. Of course, in the latter case the data may get lost...
6244 80. Added the %p formatting code to string_format() so that it can be used to
6245 print addresses in debug_print(). Adjusted all the address printing in the
6246 debugging in store.c to use %p rather than %d.
6248 81. There was a concern that a line of code in smtp_in.c could overflow a
6249 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6250 initially expressed, the concern was not well-founded, because trailing
6251 spaces are removed early. However, if the trailing spaces were followed by
6252 a NULL, they did not get removed, so the overflow was possible. Two fixes
6255 (a) I re-wrote the offending code in a cleaner fashion.
6256 (b) If an incoming SMTP command contains a NULL character, it is rejected
6259 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6260 runs initgroups(), so that if the Exim user is in any additional groups,
6261 they will be used during message reception.
6267 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob