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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
95 TL/01 When running the test suite, the README says that variables such as
96 no_msglog_check are global and can be placed anywhere in a specific
97 test's script, however it was observed that placement needed to be near
98 the beginning for it to behave that way. Changed the runtest perl
99 script to read through the entire script once to detect and set these
100 variables, reset to the beginning of the script, and then run through
101 the script parsing/test process like normal.
103 TL/02 The BSD's have an arc4random API. One of the functions to induce
104 adding randomness was arc4random_stir(), but it has been removed in
105 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
106 function when detected.
108 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
109 cause callback expansion.
111 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
112 syntax errors in an expansion can be treated as a string instead of
113 logging or causing an error, due to the internal use of bool_lax
114 instead of bool when processing it.
116 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
117 server certificates when making smtp deliveries.
119 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
121 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
123 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
125 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
126 Merged patch from Sebastian Wiedenroth.
128 JH/05 Fix results-pipe from transport process. Several recipients, combined
129 with certificate use, exposed issues where response data items split
130 over buffer boundaries were not parsed properly. This eventually
131 resulted in duplicates being sent. This issue only became common enough
132 to notice due to the introduction of conection certificate information,
133 the item size being so much larger. Found and fixed by Wolfgang Breyha.
135 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
136 size buffer was used, resulting in syntax errors when an expansion
139 JH/07 Add support for directories of certificates when compiled with a GnuTLS
140 version 3.3.6 or later.
142 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
143 is EXPERIMENTAL_EVENT, the main-configuration and transport options
144 both become "event_action", the variables become $event_name, $event_data
145 and $event_defer_errno. There is a new variable $verify_mode, usable in
146 routers, transports and related events. The tls:cert event is now also
147 raised for inbound connections, if the main configuration event_action
150 TL/06 In test suite, disable OCSP for old versions of openssl which contained
151 early OCSP support, but no stapling (appears to be less than 1.0.0).
153 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
154 server certificate names available under the smtp transport option
155 "tls_verify_cert_hostname" now do not permit multi-component wildcard
158 JH/10 Time-related extraction expansions from certificates now use the main
159 option "timezone" setting for output formatting, and are consistent
160 between OpenSSL and GnuTLS compilations. Bug 1541.
162 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
163 encoded parameter in the incoming message. Bug 1558.
165 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
166 include certificate info, eximon was claiming there were spoolfile
169 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
171 JH/14 Log delivery-related information more consistently, using the sequence
172 "H=<name> [<ip>]" wherever possible.
174 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
175 are problematic for Debian distribution, omit them from the release
178 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
180 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
183 JH/17 Fix a null-indirection in certextract expansions when a nondefault
184 output list separator was used.
189 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
190 checkers that were complaining about end of non-void function with no
193 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
194 This was a regression intruduced in 4.83 by another bugfix.
196 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
198 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
199 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
205 TF/01 Correctly close the server side of TLS when forking for delivery.
207 When a message was received over SMTP with TLS, Exim failed to clear up
208 the incoming connection properly after forking off the child process to
209 deliver the message. In some situations the subsequent outgoing
210 delivery connection happened to have the same fd number as the incoming
211 connection previously had. Exim would try to use TLS and fail, logging
212 a "Bad file descriptor" error.
214 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
215 utilities have not been installed.
217 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
218 temporary space as the ACL may create new global variables.
220 TL/01 LDAP support uses per connection or global context settings, depending
221 upon the detected version of the libraries at build time.
223 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
224 to extract and use the src ip:port in logging and expansions as if it
225 were a direct connection from the outside internet. PPv2 support was
226 updated based on HAProxy spec change in May 2014.
228 JH/02 Add ${listextract {number}{list}{success}{fail}}.
230 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
231 Properly escape header and check for NULL return.
233 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
236 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
238 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
239 characters in header names, implemented as a verify condition.
240 Contributed by Michael Fischer v. Mollard.
242 TL/05 Rename SPF condition results err_perm and err_temp to standardized
243 results permerror and temperror. Previous values are deprecated but
244 still accepted. In a future release, err_perm and err_temp will be
245 completely removed, which will be a backward incompatibility if the
246 ACL tests for either of these two old results. Patch contributed by
247 user bes-internal on the mailing list.
249 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
251 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
252 selectors, in both main and reject logs.
254 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
257 JH/07 Add malware type "sock" for talking to simple daemon.
259 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
261 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
262 routers/transports under cutthrough routing.
264 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
265 numbers. Touch up "bool" conditional to keep the same definition.
267 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
269 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
271 JH/12 Expand items in router/transport headers_add or headers_remove lists
272 individually rather than the list as a whole. Bug 1452.
274 Required for reasonable handling of multiple headers_ options when
275 they may be empty; requires that headers_remove items with embedded
276 colons must have them doubled (or the list-separator changed).
278 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
279 view the policy declared in the DMARC record. Currently, $dmarc_status
280 is a combined value of both the record presence and the result of the
283 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
285 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
286 dnslookup router and the smtp transport (applying to the forward
289 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
290 of ldap servers used for a specific lookup. Patch provided by Heiko
293 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
294 New variable $lookup_dnssec_authenticated for observability.
296 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
297 Patch submitted by Lars Timman.
299 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
301 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
302 Requires trusted mode and valid format message id, aborts otherwise.
303 Patch contributed by Heiko Schlichting.
305 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
306 certextract with support for various fields. Bug 1358.
308 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
309 is requested by default, modifiable by smtp transport option
312 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
313 operate on certificate variables to give certificate fingerprints
314 Also new ${sha256:cert_variable}.
316 JH/23 The PRDR feature is moved from being Experimental into the mainline.
318 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
319 Christian Aistleitner.
321 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
323 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
324 file. Patch from Wolfgang Breyha.
326 JH/25 Expand the coverage of the delivery $host and $host_address to
327 client authenticators run in verify callout. Bug 1476.
329 JH/26 Port service names are now accepted for tls_on_connect_ports, to
330 align with daemon_smtp_ports. Bug 72.
332 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
333 support and error reporting did not work properly.
335 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
336 and is readable. Patch from Andrew Colin Kissa.
338 TL/14 Enhance documentation of ${run expansion and how it parses the
339 commandline after expansion, particularly in the case when an
340 unquoted variable expansion results in an empty value.
342 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
344 PP/02 Fix internal collision of T_APL on systems which support RFC3123
345 by renaming away from it. Addresses GH issue 15, reported by
348 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
350 TL/15 SECURITY: prevent double expansion in math comparison functions
351 (can expand unsanitized data). Not remotely exploitable.
358 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
360 PP/02 Make -n do something, by making it not do something.
361 When combined with -bP, the name of an option is not output.
363 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
366 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
367 $sender_host_name and config options to manage this, and basic check
370 PP/05 DSCP support for outbound connections and control modifier for inbound.
372 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
373 (Only plugin which currently uses this is kerberos4, which nobody should
374 be using, but we should make it available and other future plugins might
375 conceivably use it, even though it would break NAT; stuff *should* be
376 using channel bindings instead).
378 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
379 name; added for Sendmail compatibility; requires admin caller.
380 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
381 just ignore it); requires trusted caller.
382 Also parse but ignore: -Ac -Am -X<logfile>
385 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
387 TL/02 Add +smtp_confirmation as a default logging option.
389 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
390 Patch by Magnus Holmgren from 2007-02-20.
392 TL/04 Bugzilla 1281 - Spec typo.
393 Bugzilla 1283 - Spec typo.
394 Bugzilla 1290 - Spec grammar fixes.
396 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
398 TL/06 Add Experimental DMARC support using libopendmarc libraries.
400 TL/07 Fix an out of order global option causing a segfault. Reported to dev
401 mailing list by by Dmitry Isaikin.
403 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
405 JH/02 Support "G" suffix to numbers in ${if comparisons.
407 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
409 NM/01 Bugzilla 1197 - Spec typo
410 Bugzilla 1196 - Spec examples corrections
412 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
414 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
415 gnutls_enable_pkcs11, but renamed to more accurately indicate its
418 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
419 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
421 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
422 "acl {{name}{arg}...}", and optional args on acl condition
425 JH/05 Permit multiple router/transport headers_add/remove lines.
427 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
429 JH/07 Avoid using a waiting database for a single-message-only transport.
430 Performance patch from Paul Fisher. Bugzilla 1262.
432 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
435 JH/09 Add $headers_added variable, with content from use of ACL modifier
436 add_header (but not yet added to the message). Bugzilla 199.
438 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
439 Pulled from Bugzilla 817 by Wolfgang Breyha.
441 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
443 (nb: this is the same fix as in Exim 4.80.1)
445 JH/11 Add A= logging on delivery lines, and a client_set_id option on
448 JH/12 Add optional authenticated_sender logging to A= and a log_selector
451 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
453 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
454 advertise SMTP AUTH mechanism to us, instead of a generic
455 protocol violation error. Also, make Exim more robust to bad
456 data from the Dovecot auth socket.
458 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
460 When a queue runner is handling a message, Exim first routes the
461 recipient addresses, during which it prunes them based on the retry
462 hints database. After that it attempts to deliver the message to
463 any remaining recipients. It then updates the hints database using
466 So if a recipient address works intermittently, it can get repeatedly
467 deferred at routing time. The retry hints record remains fresh so the
468 address never reaches the final cutoff time.
470 This is a fairly common occurrence when a user is bumping up against
471 their storage quota. Exim had some logic in its local delivery code
472 to deal with this. However it did not apply to per-recipient defers
473 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
475 This change adds a proper retry rule check during routing so that the
476 final cutoff time is checked against the message's age. We only do
477 this check if there is an address retry record and there is not a
478 domain retry record; this implies that previous attempts to handle
479 the address had the retry_use_local_parts option turned on. We use
480 this as an approximation for the destination being like a local
481 delivery, as in LMTP.
483 I suspect this new check makes the old local delivery cutoff check
484 redundant, but I have not verified this so I left the code in place.
486 TF/02 Correct gecos expansion when From: is a prefix of the username.
488 Test 0254 submits a message to Exim with the header
492 When I ran the test suite under the user fanf2, Exim expanded
493 the header to contain my full name, whereas it should have added
494 a Resent-Sender: header. It erroneously treats any prefix of the
495 username as equal to the username.
497 This change corrects that bug.
499 GF/01 DCC debug and logging tidyup
500 Error conditions log to paniclog rather than rejectlog.
501 Debug lines prefixed by "DCC: " to remove any ambiguity.
503 TF/03 Avoid unnecessary rebuilds of lookup-related code.
505 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
506 Bug spotted by Jeremy Harris; was flawed since initial commit.
507 Would have resulted in OCSP responses post-SNI triggering an Exim
508 NULL dereference and crash.
510 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
512 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
513 Bug detection, analysis and fix by Samuel Thibault.
514 Bugzilla 1331, Debian bug #698092.
516 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
518 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
519 Server implementation by Todd Lyons, client by JH.
520 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
521 config variable "prdr_enable" controls whether the server
522 advertises the facility. If the client requests PRDR a new
523 acl_data_smtp_prdr ACL is called once for each recipient, after
524 the body content is received and before the acl_smtp_data ACL.
525 The client is controlled by bolth of: a hosts_try_prdr option
526 on the smtp transport, and the server advertisement.
527 Default client logging of deliveries and rejections involving
528 PRDR are flagged with the string "PRDR".
530 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
531 fclose(). Diagnosis by Todd Lyons.
533 PP/17 Update configure.default to handle IPv6 localhost better.
534 Patch by Alain Williams (plus minor tweaks).
537 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
538 This is now consistent with GnuTLS, and is now documented: the
539 previous undocumented portable approach to treating the option as
540 unset was to force an expansion failure. That still works, and
541 an empty string is now equivalent.
543 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
544 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
545 not performing validation itself.
547 PP/20 Added force_command boolean option to pipe transport.
548 Patch from Nick Koston, of cPanel Inc.
550 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
553 TF/04 Added udpsend ACL modifer and hexquote expansion operator
555 PP/21 Fix eximon continuous updating with timestamped log-files.
556 Broken in a format-string cleanup in 4.80, missed when I repaired the
557 other false fix of the same issue.
558 Report and fix from Heiko Schlichting.
561 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
562 Report from Prashanth Katuri.
564 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
565 It's SecureTransport, so affects any MacOS clients which use the
566 system-integrated TLS libraries, including email clients.
568 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
569 using a MIME ACL for non-SMTP local injection.
570 Report and assistance in diagnosis by Warren Baker.
572 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
574 JH/16 Fix comparisons for 64b. Bugzilla 1385.
576 TL/09 Add expansion variable $authenticated_fail_id to keep track of
577 last id that failed so it may be referenced in subsequent ACL's.
579 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
582 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
583 ldap library initialization, allowing self-signed CA's to be
584 used. Also properly sets require_cert option later in code by
585 using NULL (global ldap config) instead of ldap handle (per
586 session). Bug diagnosis and testing by alxgomz.
588 TL/12 Enhanced documentation in the ratelimit.pl script provided in
589 the src/util/ subdirectory.
591 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
592 renamed to Transport Post Delivery Action by Jeremy Harris, as
595 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
596 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
597 redis_servers = needs to be configured which will be used by the redis
598 lookup. Patch from Warren Baker, of The Packet Hub.
600 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
602 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
603 hostname or reverse DNS when processing a host list. Used suggestions
604 from multiple comments on this bug.
606 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
608 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
609 Missed a few lines, added it to make the runtest require no keyboard
612 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
613 contains upper case chars. Make router use caseful_local_part.
615 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
616 support when GnuTLS has been built with p11-kit.
622 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
624 This, or similar/improved, will also be change PP/11 of 4.82.
630 PP/01 Handle short writes when writing local log-files.
631 In practice, only affects FreeBSD (8 onwards).
632 Bugzilla 1053, with thanks to Dmitry Isaikin.
634 NM/01 Bugzilla 949 - Documentation tweak
636 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
639 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
641 PP/02 Implemented gsasl authenticator.
643 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
645 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
646 `pkg-config foo` for cflags/libs.
648 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
649 with rest of GSASL and with heimdal_gssapi.
651 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
652 `pkg-config foo` for cflags/libs for the TLS implementation.
654 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
655 properties get this fed in as external SSF. A number of robustness
656 and debugging improvements to the cyrus_sasl authenticator.
658 PP/08 cyrus_sasl server now expands the server_realm option.
660 PP/09 Bugzilla 1214 - Log authentication information in reject log.
661 Patch by Jeremy Harris.
663 PP/10 Added dbmjz lookup type.
665 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
667 PP/12 MAIL args handles TAB as well as SP, for better interop with
668 non-compliant senders.
669 Analysis and variant patch by Todd Lyons.
671 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
672 Bug report from Lars Müller <lars@samba.org> (via SUSE),
673 Patch from Dirk Mueller <dmueller@suse.com>
675 PP/13 tls_peerdn now print-escaped for spool files.
676 Observed some $tls_peerdn in wild which contained \n, which resulted
677 in spool file corruption.
679 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
680 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
681 or write after TLS renegotiation, which otherwise led to messages
684 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
685 as a tracking header (ie: a signed header comes before the signature).
686 Patch from Wolfgang Breyha.
688 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
689 comma-sep list; embedded commas doubled.
691 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
693 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
695 Report and patch from Dmitry Banschikov.
697 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
698 Removed SSL_clear() after SSL_new() which led to protocol negotiation
699 failures. We appear to now support TLS1.1+ with Exim.
701 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
702 lets Exim select keys and certificates based upon TLS SNI from client.
703 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
704 before an outbound SMTP session. New log_selector, +tls_sni.
706 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
707 NULL dereference. Report and patch from Alun Jones.
709 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
710 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
711 Not seeing resolver debug output on NetBSD, but suspect this is a
712 resolver implementation change.
714 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
715 Left warnings. Added "eximon gdb" invocation mode.
717 PP/21 Defaulting "accept_8bitmime" to true, not false.
719 PP/22 Added -bw for inetd wait mode support.
721 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
722 locate the relevant includes and libraries. Made this the default.
724 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
725 Bugzilla 1246, report and most of solution from Tomasz Kusy.
727 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
728 This may cause build issues on older platforms.
730 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
731 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
732 gnutls_require_mac & gnutls_require_protocols (no longer supported).
733 Added SNI support via GnuTLS too.
734 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
736 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
738 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
739 Applied second patch from Janne, implementing suggestion to default
740 multiple-strings-in-record handling to match SPF spec.
742 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
744 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
745 read-only, out of scope).
746 Patch from Wolfgang Breyha, report from Stuart Northfield.
748 PP/29 Fix three issues highlighted by clang analyser static analysis.
749 Only crash-plausible issue would require the Cambridge-specific
750 iplookup router and a misconfiguration.
751 Report from Marcin Mirosław.
753 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
755 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
756 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
757 As part of this, removing so much warning spew let me fix some minor
758 real issues in debug logging.
760 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
761 assignment on my part. Fixed.
763 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
764 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
765 Janne Snabb (who went above and beyond: thank you).
767 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
768 string otherwise requires a connection and a bunch more work and it's
769 relatively easy to get wrong. Should also expose TLS library linkage
772 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
773 64-bit ${eval} (JH/03).
775 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
776 GNU libc to support some of the 64-bit stuff, should not lead to
777 conflicts. Defined before os.h is pulled in, so if a given platform
778 needs to override this, it can.
780 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
781 protection layer was required, which is not implemented.
782 Bugzilla 1254, patch from Wolfgang Breyha.
784 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
785 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
786 tls_dhparam take prime identifiers. Also unbreak combination of
787 OpenSSL+DH_params+TLSSNI.
789 PP/39 Disable SSLv2 by default in OpenSSL support.
795 PP/01 Solaris build fix for Oracle's LDAP libraries.
796 Bugzilla 1109, patch from Stephen Usher.
798 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
800 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
803 TF/02 Fix a couple more cases where we did not log the error message
804 when unlink() failed. See also change 4.74-TF/03.
806 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
807 lock up or crash if it happened to be inside a call to libc when it
808 got a SIGUSR1 from exiwhat.
810 The SIGUSR1 handler appends the current process status to the process
811 log which is later printed by exiwhat. It used to use the general
812 purpose logging code to do this, but several functions it calls are
813 not safe for signals.
815 The new output code in the SIGUSR1 handler is specific to the process
816 log, and simple enough that it's easy to inspect for signal safety.
817 Removing some special cases also simplifies the general logging code.
818 Removing the spurious timestamps from the process log simplifies
821 TF/04 Improved ratelimit ACL condition.
823 The /noupdate option has been deprecated in favour of /readonly which
824 has clearer semantics. The /leaky, /strict, and /readonly update modes
825 are mutually exclusive. The update mode is no longer included in the
826 database key; it just determines when the database is updated. (This
827 means that when you upgrde Exim will forget old rate measurements.)
829 Exim now checks that the per_* options are used with an update mode that
830 makes sense for the current ACL. For example, when Exim is processing a
831 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
832 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
833 must specify per_mail/readonly. If you omit the update mode it defaults to
834 /leaky where that makes sense (as before) or /readonly where required.
836 The /noupdate option is now undocumented but still supported for
837 backwards compatibility. It is equivalent to /readonly except that in
838 ACLs where /readonly is required you may specify /leaky/noupdate or
839 /strict/noupdate which are treated the same as /readonly.
841 A useful new feature is the /count= option. This is a generalization
842 of the per_byte option, so that you can measure the throughput of other
843 aggregate values. For example, the per_byte option is now equivalent
844 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
846 The per_rcpt option has been generalized using the /count= mechanism
847 (though it's more complicated than the per_byte equivalence). When it is
848 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
849 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
850 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
851 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
852 Note that using per_rcpt with a non-readonly update mode in more than
853 one ACL will cause the recipients to be double-counted. (The per_mail
854 and per_byte options don't have this problem.)
856 The handling of very low rates has changed slightly. If the computed rate
857 is less than the event's count (usually one) then this event is the first
858 after a long gap. In this case the rate is set to the same as this event's
859 count, so that the first message of a spam run is counted properly.
861 The major new feature is a mechanism for counting the rate of unique
862 events. The new per_addr option counts the number of different
863 recipients that someone has sent messages to in the last time period. It
864 behaves like per_rcpt if all the recipient addresses are different, but
865 duplicate recipient addresses do not increase the measured rate. Like
866 the /count= option this is a general mechanism, so the per_addr option
867 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
868 example, measure the rate that a client uses different sender addresses
869 with the options per_mail/unique=$sender_address. There are further
870 details in the main documentation.
872 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
874 TF/06 Removed a few PCRE remnants.
876 TF/07 Automatically extract Exim's version number from tags in the git
877 repository when doing development or release builds.
879 PP/02 Raise smtp_cmd_buffer_size to 16kB.
880 Bugzilla 879. Patch from Paul Fisher.
882 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
883 Heavily based on revision 40f9a89a from Simon Arlott's tree.
886 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
888 PP/05 Variable $av_failed, true if the AV scanner deferred.
889 Bugzilla 1078. Patch from John Horne.
891 PP/06 Stop make process more reliably on build failure.
892 Bugzilla 1087. Patch from Heiko Schlittermann.
894 PP/07 Make maildir_use_size_file an _expandable_ boolean.
895 Bugzilla 1089. Patch from Heiko Schlittermann.
897 PP/08 Handle ${run} returning more data than OS pipe buffer size.
898 Bugzilla 1131. Patch from Holger Weiß.
900 PP/09 Handle IPv6 addresses with SPF.
901 Bugzilla 860. Patch from Wolfgang Breyha.
903 PP/10 GnuTLS: support TLS 1.2 & 1.1.
905 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
908 PP/11 match_* no longer expand right-hand-side by default.
909 New compile-time build option, EXPAND_LISTMATCH_RHS.
910 New expansion conditions, "inlist", "inlisti".
912 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
914 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
916 PP/14 fix log_write() format string regression from TF/03.
917 Bugzilla 1152. Patch from Dmitry Isaikin.
923 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
925 PP/02 Harmonised TLS library version reporting; only show if debugging.
926 Layout now matches that introduced for other libraries in 4.74 PP/03.
928 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
930 PP/04 New "dns_use_edns0" global option.
932 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
935 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
936 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
938 TK/01 Updated PolarSSL code to 0.14.2.
939 Bugzilla 1097. Patch from Andreas Metzler.
941 PP/07 Catch divide-by-zero in ${eval:...}.
944 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
947 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
948 format-string attack -- SECURITY: remote arbitrary code execution.
950 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
951 time unintentionally subject to list matching rules, letting the header
952 cause arbitrary Exim lookups (of items which can occur in lists, *not*
953 arbitrary string expansion). This allowed for information disclosure.
955 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
956 INT_MIN/-1 -- value coerced to INT_MAX.
962 NM/01 Workround for PCRE version dependency in version reporting
965 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
966 This fixes portability to compilers other than gcc, notably
967 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
969 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
970 makefiles for portability to HP-UX and POSIX correctness.
972 PP/01 Permit LOOKUP_foo enabling on the make command-line.
973 Also via indented variable definition in the Makefile.
974 (Debugging by Oliver Heesakkers).
976 PP/02 Restore caching of spamd results with expanded spamd_address.
977 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
979 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
980 Improves build reliability. Fix from: Frank Elsner
982 NM/02 Fix wide character breakage in the rfc2047 coding
983 Fixes bug 1064. Patch from Andrey N. Oktyabrski
985 NM/03 Allow underscore in dnslist lookups
986 Fixes bug 1026. Patch from Graeme Fowler
988 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
989 Code patches from Adam Ciarcinski of NetBSD.
991 NM/04 Fixed exiqgrep to cope with mailq missing size issue
994 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
995 is logged, to avoid truncation. Patch from John Horne.
997 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
998 Patch from Jakob Hirsch.
1000 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1001 SQL string expansion failure details.
1002 Patch from Andrey Oktyabrski.
1004 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1005 Patch from Simon Arlott.
1007 PP/09 New lookups functionality failed to compile on old gcc which rejects
1008 extern declarations in function scope.
1009 Patch from Oliver Fleischmann
1011 PP/10 Use sig_atomic_t for flags set from signal handlers.
1012 Check getgroups() return and improve debugging.
1013 Fixed developed for diagnosis in bug 927 (which turned out to be
1016 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1017 Patch from Mark Zealey.
1019 PP/12 Bugzilla 1056: Improved spamd server selection.
1020 Patch from Mark Zealey.
1022 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1023 Based on patch from Heiko Schlittermann.
1025 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1026 Patch from Uwe Doering, sign-off by Michael Haardt.
1028 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1029 variable declaration deep within a block. Bug and patch from
1032 PP/15 lookups-Makefile IRIX compatibilty coercion.
1034 PP/16 Make DISABLE_DKIM build knob functional.
1036 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1037 Patch by Simon Arlott
1039 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1040 variable argument macros. Our copy now differs from upstream.
1046 TF/01 Failure to get a lock on a hints database can have serious
1047 consequences so log it to the panic log.
1049 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1050 controlled using the smtp_confirmation log selector.
1052 TF/03 Include the error message when we fail to unlink a spool file.
1054 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1055 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1056 for maintaining out-of-tree patches for some time.
1058 PP/01 Bugzilla 139: Documentation and portability issues.
1059 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1060 Handle per-OS dynamic-module compilation flags.
1062 PP/02 Let /dev/null have normal permissions.
1063 The 4.73 fixes were a little too stringent and complained about the
1064 permissions on /dev/null. Exempt it from some checks.
1065 Reported by Andreas M. Kirchwitz.
1067 PP/03 Report version information for many libraries, including
1068 Exim version information for dynamically loaded libraries. Created
1069 version.h, now support a version extension string for distributors
1070 who patch heavily. Dynamic module ABI change.
1072 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1073 privilege escalation vulnerability whereby the Exim run-time user
1074 can cause root to append content of the attacker's choosing to
1077 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1080 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1081 If dropping privileges for untrusted macros, we disabled normal logging
1082 on the basis that it would fail; for the Exim run-time user, this is not
1083 the case, and it resulted in successful deliveries going unlogged.
1084 Fixed. Reported by Andreas Metzler.
1090 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1091 only prepend for the Resent-* case. Fixes regression introduced in
1092 Exim 4.70 by NM/22 for Bugzilla 607.
1094 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1095 increasing numbers of administrators be bitten by this.
1097 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1099 PP/03 Bugzilla 994: added openssl_options main configuration option.
1101 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1103 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1105 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1107 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1108 without a peer certificate, leading to a segfault because of an
1109 assumption that peers always have certificates. Be a little more
1110 paranoid. Problem reported by Martin Tscholak.
1112 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1113 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1114 NB: ClamAV planning to remove STREAM in "middle of 2010".
1115 CL also introduces -bmalware, various -d+acl logging additions and
1116 more caution in buffer sizes.
1118 PP/09 Implemented reverse_ip expansion operator.
1120 PP/10 Bugzilla 937: provide a "debug" ACL control.
1122 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1124 PP/12 Bugzilla 973: Implement --version.
1126 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1128 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1130 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1132 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1133 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1134 ignore trailing whitespace.
1136 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1138 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1141 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1142 Notification from Dr Andrew Aitchison.
1144 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1145 ExtendedDetectionInfo response format.
1146 Notification from John Horne.
1148 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1151 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1152 XSL and documented dependency on system catalogs, with examples of how
1155 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1158 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1159 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1160 configuration file which is writeable by the Exim user or group.
1162 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1163 of configuration files to cover files specified with the -C option if
1164 they are going to be used with root privileges, not just the default
1167 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1168 option (effectively making it always true).
1170 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1171 files to be used while preserving root privileges.
1173 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1174 that rogue child processes cannot use them.
1176 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1177 run-time user, instead of root.
1179 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1180 Exim run-time user without dropping privileges.
1182 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1183 result string, instead of calling string_vformat() twice with the same
1186 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1187 for other users. Others should always drop root privileges if they use
1188 -C on the command line, even for a whitelisted configure file.
1190 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1192 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1198 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1199 $data_path, and $header_path variables; fixed documentation bugs and
1202 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1203 exipick to access non-standard spools, including the "frozen" queue
1206 NM/01 Bugzilla 965: Support mysql stored procedures.
1207 Patch from Alain Williams
1209 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1211 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1212 Patch from Andreas Metzler
1214 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1215 Patch from Kirill Miazine
1217 NM/05 Bugzilla 671: Added umask to procmail example.
1219 JJ/03 installed exipick 20100323.0, fixing doc bug
1221 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1222 directory. Notification and patch from Dan Rosenberg.
1224 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1226 TK/02 Improve log output when DKIM signing operation fails.
1228 MH/01 Treat the transport option dkim_domain as a colon separated
1229 list, not as a single string, and sign the message with each element,
1230 omitting multiple occurences of the same signer.
1232 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1233 Bugzilla 985, 986. Patch by Simon Arlott
1235 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1236 Patch by Simon Arlott
1238 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1239 MBX locking. Notification from Dan Rosenberg.
1245 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1247 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1249 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1251 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1253 NM/04 Bugzilla 915: Flag broken perl installation during build.
1259 TK/01 Added patch by Johannes Berg that expands the main option
1260 "spamd_address" if it starts with a dollar sign.
1262 TK/02 Write list of recipients to X-Envelope-Sender header when building
1263 the mbox-format spool file for content scanning (suggested by Jakob
1266 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1267 (http://www.dcc-servers.net/) support via dccifd. Activated by
1268 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1270 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1271 by Mark Daniel Reidel <mr@df.eu>.
1273 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1274 When building exim an external PCRE library is now needed -
1275 PCRE is a system library on the majority of modern systems.
1276 See entry on PCRE_LIBS in EDITME file.
1278 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1279 conversation. Added nologin parameter to request.
1280 Patch contributed by Kirill Miazine.
1282 TF/01 Do not log submission mode rewrites if they do not change the address.
1284 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1286 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1287 log files in place. Contributed by Roberto Lima.
1289 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1291 TF/03 Bugzilla 615: When checking the local_parts router precondition
1292 after a local_part_suffix or local_part_prefix option, Exim now
1293 does not use the address's named list lookup cache, since this
1294 contains cached lookups for the whole local part.
1296 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1297 Robert Millan. Documentation is in experimental-spec.txt.
1299 TF/04 Bugzilla 668: Fix parallel build (make -j).
1301 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1303 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1304 Patch provided by Jan Srzednicki.
1306 TF/05 Leading white space used to be stripped from $spam_report which
1307 wrecked the formatting. Now it is preserved.
1309 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1310 that they are available at delivery time.
1312 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1314 TF/08 TLS error reporting now respects the incoming_interface and
1315 incoming_port log selectors.
1317 TF/09 Produce a more useful error message if an SMTP transport's hosts
1318 setting expands to an empty string.
1320 NM/06 Bugzilla 744: EXPN did not work under TLS.
1321 Patch provided by Phil Pennock.
1323 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1324 Patch provided by Richard Godbee.
1326 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1327 acl_smtp_notquit, added index entry.
1329 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1330 Patch provided by Eugene Bujak.
1332 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1333 accept(). Patch provided by Maxim Dounin.
1335 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1336 Patch provided by Phil Pennock.
1338 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1340 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1341 Patch provided by Brad "anomie" Jorsch.
1343 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1344 Patch provided by Dean Brooks.
1346 TK/05 Add native DKIM support (does not depend on external libraries).
1348 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1349 Patch provided by Graeme Fowler.
1351 NM/16 Bugzilla 851: Documentation example syntax fix.
1353 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1355 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1358 NM/19 Bugzilla 745: TLS version reporting.
1359 Patch provided by Phil Pennock.
1361 NM/20 Bugzilla 167: bool: condition support.
1362 Patch provided by Phil Pennock.
1364 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1365 clients. Patch provided by Phil Pennock.
1367 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1368 Patch provided by Brad "anomie" Jorsch.
1370 NM/23 Bugzilla 687: Fix misparses in eximstats.
1371 Patch provided by Heiko Schlittermann.
1373 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1374 Patch provided by Heiko Schlittermann.
1376 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1377 plus update to original patch.
1379 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1381 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1382 Patch provided by David Brownlee.
1384 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1386 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1388 NM/30 Bugzilla 888: TLS documentation bugfixes.
1390 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1392 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1393 Unlike the original bugzilla I have changed all shell scripts in src tree.
1395 NM/33 Bugzilla 898: Transport filter timeout fix.
1396 Patch by Todd Rinaldo.
1398 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1399 Patch by Serge Demonchaux.
1401 NM/35 Bugzilla 39: Base64 decode bug fixes.
1402 Patch by Jakob Hirsch.
1404 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1406 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1408 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1410 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1416 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1417 ALT-N's libdkim that I have put here:
1418 http://duncanthrax.net/exim-experimental/
1420 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1421 were called 'true' and it seems that C99 defines that as a reserved
1422 keyword to be used with 'bool' variable types. That means you could
1423 not include C99-style headers which use bools without triggering
1424 build errors in sieve.c.
1426 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1427 as mailq or other aliases. Changed the --help handling significantly
1428 to do whats expected. exim_usage() emits usage/help information.
1430 SC/01 Added the -bylocaldomain option to eximstats.
1432 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1434 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1436 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1442 PH/01 Another patch from the Sieve maintainer.
1444 PH/02 When an IPv6 address is converted to a string for single-key lookup
1445 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1446 dots are used instead of colons so that keys in lsearch files need not
1447 contain colons. This was done some time before quoting was made available
1448 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1449 (notated using the quote facility) so as to distinguish them from IPv4
1450 keys. This meant that lookups for IP addresses in host lists did not work
1451 for iplsearch lookups.
1453 This has been fixed by arranging for IPv6 addresses to be expressed with
1454 colons if the lookup type is iplsearch. This is not incompatible, because
1455 previously such lookups could never work.
1457 The situation is now rather anomolous, since one *can* have colons in
1458 ordinary lsearch keys. However, making the change in all cases is
1459 incompatible and would probably break a number of configurations.
1461 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1464 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1465 conversion specification without a maximum field width, thereby enabling
1466 a rogue spamd server to cause a buffer overflow. While nobody in their
1467 right mind would setup Exim to query an untrusted spamd server, an
1468 attacker that gains access to a server running spamd could potentially
1469 exploit this vulnerability to run arbitrary code as the Exim user.
1471 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1472 $primary_hostname instead of what libspf2 thinks the hosts name is.
1474 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1475 a directory entry by the name of the lookup key. Previously, if a
1476 symlink pointed to a non-existing file or a file in a directory that
1477 Exim lacked permissions to read, a lookup for a key matching that
1478 symlink would fail. Now it is enough that a matching directory entry
1479 exists, symlink or not. (Bugzilla 503.)
1481 PH/03 The body_linecount and body_zerocount variables are now exported in the
1484 PH/04 Added the $dnslist_matched variable.
1486 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1487 This means they are set thereafter only if the connection becomes
1490 PH/06 Added the client_condition to authenticators so that some can be skipped
1491 by clients under certain conditions.
1493 PH/07 The error message for a badly-placed control=no_multiline_responses left
1494 "_responses" off the end of the name.
1496 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1498 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1499 (without spaces) instead of just copying the configuration text.
1501 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1503 PH/11 Added $max_received_linelength.
1505 PH/12 Added +ignore_defer and +include_defer to host lists.
1507 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1508 way in which PCRE > 7.0 is built.
1510 PH/14 Implemented queue_only_load_latch.
1512 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1513 MAIL command. The effect was to mangle the value on 64-bit systems.
1515 PH/16 Another patch from the Sieve maintainer.
1517 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1519 PH/18 If a system quota error occurred while trying to create the file for
1520 a maildir delivery, the message "Mailbox is full" was not appended to the
1521 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1522 only to a quota excession during the actual writing of the file.
1524 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1525 characters?) which causes problems in log lines. The DN values are now
1526 passed through string_printing() before being added to log lines.
1528 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1529 and InterBase are left for another time.)
1531 PH/21 Added message_body_newlines option.
1533 PH/22 Guard against possible overflow in moan_check_errorcopy().
1535 PH/23 POSIX allows open() to be a macro; guard against that.
1537 PH/24 If the recipient of an error message contained an @ in the local part
1538 (suitably quoted, of course), incorrect values were put in $domain and
1539 $local_part during the evaluation of errors_copy.
1545 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1546 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1549 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1550 issue a MAIL command.
1552 PH/02 In an ACL statement such as
1554 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1556 if a client was not listed at all, or was listed with a value other than
1557 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1558 the condition was not true (as it should be), so access was not denied.
1559 The bug was that the ! inversion was incorrectly passed on to the second
1560 item. This has been fixed.
1562 PH/03 Added additional dnslists conditions == and =& which are different from
1563 = and & when the dns lookup returns more than one IP address.
1565 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1566 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1568 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1569 FSYNC, which compiles an option called disable_fsync that allows for
1570 bypassing fsync(). The documentation is heavily laced with warnings.
1572 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1574 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1575 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1576 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1577 including adding "make clean"; (3) Added -fPIC when compiling the test
1578 dynamically loaded module, to get rid of a warning.
1580 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1581 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1582 = 0s. The bug is otherwise harmless.
1584 PH/07 There was a bug in the dovecot authenticator such that the value of
1585 $auth1 could be overwritten, and so not correctly preserved, after a
1586 successful authentication. This usually meant that the value preserved by
1587 the server_setid option was incorrect.
1589 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1591 PH/09 Installed PCRE release 7.0.
1593 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1594 run for batched SMTP input. It is now run at the start of every message
1595 in the batch. While fixing this I discovered that the process information
1596 (output by running exiwhat) was not always getting set for -bs and -bS
1597 input. This is fixed, and it now also says "batched" for BSMTP.
1599 PH/11 Added control=no_pipelining.
1601 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1602 patch, slightly modified), and move the expansion of helo_data till after
1603 the connection is made in the smtp transport (so it can use these
1606 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1608 PH/14 Added log_selector = +pid.
1610 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1612 PH/16 Add ${if forany and ${if forall.
1614 PH/17 Added dsn_from option to vary the From: line in DSNs.
1616 PH/18 Flush SMTP output before performing a callout, unless control =
1617 no_callout_flush is set.
1619 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1620 was true (the default) a successful delivery failed to delete the retry
1621 item, thus causing premature timeout of the address. The bug is now
1624 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1626 PH/21 Long custom messages for fakedefer and fakereject are now split up
1627 into multiline reponses in the same way that messages for "deny" and
1628 other ACL rejections are.
1630 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1631 with slight modification.
1633 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1634 draft, changing the syntax and factoring some duplicate code".
1636 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1637 for deliveries of the second and subsequent messages over the same SMTP
1640 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1641 ${reduce, with only minor "tidies".
1643 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1645 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1646 expansion side effects.
1648 PH/27 When a message times out after an over-quota error from an Exim-imposed
1649 quota, the bounce message says "mailbox is full". This message was not
1650 being given when it was a system quota that was exceeded. It now should
1653 MH/03 Made $recipients available in local_scan(). local_scan() already has
1654 better access to the recipient list through recipients_list[], but
1655 $recipients can be useful in postmaster-provided expansion strings.
1657 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1658 in the case of a MAIL command with additional options following the
1659 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1660 were accidentally chopped off.
1662 PH/29 SMTP synchronization checks are implemented when a command is read -
1663 there is a check that no more input is waiting when there shouldn't be
1664 any. However, for some commands, a delay in an ACL can mean that it is
1665 some time before the response is written. In this time, more input might
1666 arrive, invalidly. So now there are extra checks after an ACL has run for
1667 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1668 pipelining has not been advertised.
1670 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1672 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1673 returned in the SMTP response when smtp_return_error_details was false.
1674 This has been fixed.
1676 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1677 instead of the C I/O that was originally supplied, because problems were
1678 reported on Solaris.
1680 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1681 Exim which did not show up earlier: it was assuming that a call to
1682 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1683 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1684 was a macro that became an assignment, so it seemed to work. This has
1685 changed to a proper function call with a void return, hence the compile
1686 error. Exim's code has been fixed.
1688 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1691 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1693 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1695 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1696 because it turns out to be too big or there is a timeout) there is no
1697 "Completed" line in the log. When some messages of this type were
1698 selected by exigrep, they were listed as "not completed". Others were
1699 picked up by some special patterns. I have improved the selection
1700 criteria to be more general.
1702 PH/38 The host_find_failed option in the manualroute router can now be set
1703 to "ignore", to completely ignore a host whose IP address cannot be
1704 found. If all hosts are ignored, the behaviour is controlled by the new
1705 host_all_ignored option.
1707 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1708 homing or because of multiple MX records with /mx) generated more than
1709 one IP address, and the following item turned out to be the local host,
1710 all the secondary addresses of the first item were incorrectly removed
1711 from the list, along with the local host and any following hosts (which
1712 is what is supposed to happen).
1714 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1715 whoever called Exim into the -H file. In the case of the daemon it was
1716 behaving confusingly. When first started, it used values for whoever
1717 started the daemon, but after a SIGHUP it used the Exim user (because it
1718 calls itself on a restart). I have changed the code so that it now always
1721 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1722 message are rejected with the same error (e.g. no authentication or bad
1723 sender address), and a DATA command is nevertheless sent (as can happen
1724 with PIPELINING or a stupid MUA), the error message that was given to the
1725 RCPT commands is included in the rejection of the DATA command. This is
1726 intended to be helpful for MUAs that show only the final error to their
1729 PH/42 Another patch from the Sieve maintainer.
1731 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1732 Eximstats - Fixed some broken HTML links and added missing column headers
1734 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1735 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1737 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1739 PH/43 Yet another patch from the Sieve maintainer.
1741 PH/44 I found a way to check for a TCP/IP connection going away before sending
1742 the response to the final '.' that terminates a message, but only in the
1743 case where the client has not sent further data following the '.'
1744 (unfortunately, this is allowed). However, in many cases there won't be
1745 any further data because there won't be any more messages to send. A call
1746 to select() can be used: if it shows that the input is "ready", there is
1747 either input waiting, or the socket has been closed. An attempt to read
1748 the next input character can distinguish the two cases. Previously, Exim
1749 would have sent an OK response which the client would never have see.
1750 This could lead to message repetition. This fix should cure that, at
1751 least in a lot of common cases.
1753 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1754 advertised in response to EHLO.
1760 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1761 fixed by 4.65/MH/01 (is this a record?) are fixed:
1763 (i) An empty string was always treated as zero by the numeric comparison
1764 operators. This behaviour has been restored.
1766 (ii) It is documented that the numeric comparison operators always treat
1767 their arguments as decimal numbers. This was broken in that numbers
1768 starting with 0 were being interpreted as octal.
1770 While fixing these problems I realized that there was another issue that
1771 hadn't been noticed. Values of message_size_limit (both the global option
1772 and the transport option) were treated as octal if they started with 0.
1773 The documentation was vague. These values are now always treated as
1774 decimal, and I will make that clear in the documentation.
1780 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1781 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1784 MH/01 Don't check that the operands of numeric comparison operators are
1785 integers when their expansion is in "skipping" mode (fixes bug
1786 introduced by 4.64-PH/07).
1788 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1789 child addresses, Exim now panics and dies. Previously, because the count
1790 is held in a short int, deliveries were likely to be lost. As such a
1791 large number of recipients for a single message is ridiculous
1792 (performance will be very, very poor), I have chosen to impose a limit
1793 rather than extend the field.
1799 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1800 leftover -K file (the existence of which was triggered by #402).
1801 While we were at it, introduced process PID as part of the -K
1802 filename. This should rule out race conditions when creating
1805 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1806 processing considerably. Previous code took too long for large mails,
1807 triggering a timeout which in turn triggers #401.
1809 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1810 in the DK code in transports.c. sendfile() is not really portable,
1811 hence the _LINUX specificness.
1813 TF/01 In the add_headers option to the mail command in an Exim filter,
1814 there was a bug that Exim would claim a syntax error in any
1815 header after the first one which had an odd number of characters
1818 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1819 callout verification, Exim cached a "reject" for the entire domain. This
1820 is correct for most verifications, but it is not correct for a recipient
1821 verification with use_sender or use_postmaster set, because in that case
1822 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1823 case of MAIL FROM:<> rejection from other early rejections (e.g.
1824 rejection of HELO). When verifying a recipient using a non-null MAIL
1825 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1826 Whatever the result of the callout, the value of the domain cache is
1827 left unchanged (for any other kind of callout, getting as far as trying
1828 RCPT means that the domain itself is ok).
1830 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1833 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1834 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1835 session, the daemon ignores ECONNECT errors and logs others; it now
1836 ignores EPIPE as well.
1838 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1839 (quoted-printable decoding).
1841 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1842 later the small subsequent patch to fix an introduced bug.
1844 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1846 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1848 PH/08 An error is now given if message_size_limit is specified negative.
1850 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1851 to be given (somewhat) arbitrary names.
1853 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1856 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1857 miscellaneous code fixes
1859 PH/10 Added the log_reject_target ACL modifier to specify where to log
1862 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1863 hostname. This is wrong, because it relates to the incoming message (and
1864 probably the interface on which it is arriving) and not to the outgoing
1865 callout (which could be using a different interface). This has been
1866 changed to use the value of the helo_data option from the smtp transport
1867 instead - this is what is used when a message is actually being sent. If
1868 there is no remote transport (possible with a router that sets up host
1869 addresses), $smtp_active_hostname is used.
1871 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1872 tweaks were necessary in order to get it to work (see also 21 below):
1873 (a) The code assumed that strncpy() returns a negative number on buffer
1874 overflow, which isn't the case. Replaced with Exim's string_format()
1876 (b) There were several signed/unsigned issues. I just did the minimum
1877 hacking in of casts. There is scope for a larger refactoring.
1878 (c) The code used strcasecmp() which is not a standard C function.
1879 Replaced with Exim's strcmpic() function.
1880 (d) The code set only $1; it now sets $auth1 as well.
1881 (e) A simple test gave the error "authentication client didn't specify
1882 service in request". It would seem that Dovecot has changed its
1883 interface. Fortunately there's a specification; I followed it and
1884 changed what the client sends and it appears to be working now.
1886 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1889 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1890 address A is aliased to B and C, where B exists and C does not. Without
1891 -v the output is "A verified" because verification stops after a
1892 successful redirection if more than one address is generated. However,
1893 with -v the child addresses are also verified. Exim was outputting "A
1894 failed to verify" and then showing the successful verification for C,
1895 with its parentage. It now outputs "B failed to verify", showing B's
1896 parentage before showing the successful verification of C.
1898 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1899 look up a TXT record in a specific list after matching in a combined
1902 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1903 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1904 they consult the DNS. I had assumed they would set it the way they
1905 wanted; and indeed my experiments on Linux seem to show that in some
1906 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1907 To be on the safe side, however, I have now made the interface to
1908 host_find_byname() similar to host_find_bydns(), with an argument
1909 containing the DNS resolver options. The host_find_byname() function now
1910 sets these options at its start, just as host_find_bydns() does. The smtp
1911 transport options dns_qualify_single and dns_search_parents are passed to
1912 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1913 of host_find_byname() use the default settings of RES_DEFNAMES
1914 (qualify_single) but not RES_DNSRCH (search_parents).
1916 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1917 spool_read_header() do less string testing, by means of a preliminary
1918 switch on the second character of optional "-foo" lines. (This is
1919 overdue, caused by the large number of possibilities that now exist.
1920 Originally there were few.) While I was there, I also converted the
1921 str(n)cmp tests so they don't re-test the leading "-" and the first
1922 character, in the hope this might squeeze out yet more improvement.
1924 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1925 flag allowing group syntax was set by the header_syntax check but not
1926 turned off, possible causing trouble later; (2) The flag was not being
1927 set at all for the header_verify test, causing "group"-style headers to
1928 be rejected. I have now set it in this case, and also caused header_
1929 verify to ignore an empty address taken from a group. While doing this, I
1930 came across some other cases where the code for allowing group syntax
1931 while scanning a header line wasn't quite right (mostly, not resetting
1932 the flag correctly in the right place). These bugs could have caused
1933 trouble for malformed header lines. I hope it is now all correct.
1935 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1936 with the "reply" argument non-NULL. The code, however (which originally
1937 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1938 but it didn't always do it. This confused somebody who was copying the
1939 code for some other use. I have removed all the tests.
1941 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1942 feature that was used to support insecure browsers during the U.S. crypto
1943 embargo. It requires special client support, and Exim is probably the
1944 only MTA that supported it -- and would never use it because real RSA is
1945 always available. This code has been removed, because it had the bad
1946 effect of slowing Exim down by computing (never used) parameters for the
1947 RSA_EXPORT functionality.
1949 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1950 authenticator to fail if there's a tab character in the incoming data
1951 (there should never be unless someone is messing about, as it's supposed
1952 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1953 if the connection is using TLS or if the remote IP is the same as the
1954 local IP, and the "valid-client-cert option" if a client certificate has
1957 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1958 authenticators. This can be used for authorization after authentication
1959 succeeds. (In the case of plaintext, it servers for both authentication
1962 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1963 if any retry times were supplied.
1965 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1966 connection, where there is no client IP address to check. In this
1967 situation, the verify now always succeeds.
1969 PH/25 Applied John Jetmore's -Mset patch.
1971 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1973 PH/27 In a string expansion for a processed (not raw) header when multiple
1974 headers of the same name were present, leading whitespace was being
1975 removed from all of them, but trailing whitespace was being removed only
1976 from the last one. Now trailing whitespace is removed from each header
1977 before concatenation. Completely empty headers in a concatenation (as
1978 before) are ignored.
1980 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1981 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1983 PH/29 [Removed. This was a change that I later backed out, and forgot to
1984 correct the ChangeLog entry (that I had efficiently created) before
1985 committing the later change.]
1987 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1988 address errors (4xx response to RCPT) over the same connection as other
1989 messages routed to the same hosts. Such deliveries are always "forced",
1990 so retry times are not inspected. This resulted in far too many retries
1991 for the affected addresses. The effect occurred only when there were more
1992 hosts than the hosts_max_try setting in the smtp transport when it had
1993 the 4xx errors. Those hosts that it had tried were not added to the list
1994 of hosts for which the message was waiting, so if all were tried, there
1995 was no problem. Two fixes have been applied:
1997 (i) If there are any address or message errors in an SMTP delivery, none
1998 of the hosts (tried or untried) are now added to the list of hosts
1999 for which the message is waiting, so the message should not be a
2000 candidate for sending over the same connection that was used for a
2001 successful delivery of some other message. This seems entirely
2002 reasonable: after all the message is NOT "waiting for some host".
2003 This is so "obvious" that I'm not sure why it wasn't done
2004 previously. Hope I haven't missed anything, but it can't do any
2005 harm, as the worst effect is to miss an optimization.
2007 (ii) If, despite (i), such a delivery is accidentally attempted, the
2008 routing retry time is respected, so at least it doesn't keep
2009 hammering the server.
2011 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2012 in ${readsocket because some servers need this prod.
2014 PH/32 Added some extra debug output when updating a wait-xxx database.
2016 PH/33 The hint "could be header name not terminated by colon", which has been
2017 given for certain expansion errors for a long time, was not being given
2018 for the ${if def:h_colon_omitted{... case.
2020 PH/34 The spec says: "With one important exception, whenever a domain list is
2021 being scanned, $domain contains the subject domain." There was at least
2022 one case where this was not true.
2024 PH/35 The error "getsockname() failed: connection reset by peer" was being
2025 written to the panic log as well as the main log, but it isn't really
2026 panic-worthy as it just means the connection died rather early on. I have
2027 removed the panic log writing for the ECONNRESET error when getsockname()
2030 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2031 runs only) independently of the message's sender address. This meant
2032 that, if the 4xx error was in fact related to the sender, a different
2033 message to the same recipient with a different sender could confuse
2034 things. In particualar, this can happen when sending to a greylisting
2035 server, but other circumstances could also provoke similar problems.
2036 I have changed the default so that the retry time for these errors is now
2037 based a combination of the sender and recipient addresses. This change
2038 can be overridden by setting address_retry_include_sender=false in the
2041 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2042 remote server are returned as part of bounce messages. This was not
2043 happening for LMTP over a pipe (the lmtp transport), but now it is the
2044 same for both kinds of LMTP.
2046 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2047 in header lines that were in fact CNAMEs. This is no longer the case.
2049 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2050 and queue runs started by the daemon processed all messages. This has
2051 been fixed so that -R and -S can now usefully be given with -q<time>.
2053 PH/40 Import PCRE release 6.7 (fixes some bugs).
2055 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2057 PH/42 Give an error if -q is specified more than once.
2059 PH/43 Renamed the variables $interface_address and $interface_port as
2060 $received_ip_address and $received_port, to make it clear that these
2061 values apply to message reception, and not to the outgoing interface when
2062 a message is delivered. (The old names remain recognized, of course.)
2064 PH/44 There was no timeout on the connect() call when using a Unix domain
2065 socket in the ${readsocket expansion. There now is.
2067 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2068 be meaningful with "accept".
2070 SC/01 Eximstats V1.43
2071 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2073 SC/02 Eximstats V1.44
2074 Use a glob alias rather than an array ref in the generated
2075 parser. This improves both readability and performance.
2077 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2078 Collect SpamAssassin and rejection statistics.
2079 Don't display local sender or destination tables unless
2080 there is data to show.
2081 Added average volumes into the top table text output.
2083 SC/04 Eximstats V1.46
2084 Collect data on the number of addresses (recipients)
2085 as well as the number of messages.
2087 SC/05 Eximstats V1.47
2088 Added 'Message too big' to the list of mail rejection
2089 reasons (thanks to Marco Gaiarin).
2091 SC/06 Eximstats V1.48
2092 Mainlog lines which have GMT offsets and are too short to
2093 have a flag are now skipped.
2095 SC/07 Eximstats V1.49 (Alain Williams)
2096 Added the -emptyok flag.
2098 SC/08 Eximstats V1.50
2099 Fixes for obtaining the IP address from reject messages.
2101 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2102 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2103 whitesspace changes from 4.64-PH/27
2105 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2108 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2109 are found, allow negative numbers in numeric criteria)
2111 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2113 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2114 to match changes made in 4.64-PH/43
2116 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2118 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2119 "domain" is actually an IP address, and force a failure. This locks out
2120 those revolvers/nameservers that support "A-for-A" lookups, in
2121 contravention of the specifications.
2123 PH/48 When a host name was looked up from an IP address, and the subsequent
2124 forward lookup of the name timed out, the host name was left in
2125 $sender_host_name, contrary to the specification.
2127 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2128 restricted to single-key lookups, Exim was not diagnosing an error if
2129 * or *@ was used with a query-style lookup.
2131 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2133 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2134 long ago, but noone interested enough thought of it. Let's just say that
2135 the "1.1" means that there are some new functions that weren't there at
2136 some point in the past.
2138 PH/51 Error processing for expansion failure of helo_data from an smtp
2139 transport during callout processing was broken.
2141 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2142 tested/used via the -bh/-bhc/-bs options.
2144 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2145 bug, fixed in subsequent PCRE releases).
2147 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2148 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2150 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2156 SC/01 Use a glob alias rather than an array ref in eximstats generated
2157 parser. This improves both readability and performance.
2159 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2160 Don't display local sender or destination tables in eximstats unless
2161 there is data to show.
2162 Added average volumes into the eximstats top table text output.
2164 SC/03 Collect data on the number of addresses (recipients) as well
2165 as the number of messages in eximstats.
2167 TF/01 Correct an error in the documentation for the redirect router. Exim
2168 does (usually) call initgroups() when daemonizing.
2170 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2171 with consistent privilege compared to when running as a daemon.
2173 TF/03 Note in the spec that $authenticated_id is not set for local
2174 submissions from trusted users.
2176 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2177 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2179 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2180 by adding some example configuration directives to the default
2181 configuration file. A little bit of work is required to uncomment the
2182 directives and define how usernames and passwords are checked, but
2183 there is now a framework to start from.
2185 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2186 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2187 without this. I don't know how relevant this is to other LDAP libraries.
2189 PH/02 Add the verb name to the "unknown ACL verb" error.
2191 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2193 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2195 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2196 directory not expanded when it should be if an expanded home directory
2197 was set for the address (which is overridden by the transport).
2199 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2202 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2203 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2204 because it is too late at that time, and has no effect.
2206 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2207 security issue with \' (bugzilla #107). I could not use the
2208 PQescapeStringConn() function, because it needs a PGconn value as one of
2211 PH/08 When testing addresses using -bt, indicate those final addresses that
2212 are duplicates that would not cause an additional delivery. At least one
2213 person was confused, thinking that -bt output corresponded to deliveries.
2214 (Suppressing duplicates isn't a good idea as you lose the information
2215 about possibly different redirections that led to the duplicates.)
2217 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2218 systems where poll() doesn't work, in particular OS X.
2220 PH/10 Added more information to debugging output for retry time not reached.
2222 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2223 operations in malware.c.
2225 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2228 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2229 syslog_duplication set false, log lines that would normally be written
2230 both the the main log and to the reject log were not written to syslog at
2233 PH/14 In the default configuration, change the use of "message" in ACL warn
2234 statements to "add_header".
2236 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2237 not followed by a command (e.g. "seen endif").
2239 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2240 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2243 PH/17 Added extra conditions to the default value of delay_warning_condition
2247 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2248 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2249 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2252 The Auto-Submitted: and various List- headers are standardised, whereas I
2253 don't think Precedence: ever was.
2255 PH/18 Refactored debugging code in route_finduser() to show more information,
2256 in particular, the error code if getpwnam() issues one.
2258 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2259 This is apparently needed in addition to the PH/07 change above to avoid
2260 any possible encoding problems.
2262 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2263 but not after initializing Perl.
2265 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2266 output them only if debugging. By default they are written stderr,
2267 apparently, which is not desirable.
2269 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2272 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2275 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2277 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2278 authenticated or an ident call has been made. Suppress the default
2279 values for $authenticated_id and $authenticated_sender (but permit -oMai
2280 and -oMas) when testing with -bh.
2282 PH/24 Re-jigged the order of the tests in the default configuration so that the
2283 tests for valid domains and recipients precede the DNS black list and CSA
2284 tests, on the grounds that those ones are more expensive.
2286 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2287 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2288 This bug exists in every version of Exim that I still have, right back to
2291 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2292 However, an attempt to turn on case-sensitivity in a regex key by
2293 including (?-i) didn't work because the subject string was already
2294 lowercased, and the effects were non-intuitive. It turns out that a
2295 one-line patch can be used to allow (?-i) to work as expected.
2301 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2302 other effects) broke the use of negated acl sub-conditions.
2304 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2307 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2308 "Deny" causes Exim to reject the incoming connection with a 554 error.
2309 Unfortunately, if there is a major crisis, such as a disk failure,
2310 tcp-wrappers gives "deny", whereas what one would like would be some
2311 kind of temporary error. A kludge has been added to help with this.
2312 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2313 554 error is used if errno is still zero or contains ENOENT (which occurs
2314 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2317 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2319 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2320 errors. Otherwise a message that provokes a temporary error (when other
2321 messages do not) can cause a whole host to time out.
2323 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2324 addresses were routed directly to files or pipes from a redirect router.
2325 File deliveries just didn't batch; pipe deliveries might have suffered
2328 PH/06 A failure to get a lock for a hints database would erroneously always say
2329 "Failed to get write lock", even when it was really a read lock.
2331 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2332 of 0600. This has been changed to use the value of the lockfile_mode
2333 option (which defaults to 0600).
2335 PH/08 Applied small patch from the Sieve maintainer.
2337 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2338 folder from quota calculations, a direct delivery into this folder messed
2339 up the contents of the maildirsize file. This was because the regex was
2340 used only to exclude .Trash (or whatever) when the size of the mailbox
2341 was calculated. There was no check that a delivery was happening into an
2342 excluded directory. This bug has been fixed by ignoring all quota
2343 processing for deliveries into excluded directories.
2345 PH/10 Added the maildirfolder_create_regex option to appendfile.
2351 PH/01 The code for finding all the local interface addresses on a FreeBSD
2352 system running IPv6 was broken. This may well have applied to all BSD
2353 systems, as well as to others that have similar system calls. The broken
2354 code found IPv4 interfaces correctly, but gave incorrect values for the
2355 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2356 that it would not match correctly against @[] and not recognize the IPv6
2359 PH/02 The ipliteral router was not recognizing addresses of the form user@
2360 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2362 PH/03 Added disable_ipv6.
2364 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2365 decoded form, because it is most often used to construct To: headers
2366 lines in autoreplies, and the decoded form may well be syntactically
2367 invalid. However, $reply_address has leading white space removed, and all
2368 newlines turned into spaces so that the autoreply transport does not
2371 PH/05 If group was specified without a user on a router, and no group or user
2372 was specified on a transport, the group from the router was ignored.
2374 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2375 for visible compile-time settings that can be used to change these
2376 numbers, for those that want even more. Backwards compatibility with old
2377 spool files has been maintained. However, going back to a previous Exim
2378 release will lost any variables that are in spool files.
2380 PH/07 Two small changes when running in the test harness: increase delay when
2381 passing a TCP/IP connection to a new process, in case the original
2382 process has to generate a bounce, and remove special handling of
2383 127.0.0.2 (sic), which is no longer necessary.
2385 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2386 be the same on different OS.
2388 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2391 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2392 whether --show-vars was specified or not
2394 JJ/02 exipick: Added support for new ACL variable spool format introduced
2397 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2398 syntactically invalid From: or Reply-to: line, and a filter used this to
2399 generate an autoreply, and therefore failed to obtain an address for the
2400 autoreply, Exim could try to deliver to a non-existent relative file
2401 name, causing unrelated and misleading errors. What now happens is that
2402 it logs this as a hard delivery error, but does not attempt to create a
2405 PH/11 The exinext utility has a -C option for testing purposes, but although
2406 the given file was scanned by exinext itself; it wasn't being passed on
2407 when Exim was called.
2409 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2410 an end-of-file indication when reading a command response.
2412 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2413 compiled. In many other places in Exim, IPv6 addresses are always
2414 recognized, so I have changed this. It also means that IPv4 domain
2415 literals of the form [IPV4:n.n.n.n] are now always recognized.
2417 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2418 used if the router is not running as root, for example, when verifying at
2419 ACL time, or when using -bh. The debugging output from this situation was
2420 non-existent - all you got was a failure to exec. I have made two
2423 (a) Failures to set uid/gid, the current directory, or a process leader
2424 in a subprocess such as that created by queryprogram now generate
2425 suitable debugging ouput when -d is set.
2427 (b) The queryprogram router detects when it is not running as root,
2428 outputs suitable debugging information if -d is set, and then runs
2429 the subprocess without attempting to change uid/gid.
2431 PH/15 Minor change to Makefile for building test_host (undocumented testing
2434 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2435 additional section of a DNS packet that returns MX or SRV records.
2436 Instead, it always explicitly searches for A/AAAA records. This avoids
2437 major problems that occur when a DNS server includes only records of one
2438 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2439 fixed another bug: if SRV records were looked up and the corresponding
2440 address records were *not* found in the additional section, the port
2441 values from the SRV records were lost.
2443 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2444 using the correct key (the original address) when searching the retry
2445 rules in order to find which one to use for generating the retry hint.
2447 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2448 adding the default one. Similarly, if it contains a Reply-To: header, the
2449 errors_reply_to option, if set, is not used.
2451 PH/19 When calculating a retry time, Exim used to measure the "time since
2452 failure" by looking at the "first failed" field in the retry record. Now
2453 it does not use this if it is later than than the arrival time of the
2454 message. Instead it uses the arrival time. This makes for better
2455 behaviour in cases where some deliveries succeed, thus re-setting the
2456 "first failed" field. An example is a quota failure for a huge message
2457 when small messages continue to be delivered. Without this change, the
2458 "time since failure" will always be short, possible causing more frequent
2459 delivery attempts for the huge message than are intended.
2460 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2462 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2463 $1, $2, $3) because the numerical variables can be reset during some
2464 expansion items (e.g. "match"), thereby losing the authentication data.
2466 PH/21 Make -bV show the size of off_t variables so that the test suite can
2467 decide whether to run tests for quotas > 2G.
2469 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2470 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2471 filecount value is greater than 2G or if a quota value is greater than 2G
2472 on a system where the size of off_t is not greater than 4, a panic error
2475 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2476 never match. The debug and -bh output now contains an explicit error
2477 message indicating a malformed IPv4 address or mask.
2479 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2480 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2481 PH/23 above applies.
2483 PH/25 Do not write to syslog when running in the test harness. The only
2484 occasion when this arises is a failure to open the main or panic logs
2485 (for which there is an explicit test).
2487 PH/26 Added the /no_tell option to "control=freeze".
2489 PH/27 If a host name lookup failed very early in a connection, for example, if
2490 the IP address matched host_lookup and the reverse lookup yielded a name
2491 that did not have a forward lookup, an error message of the form "no IP
2492 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2493 could be logged. Now it outputs the IP address instead of "NULL".
2495 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2496 allows the sender and the authenticated sender to be set when
2497 submitting a message from within Exim. Since child_open_exim() is
2498 documented for local_scan(), the new function should be too.
2500 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2501 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2502 results in an empty string is now treated as unset.
2504 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2506 PH/31 Added sender_verify_fail logging option.
2508 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2509 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2510 tidied the source and removed it altogether.
2512 PH/33 When a queue run was abandoned because the load average was too high, a
2513 log line was always written; now it is written only if the queue_run log
2514 selector is set. In addition, the log line for abandonment now contains
2515 information about the queue run such as the pid. This is always present
2516 in "start" and "stop" lines but was omitted from the "abandon" line.
2518 PH/34 Omit spaces between a header name and the colon in the error message that
2519 is given when verify = headers_syntax fails (if there are lots of them,
2520 the message gets confusing).
2522 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2523 names, as there are now some PTR records that contain slashes. This check
2524 is only to protect against broken name servers that fall over on strange
2525 characters, so the fact that it applies to all lookups doesn't matter.
2527 PH/36 Now that the new test suite is complete, we can remove some of the
2528 special code in Exim that was needed for the old test suite. For example,
2529 sorting DNS records because real resolvers return them in an arbitrary
2530 order. The new test suite's fake resolver always returns records in the
2533 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2534 bounces) except when queue_only is set, to avoid logging races between
2535 the different processes.
2537 PH/38 Panic-die if .include specifies a non-absolute path.
2539 PH/39 A tweak to the "H" retry rule from its user.
2541 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2542 a label. They prevented compilation on older perls.
2544 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2545 a warning to be raised on newish perls.
2547 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2548 on queue. Changes to match documented behaviour of showing count of
2549 messages matching specified criteria.
2551 PH/40 Changed the default ident timeout from 30s to 5s.
2553 PH/41 Added support for the use of login_cap features, on those BSD systems
2554 that have them, for controlling the resources used by pipe deliveries.
2556 PH/42 The content-scanning code uses fopen() to create files in which to put
2557 message data. Previously it was not paying any attention to the mode of
2558 the files. Exim runs with umask(0) because the rest of the code creates
2559 files with open(), and sets the required mode explicitly. Thus, these
2560 files were ending up world-writeable. This was not a big issue, because,
2561 being within the spool directory, they were not world-accessible. I have
2562 created a function called modefopen, which takes an additional mode
2563 argument. It sets umask(777), creates the file, chmods it to the required
2564 mode, then resets the umask. All the relevant calls to fopen() in the
2565 content scanning code have been changed to use this function.
2567 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2568 to 24 hours. This avoids potential overflow problems when processing G
2569 and H retry rules. I suspect nobody ever tinkers with this value.
2571 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2573 PH/45 When the plaintext authenticator is running as a client, the server's
2574 challenges are checked to ensure they are valid base64 strings. By
2575 default, the authentication attempt is cancelled if an invalid string is
2576 received. Setting client_ignore_invalid_base64 true ignores these errors.
2577 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2578 they are received. Thus, the responses can be made to depend on the
2579 challenges. If an invalid string is ignored, an empty string is placed in
2582 PH/46 Messages that are created by the autoreply transport now contains a
2583 References: header, in accordance with RFCs 2822 and 3834.
2585 PH/47 Added authenticated_sender_force to the smtp transport.
2587 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2589 PH/49 Installed latest patch from the Sieve maintainer.
2591 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2592 was also set, the appendfile transport was unnecessarily scanning a
2593 directory of message files (e.g. for maildir delivery) to find the count
2594 of files (along with the size), even though it did not need this
2595 information. It now does the scan only if it needs to find either the
2596 size of the count of files.
2598 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2600 PH/52 Two bugs concerned with error handling when the smtp transport is
2603 (i) Exim was not creating retry information for temporary errors given
2604 for individual recipients after the DATA command when the smtp transport
2605 was used in LMTP mode. This meant that they could be retried too
2606 frequently, and not timed out correctly.
2608 (ii) Exim was setting the flag that allows error details to be returned
2609 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2610 recipients that were returned after the DATA command.
2612 PH/53 This is related to PH/52, but is more general: for any failing address,
2613 when detailed error information was permitted to be returned to the
2614 sender, but the error was temporary, then after the final timeout, only
2615 "retry timeout exceeded" was returned. Now it returns the full error as
2616 well as "retry timeout exceeded".
2618 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2619 do this, and (what is worse) MTAs that accept it.
2621 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2622 will now be deprecated.
2624 PH/56 New os.c-cygwin from the Cygwin maintainer.
2626 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2627 formats (previously only available in exim formats via -bpr, -bpru,
2628 and -bpra. Now also available in native and exiqgrep formats)
2630 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2631 with very large, slow to parse queues
2633 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2635 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2637 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2638 responses to authentication challenges, though it was showing the
2639 challenges; (ii) I've removed the CR characters from the debug output for
2642 PH/58 Allow for the insertion of a newline as well as a space when a string
2643 is turned into more than one encoded-word during RFC 2047 encoding. The
2644 Sieve code now uses this.
2646 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2647 data_4xx, lost_connection, tls_required.
2649 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2650 message was being sent as an SMTP response.
2652 PH/61 Add -l and -k options to exicyclog.
2654 PH/62 When verifying, if an address was redirected to one new address, so that
2655 verification continued, and the new address failed or deferred after
2656 having set something in $address_data, the value of $address_data was not
2657 passed back to the ACL. This was different to the case when no
2658 redirection occurred. The value is now passed back in both cases.
2660 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2661 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2662 use login_cap.h, so on its own it isn't the distinguishing feature. The
2663 new name refers directly to the setclassresources() function.
2665 PH/65 Added configuration files for NetBSD3.
2667 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2669 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2670 is preferred over IPv4.
2672 PH/68 The bounce_return_message and bounce_return_body options were not being
2673 honoured for bounces generated during the reception of non-SMTP messages.
2674 In particular, this applied to messages rejected by the ACL. This bug has
2675 been fixed. However, if bounce_return_message is true and bounce_return_
2676 body is false, the headers that are returned for a non-SMTP message
2677 include only those that have been read before the error was detected.
2678 (In the case of an ACL rejection, they have all been read.)
2680 PH/69 The HTML version of the specification is now built in a directory called
2681 spec_html instead of spec.html, because the latter looks like a path with
2682 a MIME-type, and this confuses some software.
2684 PH/70 Catch two compiler warnings in sieve.c.
2686 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2687 function verify_get_ident() calls ip_connect() to connect a socket, but
2688 if the "connect()" function timed out, ip_connect() used to close the
2689 socket. However, verify_get_ident() also closes the socket later, and in
2690 between Exim writes to the log, which may get opened at this point. When
2691 the socket was closed in ip_connect(), the log could get the same file
2692 descriptor number as the socket. This naturally causes chaos. The fix is
2693 not to close the socket in ip_connect(); the socket should be closed by
2694 the function that creates it. There was only one place in the code where
2695 this was missing, in the iplookup router, which I don't think anybody now
2696 uses, but I've fixed it anyway.
2698 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2699 well as to direct DNS lookups. Otherwise the handling of names in host
2700 lists is inconsistent and therefore confusing.
2706 PH/01 Two changes to the default runtime configuration:
2708 (1) Move the checks for relay_from_hosts and authenticated clients from
2709 after to before the (commented out) DNS black list checks.
2711 (2) Add control=submission to the relay_from_hosts and authenticated
2712 clients checks, on the grounds that messages accepted by these
2713 statements are most likely to be submissions.
2715 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2717 (1) Generate an error if the third argument for the ${prvs expansion is
2720 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2723 (3) Reset the variables that are obtained from the first argument of
2724 ${prvscheck and used in the second argument before leaving the code,
2725 because their memory is reclaimed, so using them afterwards may do
2728 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2729 one (it's much easier than Tom thought :-).
2731 (5) Because of (4), we can now allow for the use of $prvscheck_result
2732 inside the third argument.
2734 PH/03 For some reason, the default setting of PATH when running a command from
2735 a pipe transport was just "/usr/bin". I have changed it to
2738 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2739 anything to be listed in the output from -bV.
2741 PH/05 When a filter generated an autoreply, the entire To: header line was
2742 quoted in the delivery log line, like this:
2744 => >A.N.Other <ano@some.domain> <original@ddress> ...
2746 This has been changed so that it extracts the operative address. There
2747 may be more than one such address. If so, they are comma-separated, like
2750 => >ano@some.domain,ona@other.domain <original@ddress> ...
2752 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2753 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2754 not being looked up in the rDNS to get a host name, Exim was showing the
2755 IP address twice in Received: lines, even though the IP addresses were
2756 identical. For example:
2758 Received: from [1.2.3.4] (helo=[1.2.3.4])
2760 However, if the real host name was known, it was omitting the HELO data
2761 if it matched the actual IP address. This has been tidied up so that it
2762 doesn't show the same IP address twice.
2764 PH/07 When both +timestamp and +memory debugging was on, the value given by
2765 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2766 called by the debug printing, thereby overwriting the timestamp buffer.
2767 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2769 PH/08 When the original message was included in an autoreply transport, it
2770 always said "this is a copy of the message, including all the headers",
2771 even if body_only or headers_only was set. It now gives an appropriate
2774 PH/09 Applied a patch from the Sieve maintainer which:
2776 o fixes some comments
2777 o adds the (disabled) notify extension core
2778 o adds some debug output for the result of if/elsif tests
2779 o points to the current vacation draft in the documentation
2780 and documents the missing references header update
2784 o fixes a bug in processing the envelope test (when testing
2785 multiple envelope elements, the last element determinted the
2788 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2789 Electronic Mail") by including:
2791 Auto-submitted: auto-generated
2793 in the messages that it generates (bounce messages and others, such as
2794 warnings). In the case of bounce messages for non-SMTP mesages, there was
2795 also a typo: it was using "Auto_submitted" (underscore instead of
2796 hyphen). Since every message generated by Exim is necessarily in response
2797 to another message, thes have all been changed to:
2799 Auto-Submitted: auto-replied
2801 in accordance with these statements in the RFC:
2803 The auto-replied keyword:
2805 - SHOULD be used on messages sent in direct response to another
2806 message by an automatic process,
2808 - MUST NOT be used on manually-generated messages,
2810 - MAY be used on Delivery Status Notifications (DSNs) and Message
2811 Disposition Notifications (MDNs),
2813 - MUST NOT be used on messages generated by automatic or periodic
2814 processes, except for messages which are automatic responses to
2817 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2818 to the default Received: header definition.
2820 PH/12 Added log selector acl_warn_skipped (default on).
2822 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2823 variables because (a) they are in the wrong storage pool and (b) even if
2824 they were copied, it wouldn't work properly because of the caching.
2826 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2827 checking when decoding. Apparently there are clients that generate
2828 overlong encoded strings. Why am I not surprised?
2830 PH/15 If the first argument of "${if match_address" was not empty, but did not
2831 contain an "@" character, Exim crashed. Now it writes a panic log message
2832 and treats the condition as false.
2834 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2836 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2837 extension "envelope-auth". The code is finished and in agreement with
2838 other implementations, but there is no documentation so far and in fact,
2839 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2840 not changing the active code.
2842 Print executed "if" and "elsif" statements when debugging is used. This
2843 helps a great deal to understand what a filter does.
2845 Document more things not specified clearly in RFC3028. I had all this
2846 sorted out, when out of a sudden new issues came to my mind. Oops."
2848 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2851 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2852 canonical form (as documented). However, after a host name lookup from
2853 the IP address, check_host() was doing a simple string comparison with
2854 addresses acquired from the DNS when checking that the found name did
2855 have the original IP as one of its addresses. Since any found IPv6
2856 addresses are likely to be in abbreviated form, the comparison could
2857 fail. Luckily, there already exists a function for doing the comparison
2858 by converting both addresses to binary, so now that is used instead of
2859 the text comparison.
2861 PH/20 There was another similar case to PH/19, when a complete host name was
2862 given in a host list; looking up its IP address could give an abbreviated
2863 form, whereas the current host's name might or might not be abbreviated.
2864 The same fix has been applied.
2870 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2871 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2874 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2875 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2877 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2879 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2880 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2881 Also, for "Auto-Submitted", treat anything other than "no" as
2882 non-personal, in accordance with RFC 3834. (Previously it treated
2883 anything starting "auto-" as non-personal.)
2885 TF/01 The control=submission/name=... option had a problem with syntax
2886 errors if the name included a slash character. The /name= option
2887 now slurps the rest of the string, so it can include any characters
2888 but it must come last in the list of options (after /sender_retain
2891 PH/05 Some modifications to the interface to the fake nameserver for the new
2899 TK/01 Added the "success_on_redirect" address verification option. See
2900 NewStuff for rationale and an example.
2902 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2904 PH/02 Patch to exigrep to allow it to work on syslog lines.
2906 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2907 fread() to skip over the body file's header line, because in Cygwin the
2908 header line is locked and is inaccessible.
2910 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2911 co-exist for some time) to make it clear that it is the Exim ID that is
2912 referenced, not the Message-ID: header line.
2914 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2915 string_format() function, because snprintf() does not exist on all
2918 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2919 ${stat: expansion item.
2921 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2922 protocol synchronization error", to keep the pedants happy.
2924 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2925 well as for IRIX systems, when gcc is being used. See the host.c source
2928 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2930 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2933 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2934 transport and to the smtp transport in LMTP mode.
2936 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2938 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2939 run a filter in a subprocess. This could lead to confusion in subsequent
2940 lookups in the parent process. There should also be a search_tidyup() at
2941 the end of the subprocess.
2943 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2944 only if the host matched helo_try_verify_hosts, which caused the
2945 verification to occur when the EHLO/HELO command was issued. The ACL just
2946 tested the remembered result. Now, if a previous verification attempt has
2947 not happened, "verify = helo" does it there and then.
2949 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2951 TK/03 Fix log output including CR from clamd.
2953 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2954 exist provoked a memory error which could cause a segfault.
2956 PH/15 Installed PCRE 6.2
2958 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2960 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2961 of the problem. Specifically, suggested +O2 rather than +O1 for the
2964 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2966 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2967 record, the port was not correctly logged when the outgoing_port log
2968 selector was set (it logged the transort's default port).
2970 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2971 fallback_hosts, and "hosts" in the smtp transport.
2973 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2974 host errors such as "Connection refused".
2976 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2977 authentication with radiusclient 0.4.9:
2979 - Error returned from rc_read_config was caught wrongly
2980 - Username/password not passed on to radius server due to wrong length.
2982 The presumption is that some radiusclient API changes for 4.51/PH/17
2983 were not taken care of correctly. The code is still untested by me (my
2984 Linux distribution still has 0.3.2 of radiusclient), but it was
2985 contributed by a Radius user.
2987 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2988 expanding the "port" option of the smtp transport.
2990 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2991 while reading a MIME header. Thanks to Tom Hughes for a patch.
2993 PH/24 Include config.h inside local_scan.h so that configuration settings are
2996 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2997 that in an ACL for RCPT (for example), you can examine exactly what was
3000 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3001 commands, but it was not correctly comparing the address with the actual
3002 client host address. Thus, it would show the EHLO address in Received:
3003 header lines when this was not necessary.
3005 PH/27 Added the % operator to ${eval:}.
3007 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3008 it should be ignoring failures (because with -C, for example, it has lost
3009 privilege). It wasn't ignoring creation failures other than "already
3012 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3015 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3016 because an input line was too long, either on its own, or by virtue of
3017 too many continuations, the temporary file was not being removed, and the
3018 return code was incorrect.
3020 PH/31 Missing "BOOL" in function definition in filtertest.c.
3022 PH/32 Applied Sieve patches from the maintainer.
3024 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3026 PH/33 Added "verify = not_blind".
3028 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3029 Local/Makefile (with some defaults set). These are used in built scripts
3030 such as exicyclog, but they have never been used in the exim_install
3031 script (though there are many overriding facilities there). I have
3032 arranged that the exim_install script now takes note of these two
3035 PH/35 Installed configuration files for Dragonfly.
3037 PH/36 When a locally submitted message by a trusted user did not contain a
3038 From: header, and the sender address was obtained from -f or from an SMTP
3039 MAIL command, and the trusted user did not use -F to supply a sender
3040 name, $originator_name was incorrectly used when constructing a From:
3041 header. Furthermore, $originator_name was used for submission mode
3042 messages from external hosts without From: headers in a similar way,
3043 which is clearly wrong.
3045 PH/37 Added control=suppress_local_fixups.
3047 PH/38 When log_selector = +received_sender was set, and the addition of the
3048 sender made the log line's construction buffer exactly full, or one byte
3049 less than full, an overflow happened when the terminating "\n" was
3052 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3053 when the result of a list match is failure because a DNS lookup failed.
3055 PH/40 RM_COMMAND is now used in the building process.
3057 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3058 the "build-* directories that it finds.
3060 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3061 address in a domain literal was a prefix of an interface address.
3063 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3064 when verifying a sender address, unless rewrite_headers is false.
3066 PH/44 Wrote a long comment about why errors_to addresses are verified as
3067 recipients, not senders.
3069 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3070 the ratelimit ACL was added.
3072 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3074 PH/46 Added extra information about PostgreSQL errors to the error string.
3076 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3077 avoiding the need to install special zones in a real server. This is
3078 backwards compatible; if it can't find the fake resolver, it drops back.
3079 Thus, both old and new test suites can be run.
3081 TF/02 Added util/ratelimit.pl
3083 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3084 clock is set back in time.
3086 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3087 Candler <B.Candler@pobox.com>.
3089 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3090 OK for addresses that are the result of redirections.
3092 PH/48 A number of further additions for the benefit of the new test suite,
3093 including a fake gethostbyname() that interfaces to the fake DNS resolver
3096 TF/06 The fix for widen_domains has also been applied to qualify_single and
3097 search_parents which are the other dnslookup options that can cause
3100 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3103 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3105 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3106 service" if it encountered an MX record with an empty target hostname.
3107 The message is now "an MX or SRV record indicated no SMTP service".
3109 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3110 if the DNS of the sending site is misconfigured. This is quite a
3111 common situation. This change restores the behaviour of treating a
3112 helo verification defer as a failure.
3114 PH/51 If self=fail was set on a router, the bounce message did not include the
3115 actual error message.
3121 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3123 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3124 command itself ended in error, the underlying message about the transport
3125 filter timeout was being overwritten with the pipe command error. Now the
3126 underlying error message should be appended to the second error message.
3128 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3130 PH/02 Reduce the amount of output that "make" produces by default. Full output
3131 can still be requested.
3133 PH/03 The warning log line about a condition test deferring for a "warn" verb
3134 was being output only once per connection, rather than after each
3135 occurrence (because it was using the same function as for successful
3136 "warn" verbs). This seems wrong, so I have changed it.
3138 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3139 it should not have, which might have caused a crash in the right
3140 circumstances, but probably never did.
3142 PH/04 Installed a modified version of Tony Finch's patch to make submission
3143 mode fix the return path as well as the Sender: header line, and to
3144 add a /name= option so that you can make the user's friendly name appear
3147 TF/03 Added the control = fakedefer ACL modifier.
3149 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3150 Mark Lowes for thorough testing.
3152 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3154 TK/03 Merged latest SRS patch from Miles Wilton.
3156 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3157 with the definition in sysexits.h (which is #included earlier).
3158 Fortunately, Exim does not actually use EX_OK. The code used to try to
3159 preserve the sysexits.h value, by assumimg that macro definitions were
3160 scanned for macro replacements. I have been disabused of this notion,
3161 so now the code just undefines EX_OK before #including unistd.h.
3163 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3164 in the smtp transport. When a block could not be written in a single
3165 write() function, the timeout was being re-applied to each part-write.
3166 This seems wrong - if the receiver was accepting one byte at a time it
3167 would take for ever. The timeout is now adjusted when this happens. It
3168 doesn't have to be particularly precise.
3170 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3171 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3173 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3174 without a domain if the check to <postmaster@domain> fails.
3176 SC/01 Eximstats: added -xls and the ability to specify output files
3177 (patch written by Frank Heydlauf).
3179 SC/02 Eximstats: use FileHandles for outputing results.
3181 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3183 SC/04 Eximstats: fixed display of large numbers with -nvr option
3185 SC/05 Eximstats: fixed merging of reports with empty tables.
3187 SC/06 Eximstats: added the -include_original_destination flag
3189 SC/07 Eximstats: removed tabs and trailing whitespace.
3191 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3193 TK/06 MBOX spool code: Add real "From " MBOX separator line
3194 so the .eml file is really in mbox format (even though
3195 most programs do not really care). Patch from Alex Miller.
3197 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3198 The latter is generated from $received_to and is only set if the
3199 message has one envelope recipient. SA can use these headers,
3200 obviously out-of-the-box. Patch from Alex Miller.
3202 PH/08 The ${def test on a variable was returning false if the variable's
3203 value was "0", contrary to what the specification has always said!
3204 The result should be true unless the variable is empty.
3206 PH/09 The syntax error of a character other than { following "${if
3207 def:variable_name" (after optional whitespace) was not being diagnosed.
3208 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3209 accidental colon was present, for example, could give incorrect results.
3211 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3212 result is used (not including appendfile, where other changes are about
3215 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3216 This involved changing a lot of size variables from int to off_t. It
3217 should work with maildirs and everything.
3219 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3220 spamd dying while we are connected to it.
3222 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3225 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3226 function for BDB 4.3.
3228 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3230 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3231 the sources for PCRE have been re-arranged and more files are now
3234 PH/15 The code I had for printing potentially long long variables in PH/11
3235 above was not the best (it lost precision). The length of off_t variables
3236 is now inspected at build time, and an appropriate printing format (%ld
3237 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3238 to be "long long int" or "long int". This is needed for the internal
3239 formatting function string_vformat().
3241 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3242 the configuration file to be ":syslog", then the script "guesses" where
3243 the logs files are, rather than using the compiled in default. In our
3244 case the guess is not the same as the compiled default, so the script
3245 suddenly stopped working when I started to use syslog. The patch checks
3246 to see if log_file_path is "". If so, it attempts to read it from exim
3247 with no configuration file to get the compiled in version, before it
3248 falls back to the previous guessing code."
3250 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3251 implementing BATV in an Exim configuration. See NewStuff for the gory
3254 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3255 Makefile that are specific to HP-UX.
3257 PH/18 If the "use_postmaster" option was set for a recipient callout together
3258 with the "random" option, the postmaster address was used as the MAIL
3259 FROM address for the random test, but not for the subsequent recipient
3260 test. It is now used for both.
3262 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3263 patch removes a few documentation additions to RFC 3028, because the
3264 latest draft now contains them. It adds the new en;ascii-case comparator
3265 and a new error check for 8bit text in MIME parts. Comparator and
3266 require names are now matched exactly. I enabled the subaddress
3267 extension, but it is not well tested yet (read: it works for me)."
3269 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3270 rework some of the code of TK/09 above to avoid the hardwired use of
3271 "%lld" and "long long". Replaced the call to snprintf() with a call to
3274 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3275 records point to non-existent hosts", "retry timeout exceeded", and
3276 "retry time not reached for any host after a long failure period".
3278 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3279 experimental DomainKeys support:
3281 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3282 (2) On an error such as an illegally used "control", the wrong name for
3283 the control was given.
3285 These problems did NOT occur unless DomainKeys support was compiled.
3287 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3289 PH/24 Added ${if match_ip condition.
3291 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3292 databases so that it will be absolutely obvious if a crash occurs in the
3293 DB library. This is a regular occurrence (often caused by mis-matched
3296 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3297 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3298 picked up on a user's system that detects such things. There doesn't seem
3299 to be a gcc warning option for this - only an attribute that has to be
3300 put on the function's prototype. It seems that in Fedora Core 4 they have
3301 set this on a number of new functions. No doubt there will be more in due
3304 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3305 specify a transport. However, if an address that was verified by such a
3306 router was the subject of a callout, Exim crashed because it tried to
3307 read the rcpt_include_affixes from the non-existent transport. Now it
3308 just assumes that the setting of that option is false. This bug was
3309 introduced by 4.51/PH/31.
3311 PH/28 Changed -d+all to exclude +memory, because that information is very
3312 rarely of interest, but it makes the output a lot bigger. People tend to
3313 do -d+all out of habit.
3315 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3316 code in os-type was giving problems when libc.so lives in lib64, like on
3319 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3320 aren't the modern standard, and it seems that some systems' include files
3321 don't always have them. Exim was already checking for some of the newer
3322 ones like T_AAAA, and defining it itself. I've added checks for all the
3323 record types that Exim uses.
3325 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3326 not automatically generating a new one, as it is supposed to. This
3327 prevented TLS from working. If the file did exist, but contained invalid
3328 data, a new version was generated, as expected. It was only the case of a
3329 non-existent file that was broken.
3331 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3332 with a change in libdomainkeys > 0.64.
3334 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3335 from DNS. If the selector record carries the flag, it now has
3336 precedence over the domain-wide flag.
3338 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3340 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3341 the use of a port name that isn't defined in /etc/services) occurred, the
3342 message was deferred as in a normal delivery, and thus remained on the
3343 spool, instead of being failed because of the mua_wrapper setting. This
3344 is now fixed, and I tidied up some of the mua_wrapper messages at the
3347 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3348 the messages in a hash of arrays rather than using individual hashes.
3349 This is a bit cleaner and results in dramatic memory savings, albeit
3350 at a slight CPU cost.
3352 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3353 as requested by Marc Sherman.
3355 SC/10 Eximstats: added histograms for user specified patterns as requested
3358 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3360 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3361 fopen() in the content-scanning modules that did not already have it.
3367 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3368 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3370 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3372 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3374 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3375 to test Sieve filters that use "vacation".
3377 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3378 that changes the way the GnuTLS parameters are stored in the cache file.
3379 The new format can be generated externally. For backward compatibility,
3380 if the data in the cache doesn't make sense, Exim assumes it has read an
3381 old-format file, and it generates new data and writes a new file. This
3382 means that you can't go back to an older release without removing the
3385 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3386 work if there are any delivery delays because "one_time" forces the
3387 parent to be marked "delivered", so its unseen clone is never tried
3388 again. For this reason, Exim now forbids the simultaneous setting of
3391 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3392 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3393 entry if you want to know the details. The fix, however, neglected to
3394 consider the case when local delivery batching is involved. The test for
3395 "previously delivered" was not happening when checking to see if an
3396 address could be batched with a previous (undelivered) one; under
3397 certain circumstances this could lead to multiple deliveries to the same
3400 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3401 in its include files, and this causes problems building Exim.
3403 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3404 header_syntax) but Exim was just ignoring anything given after a slash.
3405 In particular, this caused confusion with an attempt to use "verify =
3406 reverse_host_lookup/defer_ok". An error is now given when options are
3407 supplied for verify items that do not have them. (Maybe reverse_host_
3408 lookup should have a defer_ok option, but that's a different point.)
3410 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3411 defined by RFC 821) to 2048, because there were problems with some AUTH
3412 commands, and RFC 1869 says the size should be increased for extended
3413 SMTP commands that take arguments.
3415 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3418 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3419 "unknown" error; now it says that the functionality isn't in the binary.
3421 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3422 an address' error message when a string expansion fails (syntax or
3423 whatever). Otherwise the password may appear in the log. Following change
3424 PH/42 below, there is no longer a chance of it appearing in a bounce
3427 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3429 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3430 of its addresses was ever tried. (Bugzilla bug #2.)
3432 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3433 the result incorrectly in the debug output. (It correctly added a newline
3434 to what was transported.)
3436 TF/01 Added $received_time.
3438 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3439 commented out examples of how to interface to a virus scanner and to
3440 SpamAssassin. Also added commented examples of av_scanner and
3441 spamd_address settings.
3443 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3444 and controls are allowed in which ACLs. There were a couple of minor
3445 errors. Some of the entries in the conditions table (which is a table of
3446 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3447 negation of where the condition IS allowed.
3449 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3451 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3452 header file does not have a version number, so I've had to invent a new
3453 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3454 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3455 radiusclient), but it was contributed by a Radius user.
3457 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3458 files or directories, for OpenSSL.
3460 PH/19 When an Exim process that is running as root has to create an Exim log
3461 file, it does so in a subprocess that runs as exim:exim so as to get the
3462 ownership right at creation (otherwise, other Exim processes might see
3463 the file with the wrong ownership). There was no test for failure of this
3464 fork() call, which would lead to the process getting stuck as it waited
3465 for a non-existent subprocess. Forks do occasionally fail when resources
3466 run out. I reviewed all the other calls to fork(); they all seem to check
3469 PH/20 When checking for unexpected SMTP input at connect time (before writing
3470 the banner), Exim was not dealing correctly with a non-positive return
3471 from the read() function. If the client had disconnected by this time,
3472 the result was a log entry for a synchronization error with an empty
3473 string after "input=" when read() returned zero. If read() returned -1
3474 (an event I could not check), uninitialized data bytes were printed.
3475 There were reports of junk text (parts of files, etc) appearing after
3478 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3480 PH/22 Added support for macro redefinition, and (re)definition in between
3481 driver and ACL definitions.
3483 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3484 forgetting to use the resulting value; it was using the unexpanded value.
3486 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3487 hadn't been configured. The fix is from Juergen Kreileder, who
3488 understands it better than I do:
3490 "Here's what I see happening with three configured cyrus_sasl
3491 authenticators configured (plain, login, cram-md5):
3493 On startup auth_cyrus_sasl_init() gets called for each of these.
3494 This means three calls to sasl_listmech() without a specified mech_list.
3495 => SASL tests which mechs of all available mechs actually work
3496 => three warnings about OTP not working
3497 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3499 With the patch, sasl_listmech() also gets called three times. But now
3500 SASL's mech_list option is set to the server_mech specified in the the
3501 authenticator. Or in other words, the answer from sasl_listmech()
3502 gets limited to just the mech you're testing for (which is different
3504 => the return list contains just 'plain' or 'login', 'cram-md5' or
3505 nothing depending on the value of ob->server_mech.
3507 I've just tested the patch: Authentication still works fine,
3508 unavailable mechs specified in the exim configuration are still
3509 caught, and the auth.log warnings about OTP are gone."
3511 PH/25 When debugging is enabled, the contents of the command line are added
3512 to the debugging output, even when log_selector=+arguments is not
3515 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3516 answer is "GNU", and only if the return is "GNU/something" is the answer
3519 PH/27 $acl_verify_message is now set immediately after the failure of a
3520 verification in an ACL, and so is available in subsequent modifiers. In
3521 particular, the message can be preserved by coding like this:
3523 warn !verify = sender
3524 set acl_m0 = $acl_verify_message
3526 Previously, $acl_verify_message was set only while expanding "message"
3527 and "log_message" when a very denied access.
3529 PH/28 Modified OS/os.c-Linux with
3531 -#ifndef OS_LOAD_AVERAGE
3532 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3534 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3535 nomenclature these days.)
3537 PH/29 Installed patch from the Sieve maintainer that adds the options
3538 sieve_useraddress and sieve_subaddress to the redirect router.
3540 PH/30 In these circumstances:
3541 . Two addresses routed to the same list of hosts;
3542 . First host does not offer TLS;
3543 . First host accepts first address;
3544 . First host gives temporary error to second address;
3545 . Second host offers TLS and a TLS session is established;
3546 . Second host accepts second address.
3547 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3548 and peerdn, if requested) that were in fact used only for the second
3551 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3552 attention to any local part prefix or suffix that was matched by the
3553 router that accepted the address. It now behaves in the same way as it
3554 does for delivery: the affixes are removed from the local part unless
3555 rcpt_include_affixes is set on the transport.
3557 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3558 timeout during the DATA phase of an incoming message.
3560 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3561 applied a patch sanctioned by the Sieve maintainer.
3563 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3564 the uid or gid is negative. A case of a negative gid caused this to be
3565 noticed. The fix allows for either to be negative.
3567 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3568 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3571 PH/36 The change PH/12 above was broken. Fixed it.
3573 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3574 the grounds that routing the same address twice would always produce the
3575 same answer. This might have been true once, but it is certainly no
3576 longer true now. Routing a child address may depend on the previous
3577 routing that produced that child. Some complicated redirection strategies
3578 went wrong when messages had multiple recipients, and made Exim's
3579 behaviour dependent on the order in which the addresses were given.
3581 I have moved the duplicate checking until after the routing is complete.
3582 Exim scans the addresses that are assigned to local and remote
3583 transports, and removes any duplicates. This means that more work will be
3584 done, as duplicates will always all be routed, but duplicates are
3585 presumably rare, so I don't expect this is of any significance.
3587 For deliveries to pipes, files, and autoreplies, the duplicate checking
3588 still happens during the routing process, since they are not going to be
3591 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3592 It corrects a timeout issue with spamd. This is Ian's comment: "The
3593 background is that sometimes spamd either never reads data from a
3594 connection it has accepted, or it never writes response data. The exiscan
3595 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3596 blindly assumes that writes won't block so it may never time out."
3598 PH/39 Allow G after quota size as well as K and M.
3600 PH/40 The value set for $authenticated_id in an authenticator may not contain
3601 binary zeroes or newlines because the value is written to log lines and
3602 to spool files. There was no check on this. Now the value is run through
3603 the string_printing() function so that such characters are converted to
3604 printable escape sequences.
3606 PH/41 $message_linecount is a new variable that contains the total number of
3607 lines in the message. Compare $body_linecount, which is the count for the
3610 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3611 bounce and delay warning messages, except in certain special cases, which
3614 (a) An SMTP error message from a remote host;
3615 (b) A message specified in a :fail: redirection;
3616 (c) A message specified in a "fail" command in a system filter;
3617 (d) A message specified in a FAIL return from the queryprogram router;
3618 (e) A message specified by the cannot_route_message router option.
3620 In these cases only, Exim does include the error details in bounce and
3621 warning messages. There are also a few cases where bland messages such
3622 as "unrouteable address" or "local delivery error" are given.
3624 PH/43 $value is now also set for the "else" part of a ${run expansion.
3626 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3627 being worked on, but at least Exim now implements the latest version to
3630 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3631 process to complete was treated as a delivery failure, a timeout while
3632 writing the message to the pipe was logged, but erroneously treated as a
3633 successful delivery. Such timeouts include transport filter timeouts. For
3634 consistency with the overall process timeout, these timeouts are now
3635 treated as errors, giving rise to delivery failures by default. However,
3636 there is now a new Boolean option for the pipe transport called
3637 timeout_defer, which, if set TRUE, converts the failures into defers for
3638 both kinds of timeout. A transport filter timeout is now identified in
3641 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3642 systems where "make" and "gmake" are different, calling "gmake" at top
3643 level broke things. I've arranged for the value of $(MAKE) to be passed
3644 from the Makefile to this script so that it can call the same version of
3648 A note about Exim versions 4.44 and 4.50
3649 ----------------------------------------
3651 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3652 changes of various kinds. As a consequence, a big documentation update was
3653 needed. This delayed the release for rather longer than seemed good, especially
3654 in the light of a couple of (minor) security issues. Therefore, the changes
3655 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3656 release. So 4.44 and 4.50 are in effect two different branches that both start
3659 I have left the 4.50 change log unchanged; it contains all the changes since
3660 4.43. The change log for 4.44 is below; many of its items are identical to
3661 those for 4.50. This seems to be the most sensible way to preserve the
3662 historical information.
3668 1. Minor wording change to the doc/README.SIEVE file.
3670 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3671 computation of the current number of files was incorrect.
3673 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3674 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3677 4. Give more explanation in the error message when the command for a transport
3678 filter fails to execute.
3680 5. There are several places where Exim runs a non-Exim command in a
3681 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3682 was being done only for the command run by the queryprogram router. It is
3683 now done for all such subprocesses. The other cases are: ${run, transport
3684 filters, and the commands run by the lmtp and pipe transports.
3686 6. Added CONFIGURE_GROUP build-time option.
3688 7. Some older OS have a limit of 256 on the maximum number of file
3689 descriptors. Exim was using setrlimit() to set 1000 as a large value
3690 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3691 systems. I've change it so that if it can't get 1000, it tries for 256.
3693 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3694 was an oversight, and furthermore, ever since the addition of extra
3695 controls (e.g. 4.43/32), the checks on when to allow different forms of
3696 "control" were broken. There should now be diagnostics for all cases when a
3697 control that does not make sense is encountered.
3699 9. Added the /retain_sender option to "control=submission".
3701 10. $recipients is now available in the predata ACL (oversight).
3703 11. Tidy the search cache before the fork to do a delivery from a message
3704 received from the command line. Otherwise the child will trigger a lookup
3705 failure and thereby defer the delivery if it tries to use (for example) a
3706 cached ldap connection that the parent has called unbind on.
3708 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3709 of $address_data from the recipient verification was clobbered by the
3710 sender verification.
3712 13. The value of address_data from a sender verification is now available in
3713 $sender_address_data in subsequent conditions in the ACL statement.
3715 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3717 15. Added a new option "connect=<time>" to callout options, to set a different
3720 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3721 was its contents. (It was OK if the option was not defined at all.)
3723 17. A "Completed" log line is now written for messages that are removed from
3724 the spool by the -Mrm option.
3726 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3727 information about exactly what failed.
3729 19. Added -dd to debug only the daemon process.
3731 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3732 handles timeouts, both on the server side and network timeouts. Renamed the
3733 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3735 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3736 It is now set to "smtps".
3738 22. $host_address is now set to the target address during the checking of
3739 ignore_target_hosts.
3741 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3742 being passed; this would have caused $sender_host_name to have been used if
3743 matching the list had actually called for a host name (not very likely,
3744 since this list is usually IP addresses). A host name is now passed as
3747 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3748 code that specifies a non-restarting handler (typically sigaction() in
3749 modern systems) in an attempt to fix a rare and obscure crash bug.
3751 25. Narrowed the window for a race in the daemon that could cause it to ignore
3752 SIGCHLD signals. This is not a major problem, because they are used only to
3753 wake it up if nothing else does.
3755 26. A malformed maildirsize file could cause Exim to calculate negative values
3756 for the mailbox size or file count. Odd effects could occur as a result.
3757 The maildirsize information is now recalculated if the size or filecount
3760 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3761 support for a long time. Removed HAVE_SYS_VFS_H.
3763 28. Installed the latest version of exipick from John Jetmore.
3765 29. In an address list, if the pattern was not a regular expression, an empty
3766 subject address (from a bounce message) matched only if the pattern was an
3767 empty string. Non-empty patterns were not even tested. This was the wrong
3768 because it is perfectly reasonable to use an empty address as part of a
3769 database query. An empty address is now tested by patterns that are
3770 lookups. However, all the other forms of pattern expect the subject to
3771 contain a local part and a domain, and therefore, for them, an empty
3772 address still always fails if the pattern is not itself empty.
3774 30. Exim went into a mad DNS loop when attempting to do a callout where the
3775 host was specified on an smtp transport, and looking it up yielded more
3776 than one IP address.
3778 31. Re-factored the code for checking spool and log partition space into a
3779 function that finds that data and another that does the check. The former
3780 is then used to implement four new variables: $spool_space, $log_space,
3781 $spool_inodes, and $log_inodes.
3783 32. The RFC2047 encoding function was originally intended for short strings
3784 such as real names; it was not keeping to the 75-character limit for
3785 encoded words that the RFC imposes. It now respects the limit, and
3786 generates multiple encoded words if necessary. To be on the safe side, I
3787 have increased the buffer size for the ${rfc2047: expansion operator from
3790 33. It is now permitted to omit both strings after an "if" condition; if the
3791 condition is true, the result is "true". As before, when the second string
3792 is omitted, a false condition yields an empty string. This makes it less
3793 cumbersome to write custom ACL and router conditions.
3795 34. Failure to deliver a bounce message always caused it to be frozen, even if
3796 there was an errors_to setting on the router. The errors_to setting is now
3799 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3800 canonical form (fully expanded) before being placed in
3801 $sender_host_address.
3803 36. The table in the code that translates DNS record types into text (T_A to
3804 "A" for instance) was missing entries for NS and CNAME. It is just possible
3805 that this could have caused confusion if both these types were looked up
3806 for the same domain, because the text type is used as part of Exim's
3807 per-process caching. But the chance of anyone hitting this buglet seems
3810 37. The dnsdb lookup has been extended in a number of ways.
3812 (1) There is a new type, "zns", which walks up the domain tree until it
3813 finds some nameserver records. It should be used with care.
3815 (2) There is a new type, "mxh", which is like "mx" except that it returns
3816 just the host names, not the priorities.
3818 (3) It is now possible to give a list of domains (or IP addresses) to be
3819 looked up. The behaviour when one of the lookups defers can be
3820 controlled by a keyword.
3822 (4) It is now possible to specify the separator character for use when
3823 multiple records are returned.
3825 38. The dnslists ACL condition has been extended: it is now possible to supply
3826 a list of IP addresses and/or domains to be looked up in a particular DNS
3829 39. Added log_selector=+queue_time_overall.
3831 40. When running the queue in the test harness, wait just a tad after forking a
3832 delivery process, to get repeatability of debugging output.
3834 41. Include certificate and key file names in error message when GnuTLS fails
3835 to set them up, because the GnuTLS error message doesn't include the name
3836 of the failing file when there is a problem reading it.
3838 42. Allow both -bf and -bF in the same test run.
3840 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3842 44. The "Exiscan patch" is now merged into the mainline Exim source.
3844 45. Sometimes the final signoff response after QUIT could fail to get
3845 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3846 < 0 before doing a fflush(). This bug looks as though it goes back to the
3847 introduction of TLS in release 3.20, but "sometimes" must have been rare
3848 because the tests only now provoked it.
3850 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3851 (this can affect the format of dates).
3853 47. exim_tidydb, when checking for the continued existence of a message for
3854 which it has found a message-specific retry record, was not finding
3855 messages that were in split spool directories. Consequently, it was
3856 deleting retry records that should have stayed in existence.
3858 48. Steve fixed some bugs in eximstats.
3860 49. The SPA authentication driver was not abandoning authentication and moving
3861 on to the next authenticator when an expansion was forced to fail,
3862 contradicting the general specification for all authenticators. Instead it
3863 was generating a temporary error. It now behaves as specified.
3865 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3866 (the order specifies the preference for clients). The order is now AES256,
3867 AES128, 3DES, ARCFOUR128.
3869 51. Small patch to Sieve code - explicitly set From: when generating an
3872 52. Exim crashed if a remote delivery caused a very long error message to be
3873 recorded - for instance if somebody sent an entire SpamAssassin report back
3874 as a large number of 550 error lines. This bug was coincidentally fixed by
3875 increasing the size of one of Exim's internal buffers (big_buffer) that
3876 happened as part of the Exiscan merge. However, to be on the safe side, I
3877 have made the code more robust (and fixed the comments that describe what
3880 53. Now that there can be additional text after "Completed" in log lines (if
3881 the queue_time_overall log selector is set), a one-byte patch to exigrep
3882 was needed to allow it to recognize "Completed" as not the last thing in
3885 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3886 patch that reportedly fixes this has been added. I am not expert enough to
3887 create a test for it. This is what the patch creator wrote:
3889 "I found a little strange behaviour of ldap code when working with
3890 Windows 2003 AD Domain, where users was placed in more than one
3891 Organization Units. When I tried to give exim partial DN, the exit code
3892 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3893 But simultaneously result of request was absolutely normal ldap result,
3894 so I produce this patch..."
3896 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3897 REFERENCE, so I have modified the code to exclude the patch when that macro
3900 55. Some experimental protocols are using DNS PTR records for new purposes. The
3901 keys for these records are domain names, not reversed IP addresses. The
3902 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3903 leaves it alone. Component reversal etc. now happens only for IP addresses.
3906 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3908 57. Double the size of the debug message buffer (to 2048) so that more of very
3909 long debug lines gets shown.
3911 58. The exicyclog utility now does better if the number of log files to keep
3912 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3914 59. Two changes related to the smtp_active_hostname option:
3916 (1) $smtp_active_hostname is now available as a variable.
3917 (2) The default for smtp_banner uses $smtp_active_hostname instead
3918 of $primary_hostname.
3920 60. The host_aton() function is supposed to be passed a string that is known
3921 to be a valid IP address. However, in the case of IPv6 addresses, it was
3922 not checking this. This is a hostage to fortune. Exim now panics and dies
3923 if the condition is not met. A case was found where this could be provoked
3924 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3925 components; fortuitously, this particular loophole had already been fixed
3926 by change 4.50/55 above.
3928 If there are any other similar loopholes, the new check in host_aton()
3929 itself should stop them being exploited. The report I received stated that
3930 data on the command line could provoke the exploit when Exim was running as
3931 exim, but did not say which command line option was involved. All I could
3932 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3933 running as the user.
3936 61. There was a buffer overflow vulnerability in the SPA authentication code
3937 (which came originally from the Samba project). I have added a test to the
3938 spa_base64_to_bits() function which I hope fixes it.
3941 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3942 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3944 63. The daemon start-up calls getloadavg() while still root for those OS that
3945 need the first call to be done as root, but it missed one case: when
3946 deliver_queue_load_max is set with deliver_drop_privilege. This is
3947 necessary for the benefit of the queue runner, because there is no re-exec
3948 when deliver_drop_privilege is set.
3950 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3951 This has been fixed.
3953 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3954 was in use, was not putting the data itself into the right store pool;
3955 consequently, it could be overwritten for a subsequent message in the same
3956 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3959 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3961 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3962 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3963 the function were treating the return as a boolean value, which happened to
3964 work because 0=false and not-0=true, but is not correct code.
3966 68. The host_aton() function was not handling scoped IPv6 addresses (those
3967 with, for example, "%eth0" on the end) correctly.
3969 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3970 negated items (that is, ~something) in unsigned ints. Some compilers
3971 apparently mutter when there is no cast.
3973 70. If an address verification called from an ACL failed, and did not produce a
3974 user-specific message (i.e. there was only a "system" message), nothing was
3975 put in $acl_verify_message. In this situation, it now puts the system
3978 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3979 session; change 4.31/43 added the unwanted input to the log line - except
3980 that it did not do this in the start of session case. It now does.
3982 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3983 This is wrong and can cause the other end to generate a synchronization
3984 error if it is another Exim or anything else that does the synchronization
3985 check. A QUIT command is no longer sent after a timeout.
3987 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3988 during host lookups.
3990 74. The defer_ok option of callout verification was not working if it was used
3991 when verifying addresses in header lines, that is, for this case:
3993 verify = header_sender/callout=defer_ok
3995 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3996 those file descriptors could be used for SMTP connections. If anything
3997 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3998 could be sent to the SMTP client, causing chaos. The daemon now opens
3999 stdin, stdout, and stderr to /dev/null when it puts itself into the
4002 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4003 log by default. The user can override this with suitable Perl magic.
4005 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4006 the log message when discard triggers, was not working for the DATA ACL or
4007 for the non-SMTP ACL.
4009 78. Error message wording change in sieve.c.
4011 79. If smtp_accept_max_per_host was set, the number of connections could be
4012 restricted to fewer than expected, because the daemon was trying to set up
4013 a new connection before checking whether the processes handling previous
4014 connections had finished. The check for completed processes is now done
4015 earlier. On busy systems, this bug wouldn't be noticed because something
4016 else would have woken the daemon, and it would have reaped the completed
4019 80. If a message was submitted locally by a user whose login name contained one
4020 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4021 It caused a spool format error. I have fixed the spool reading code. A
4022 related problem was that the "from" clause in the Received: line became
4023 illegal because of the space(s). It is now covered by ${quote_local_part.
4025 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4028 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4030 83. Patch from Sieve maintainer to fix unterminated string problem in
4031 "vacation" handling.
4033 84. Some minor changes to the Linux configuration files to help with other
4034 OS variants using glibc.
4036 85. One more patch for Sieve to update vacation handling to latest spec.
4039 ----------------------------------------------------
4040 See the note above about the 4.44 and 4.50 releases.
4041 ----------------------------------------------------
4047 1. Change 4.43/35 introduced a bug that caused file counts to be
4048 incorrectly computed when quota_filecount was set in an appendfile
4051 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4052 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4055 3. Give more explanation in the error message when the command for a transport
4056 filter fails to execute.
4058 4. There are several places where Exim runs a non-Exim command in a
4059 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4060 was being done only for the command run by the queryprogram router. It is
4061 now done for all such subprocesses. The other cases are: ${run, transport
4062 filters, and the commands run by the lmtp and pipe transports.
4064 5. Some older OS have a limit of 256 on the maximum number of file
4065 descriptors. Exim was using setrlimit() to set 1000 as a large value
4066 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4067 systems. I've change it so that if it can't get 1000, it tries for 256.
4069 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4070 was an oversight, and furthermore, ever since the addition of extra
4071 controls (e.g. 4.43/32), the checks on when to allow different forms of
4072 "control" were broken. There should now be diagnostics for all cases when a
4073 control that does not make sense is encountered.
4075 7. $recipients is now available in the predata ACL (oversight).
4077 8. Tidy the search cache before the fork to do a delivery from a message
4078 received from the command line. Otherwise the child will trigger a lookup
4079 failure and thereby defer the delivery if it tries to use (for example) a
4080 cached ldap connection that the parent has called unbind on.
4082 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4083 of $address_data from the recipient verification was clobbered by the
4084 sender verification.
4086 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4087 was its contents. (It was OK if the option was not defined at all.)
4089 11. A "Completed" log line is now written for messages that are removed from
4090 the spool by the -Mrm option.
4092 12. $host_address is now set to the target address during the checking of
4093 ignore_target_hosts.
4095 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4096 being passed; this would have caused $sender_host_name to have been used if
4097 matching the list had actually called for a host name (not very likely,
4098 since this list is usually IP addresses). A host name is now passed as
4101 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4102 code that specifies a non-restarting handler (typically sigaction() in
4103 modern systems) in an attempt to fix a rare and obscure crash bug.
4105 15. Narrowed the window for a race in the daemon that could cause it to ignore
4106 SIGCHLD signals. This is not a major problem, because they are used only to
4107 wake it up if nothing else does.
4109 16. A malformed maildirsize file could cause Exim to calculate negative values
4110 for the mailbox size or file count. Odd effects could occur as a result.
4111 The maildirsize information is now recalculated if the size or filecount
4114 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4115 support for a long time. Removed HAVE_SYS_VFS_H.
4117 18. Updated exipick to current release from John Jetmore.
4119 19. Allow an empty sender to be matched against a lookup in an address list.
4120 Previously the only cases considered were a regular expression, or an
4123 20. Exim went into a mad DNS lookup loop when doing a callout where the
4124 host was specified on the transport, if the DNS lookup yielded more than
4127 21. The RFC2047 encoding function was originally intended for short strings
4128 such as real names; it was not keeping to the 75-character limit for
4129 encoded words that the RFC imposes. It now respects the limit, and
4130 generates multiple encoded words if necessary. To be on the safe side, I
4131 have increased the buffer size for the ${rfc2047: expansion operator from
4134 22. Failure to deliver a bounce message always caused it to be frozen, even if
4135 there was an errors_to setting on the router. The errors_to setting is now
4138 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4139 canonical form (fully expanded) before being placed in
4140 $sender_host_address.
4142 24. Updated eximstats to version 1.33
4144 25. Include certificate and key file names in error message when GnuTLS fails
4145 to set them up, because the GnuTLS error message doesn't include the name
4146 of the failing file when there is a problem reading it.
4148 26. Expand error message when OpenSSL has problems setting up cert/key files.
4151 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4152 (this can affect the format of dates).
4154 28. exim_tidydb, when checking for the continued existence of a message for
4155 which it has found a message-specific retry record, was not finding
4156 messages that were in split spool directories. Consequently, it was
4157 deleting retry records that should have stayed in existence.
4159 29. eximstats updated to version 1.35
4160 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4161 1.35 - bugfix such that pie charts by volume are generated correctly
4163 30. The SPA authentication driver was not abandoning authentication and moving
4164 on to the next authenticator when an expansion was forced to fail,
4165 contradicting the general specification for all authenticators. Instead it
4166 was generating a temporary error. It now behaves as specified.
4168 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4169 (the order specifies the preference for clients). The order is now AES256,
4170 AES128, 3DES, ARCFOUR128.
4172 31. Small patch to Sieve code - explicitly set From: when generating an
4175 32. Exim crashed if a remote delivery caused a very long error message to be
4176 recorded - for instance if somebody sent an entire SpamAssassin report back
4177 as a large number of 550 error lines. This bug was coincidentally fixed by
4178 increasing the size of one of Exim's internal buffers (big_buffer) that
4179 happened as part of the Exiscan merge. However, to be on the safe side, I
4180 have made the code more robust (and fixed the comments that describe what
4183 33. Some experimental protocols are using DNS PTR records for new purposes. The
4184 keys for these records are domain names, not reversed IP addresses. The
4185 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4186 leaves it alone. Component reversal etc. now happens only for IP addresses.
4189 34. The host_aton() function is supposed to be passed a string that is known
4190 to be a valid IP address. However, in the case of IPv6 addresses, it was
4191 not checking this. This is a hostage to fortune. Exim now panics and dies
4192 if the condition is not met. A case was found where this could be provoked
4193 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4194 components; fortuitously, this particular loophole had already been fixed
4195 by change 4.50/55 or 4.44/33 above.
4197 If there are any other similar loopholes, the new check in host_aton()
4198 itself should stop them being exploited. The report I received stated that
4199 data on the command line could provoke the exploit when Exim was running as
4200 exim, but did not say which command line option was involved. All I could
4201 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4202 running as the user.
4205 35. There was a buffer overflow vulnerability in the SPA authentication code
4206 (which came originally from the Samba project). I have added a test to the
4207 spa_base64_to_bits() function which I hope fixes it.
4210 36. The daemon start-up calls getloadavg() while still root for those OS that
4211 need the first call to be done as root, but it missed one case: when
4212 deliver_queue_load_max is set with deliver_drop_privilege. This is
4213 necessary for the benefit of the queue runner, because there is no re-exec
4214 when deliver_drop_privilege is set.
4216 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4217 was in use, was not putting the data itself into the right store pool;
4218 consequently, it could be overwritten for a subsequent message in the same
4219 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4222 38. Sometimes the final signoff response after QUIT could fail to get
4223 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4224 < 0 before doing a fflush(). This bug looks as though it goes back to the
4225 introduction of TLS in release 3.20, but "sometimes" must have been rare
4226 because the tests only now provoked it.
4232 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4233 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4234 Now it returns an integer. A number of calls were still expecting a T/F
4235 return. Fortuitously, in all cases, the tests worked in OK situations,
4236 which is the norm. However, things would have gone wrong on any write
4237 failures on the smtp file descriptor. This function is used when sending
4238 messages over SMTP and also when doing verify callouts.
4240 2. When Exim is called to do synchronous delivery of a locally submitted
4241 message (the -odf or -odi options), it no longer closes stderr before doing
4244 3. Implemented the mua_wrapper option.
4246 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4248 5. Implemented the functions header_remove(), header_testname(),
4249 header_add_at_position(), and receive_remove_recipient(), and exported them
4252 6. If an ACL "warn" statement specified the addition of headers, Exim already
4253 inserted X-ACL-Warn: at the start if there was no header name. However, it
4254 was not making this test for the second and subsequent header lines if
4255 there were newlines in the string. This meant that an invalid header could
4256 be inserted if Exim was badly configured.
4258 7. Allow an ACL "warn" statement to add header lines at the start or after all
4259 the Received: headers, as well as at the end.
4261 8. Added the rcpt_4xx retry error code.
4263 9. Added postmaster_mailfrom=xxx to callout verification option.
4265 10. Added mailfrom=xxxx to the callout verification option, for verify=
4268 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4269 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4271 12. Inserted some casts to stop certain compilers warning when using pointer
4272 differences as field lengths or precisions in printf-type calls (mostly
4273 affecting debugging statements).
4275 13. Added optional readline() support for -be (dynamically loaded).
4277 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4278 same clock tick as a message's arrival, so that its received time was the
4279 same as the "first fail" time on the retry record, and that message
4280 remained on the queue past the ultimate address timeout, every queue runner
4281 would try a delivery (because it was past the ultimate address timeout) but
4282 after another failure, the ultimate address timeout, which should have then
4283 bounced the address, did not kick in. This was a "< instead of <=" error;
4284 in most cases the first failure would have been in the next clock tick
4285 after the received time, and all would be well.
4287 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4288 being recognized when the domain list was tested by the match_domain
4289 condition in an expansion string.
4291 16. Added the ${str2b64: operator.
4293 17. Exim was always calling setrlimit() to set a large limit for the number of
4294 processes, without checking whether the existing limit was already
4295 adequate. (It did check for the limit on file descriptors.) Furthermore,
4296 errors from getrlimit() and setrlimit() were being ignored. Now they are
4297 logged to the main and panic logs, but Exim does carry on, to try to do its
4298 job under whatever limits there are.
4300 18. Imported PCRE 5.0.
4302 19. Trivial typo in log message " temporarily refused connection" (the leading
4305 20. If the log selector return_path_on_delivery was set and an address was
4306 redirected to /dev/null, the delivery process crashed because it assumed
4307 that a return path would always be set for a "successful" delivery. In this
4308 case, the whole delivery is bypassed as an optimization, and therefore no
4311 21. Internal re-arrangement: the function for sending a challenge and reading
4312 a response while authentication was assuming a zero-terminated challenge
4313 string. It's now changed to take a pointer and a length, to allow for
4314 binary data in such strings.
4316 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4318 23. Exim was not respecting finduser_retries when seeking the login of the
4319 uid under which it was called; it was always trying 10 times. (The default
4320 setting of finduser_retries is zero.) Also, it was sleeping after the final
4321 failure, which is pointless.
4323 24. Implemented tls_on_connect_ports.
4325 25. Implemented acl_smtp_predata.
4327 26. If the domain in control=submission is set empty, Exim assumes that the
4328 authenticated id is a complete email address when it generates From: or
4329 Sender: header lines.
4331 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4332 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4333 chown and chgrp in /bin and hostname in /usr/bin.
4335 28. Exim was keeping the "process log" file open after each use, just as it
4336 does for the main log. This opens the possibility of it remaining open for
4337 long periods when the USR1 signal hits a daemon. Occasional processlog
4338 errors were reported, that could have been caused by this. Anyway, it seems
4339 much more sensible not to leave this file open at all, so that is what now
4342 29. The long-running daemon process does not normally write to the log once it
4343 has entered its main loop, and it closes the log before doing so. This is
4344 so that log files can straightforwardly be renamed and moved. However,
4345 there are a couple of unusual error situations where the daemon does write
4346 log entries, and I had neglected to close the log afterwards.
4348 30. The text of an SMTP error response that was received during a remote
4349 delivery was being truncated at 512 bytes. This is too short for some of
4350 the long messages that one sometimes sees. I've increased the limit to
4353 31. It is now possible to make retry rules that apply only when a message has a
4354 specific sender, in particular, an empty sender.
4356 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4357 it possible to be selective about when SMTP synchronization is enforced.
4359 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4361 32. Implemented hosts_connection_nolog.
4363 33. Added an ACL for QUIT.
4365 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4368 35. Added mailbox_size and mailbox_filecount to appendfile.
4370 36. Added control = no_multiline_responses to ACLs.
4372 37. There was a bug in the logic of the code that waits for the clock to tick
4373 in the case where the clock went backwards by a substantial amount such
4374 that the microsecond fraction of "now" was more than the microsecond
4375 fraction of "then" (but the whole seconds number was less).
4377 38. Added support for the libradius Radius client library this is found on
4378 FreeBSD (previously only the radiusclient library was supported).
4384 1. When certain lookups returned multiple values in the form name=value, the
4385 quoting of the values was not always being done properly. Specifically:
4386 (a) If the value started with a double quote, but contained no whitespace,
4388 (b) If the value contained whitespace other than a space character (i.e.
4389 tabs or newlines or carriage returns) it was not quoted.
4390 This fix has been applied to the mysql and pgsql lookups by writing a
4391 separate quoting function and calling it from the lookup code. The fix
4392 should probably also be applied to nisplus, ibase and oracle lookups, but
4393 since I cannot test any of those, I have not disturbed their existing code.
4395 2. A hit in the callout cache for a specific address caused a log line with no
4396 reason for rejecting RCPT. Now it says "Previous (cached) callout
4397 verification failure".
4399 3. There was an off-by-one bug in the queryprogram router. An over-long
4400 return line was truncated at 256 instead of 255 characters, thereby
4401 overflowing its buffer with the terminating zero. As well as fixing this, I
4402 have increased the buffer size to 1024 (and made a note to document this).
4404 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4405 when Exim is waiting for an SMTP response from a remote server, Exim
4406 restarts its select() call on the socket, thereby resetting its timeout.
4407 This is not a problem when such interrupts are rare. Somebody set up a cron
4408 job to run exiwhat every 2 minutes, which is less than the normal select()
4409 timeout (5 or 10 minutes). This meant that the select() timeout never
4410 kicked in because it was always reset. I have fixed this by comparing the
4411 time when an interrupt arrives with the time at the start of the first call
4412 to select(). If more time than the timeout has elapsed, the interrupt is
4413 treated as a timeout.
4415 5. Some internal re-factoring in preparation for the addition of Sieve
4416 extensions (by MH). In particular, the "personal" test is moved to a
4417 separate function, and given an option for scanning Cc: and Bcc: (which is
4418 not set for Exim filters).
4420 6. When Exim created an email address using the login of the caller as the
4421 local part (e.g. when creating a From: or Sender: header line), it was not
4422 quoting the local part when it contained special characters such as @.
4424 7. Installed new OpenBSD configuration files.
4426 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4427 try to make them clearer.
4429 9. Callout options, other than the timeout value, were being ignored when
4430 verifying sender addresses in header lines. For example, when using
4432 verify = header_sender/callout=no_cache
4434 the cache was (incorrectly) being used.
4436 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4437 only the Cygwin environment.
4439 11. When return_path_on_delivery was set as a log selector, if different remote
4440 addresses in the same message used different return paths and parallel
4441 remote delivery occurred, the wrong values would sometimes be logged.
4442 (Whenever a remote delivery process finished, the return path value from
4443 the most recently started remote delivery process was logged.)
4445 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4446 lines when AUTH and/or TLS are in use. This is the "received protocol"
4447 field. Exim used to use "asmtp" for authenticated SMTP, without any
4448 indication (in the protocol name) for TLS use. Now it follows the RFC and
4449 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4450 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4451 names appear in log lines as well as in Received: header lines.
4453 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4454 extensions, and comparison tests, and to fix some bugs.
4456 14. Changes to the "personal" filter test:
4458 (1) The test was buggy in that it was just doing the equivalent of
4459 "contains" tests on header lines. For example, if a user's address was
4460 anne@some.where, the "personal" test would incorrectly be true for
4462 To: susanne@some.where
4464 This test is now done by extracting each address from the header in turn,
4465 and checking the entire address. Other tests that are part of "personal"
4466 are now done using regular expressions (for example, to check local parts
4467 of addresses in From: header lines).
4469 (2) The list of non-personal local parts in From: addresses has been
4470 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4471 taken from the Sieve specification recommendations.
4473 (3) If the message contains any header line starting with "List-" it is
4474 treated as non-personal.
4476 (4) The test for "circular" in the Subject: header line has been removed
4477 because it now seems ill-conceived.
4479 15. Minor typos in src/EDITME comments corrected.
4481 16. Installed latest exipick from John Jetmore.
4483 17. If headers_add on a router specified a text string that was too long for
4484 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4485 of string_sprintf() is now avoided.
4487 18. $message_body_size was not set (it was always zero) when running the DATA
4488 ACL and the local_scan() function.
4490 19. For the "mail" command in an Exim filter, no default was being set for
4491 the once_repeat time, causing a random time value to be used if "once" was
4492 specified. (If the value happened to be <= 0, no repeat happened.) The
4493 default is now 0s, meaning "never repeat". The "vacation" command was OK
4494 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4495 (I found it when inspecting the code).
4497 20. There is now an overall timeout for performing a callout verification. It
4498 defaults to 4 times the callout timeout, which applies to individual SMTP
4499 commands during the callout. The overall timeout applies when there is more
4500 than one host that can be tried. The timeout is checked before trying the
4501 next host. This prevents very long delays if there are a large number of
4502 hosts and all are timing out (e.g. when the network connections are timing
4503 out). The value of the overall timeout can be changed by specifying an
4504 additional sub-option for "callout", called "maxwait". For example:
4506 verify = sender/callout=5s,maxwait=20s
4508 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4509 to the end before writing, but this should make it even safer).
4511 22. Exim was forgetting that it had advertised PIPELINING for the second and
4512 subsequent messages on an SMTP connection. It was also not resetting its
4513 memory on STARTTLS and an internal HELO.
4515 23. When Exim logs an SMTP synchronization error within a session, it now
4516 records whether PIPELINING has been advertised or not.
4518 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4519 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4520 rather than long int.
4522 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4524 26. Added the never_mail option to autoreply.
4530 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4531 crash if the getsockname() call failed; this can happen if a connection is
4532 closed very soon after it is established. The problem was simply in the
4533 order in which certain operations were done, causing Exim to try to write
4534 to the SMTP stream before it had set up the file descriptor. The bug has
4535 been fixed by making things happen in the correct order.
4541 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4542 before the connection was closed, thus losing the rejection response.
4544 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4545 some early Solaris releases, but causes trouble in current releases where
4546 socklen_t is defined.
4548 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4551 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4554 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4555 that releasing the top of it at the end releases what was used for sub-
4556 expansions (unless the block got too big). However, discard this block if
4557 the first thing is a variable or header, so that we can use its block when
4558 it is dynamic (useful for very large $message_headers, for example).
4560 6. Lookups now cache *every* query, not just the most recent. A new, separate
4561 store pool is used for this. It can be recovered when all lookup caches are
4562 flushed. Lookups now release memory at the end of their result strings.
4563 This has involved some general refactoring of the lookup sources.
4565 7. Some code has been added to the store_xxx() functions to reduce the amount
4566 of flapping under certain conditions.
4568 8. log_incoming_interface used to affect only the <= reception log lines. Now
4569 it causes the local interface and port to be added to several more SMTP log
4570 lines, for example "SMTP connection from", and rejection lines.
4572 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4574 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4576 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4577 could be overwritten at the end of the current message (or the start of a
4578 new message if it was set in a HELO ACL). The value is now preserved for
4579 the duration of the SMTP connection.
4581 12. If a transport had a headers_rewrite setting, and a matching header line
4582 contained an unqualified address, that address was qualified, even if it
4583 did not match any rewriting rules. The underlying bug was that the values
4584 of the flags that permit the existence of unqualified sender and recipient
4585 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4586 non-local messages, and by -bnq for local messages) were not being
4587 preserved with the message after it was received.
4589 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4590 "next input=" as part of the text comprising the host identity instead of
4591 the correct text. The code was using the same buffer for two different
4592 strings. However, depending on which order the printing function evaluated
4593 its arguments, the bug did not always show up. Under Linux, for example, my
4594 test suite worked just fine.
4596 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4597 doesn't work with some older versions of Perl. It has been changed to "my",
4598 which in any case is probably the better facility to use.
4600 15. A really picky compiler found some instances of statements for creating
4601 error messages that either had too many or two few arguments for the format
4604 16. The size of the buffer for calls to the DNS resolver has been increased
4605 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4606 for addresses that have a lot of PTR records. This alleviates a problem; it
4607 does not fully solve it.
4609 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4610 buffer now truncates the list and logs the incident, which is the same
4611 action as happens when Exim is looking up a host name and its aliases.
4612 Previously in this situation something unpredictable would happen;
4613 sometimes it was "internal error: store_reset failed".
4615 18. If a server dropped the connection unexpectedly when an Exim client was
4616 using GnuTLS and trying to read a response, the client delivery process
4617 crashed while trying to generate an error log message.
4619 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4620 string, for example:
4622 warn message = H1: something\nH2: something
4624 the text was added as a single header line from Exim's point of view
4625 though it ended up OK in the delivered message. However, searching for the
4626 second and subsequent header lines using $h_h2: did not work. This has been
4627 fixed. Similarly, if a system filter added multiple headers in this way,
4628 the routers could not see them.
4630 20. Expanded the error message when iplsearch is called with an invalid key to
4631 suggest using net-iplsearch in a host list.
4633 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4634 ACLs are no longer actually imposed (and a message to that effect is
4637 22. If a "gecos" field in a passwd entry contained escaped characters, in
4638 particular, if it contained a \" sequence, Exim got it wrong when building
4639 a From: or a Sender: header from that name. A second bug also caused
4640 incorrect handling when an unquoted " was present following a character
4641 that needed quoting.
4643 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4644 was not being matched caselessly.
4646 24. Arranged for all hyphens in the exim.8 source to be escaped with
4649 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4650 properly. Recipient callout cache records were still being keyed to include
4651 the sender, even when use_sender was set false. This led to far more
4652 callouts that were necessary. The sender is no longer included in the key
4653 when use_sender is false.
4655 26. Added "control = submission" modifier to ACLs.
4657 27. Added the ${base62d: operator to decode base 62 numbers.
4659 28. dnsdb lookups can now access SRV records.
4661 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4662 the configuration file.
4664 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4665 (-v) mode. This makes the output for a verbose queue run more intelligible.
4667 31. Added a use_postmaster feature to recipient callouts.
4669 32. Added the $body_zerocount variable, containing the number of binary zero
4670 bytes in the message body.
4672 33. The time of last modification of the "new" subdirectory is now used as the
4673 "mailbox time last read" when there is a quota error for a maildir
4676 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4678 35. Added +ignore_unknown as a special item in host lists.
4680 36. Code for decoding IPv6 addresses in host lists is now included, even if
4681 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4682 address was recognized as an IP address, but was then not correctly decoded
4683 into binary, causing unexpected and incorrect effects when compared with
4690 1. Very minor rewording of debugging text in manualroute to say "list of
4691 hosts" instead of "hostlist".
4693 2. If verify=header_syntax was set, and a header line with an unqualified
4694 address (no domain) and a large number of spaces between the end of the
4695 name and the colon was received, the reception process suffered a buffer
4696 overflow, and (when I tested it) crashed. This was caused by some obsolete
4697 code that should have been removed. The fix is to remove it!
4699 3. When running in the test harness, delay a bit after writing a bounce
4700 message to get a bit more predictability in the log output.
4702 4. Added a call to search_tidyup() just before forking a reception process. In
4703 theory, someone could use a lookup in the expansion of smtp_accept_max_
4704 per_host which, without the tidyup, could leave open a database connection.
4706 5. Added the variables $recipient_data and $sender_data which get set from a
4707 lookup success in an ACL "recipients" or "senders" condition, or a router
4708 "senders" option, similar to $domain_data and $local_part_data.
4710 6. Moved the writing of debug_print from before to after the "senders" test
4713 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4714 problems for message scanning, either using a data ACL, or using
4715 local_scan() because the Received: header was not generated till after they
4716 were called (in order to set the time as the time of reception completion).
4717 I have revised the way this works. The header is now generated after the
4718 body is received, but before the ACL or local_scan() are called. After they
4719 are run, the timestamp in the header is updated.
4725 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4726 before starting a queue runner without re-exec. This happened only when
4727 deliver_drop_privilege was set or when the Exim user was set to root. The
4728 effect of the bug was that timeouts during subsequent deliveries caused
4729 crashes instead of being properly handled. The handler is now left at its
4730 default (and expected) setting.
4732 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4733 message, again when deliver_drop_privilege is set or Exim is run as root.
4734 The bug described in (1) was not present in this case, but the tidying up
4735 of the other signals was missing. I have made the two cases consistent.
4737 3. The ignore_target_hosts setting on a manualroute router was being ignored
4738 for hosts that were looked up using the /MX notation.
4740 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4743 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4744 operated on the sender address. After changing the $sender_address to <>
4745 for the sender address verify, Exim was re-instated it as the original
4746 (before rewriting) address, but remembering that it had rewritten it, so it
4747 wasn't rewriting it again. This bug also had the effect of breaking the
4748 sender address verification caching when the sender address was rewritten.
4750 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4751 This has been changed so that if the ip literal address matches
4752 ignore_target_hosts, the router declines.
4754 7. Added expansion conditions match_domain, match_address, and match_local_
4755 part (NOT match_host).
4757 8. The placeholder for the Received: header didn't have a length field set.
4759 9. Added code to Exim itself and to exim_lock to test for a specific race
4760 condition that could lead to file corruption when using MBX delivery. The
4761 issue is with the lockfile that is created in /tmp. If this file is removed
4762 after a process has opened it but before that process has acquired a lock,
4763 there is the potential for a second process to recreate the file and also
4764 acquire a lock. This could lead to two Exim processes writing to the file
4765 at the same time. The added code performs the same test as UW imapd; it
4766 checks after acquiring the lock that its file descriptor still refers to
4767 the same named file.
4769 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4770 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4773 11. Added the smtp_active_hostname option. If used, this will typically be made
4774 to depend on the incoming interface address. Because $interface_address is
4775 not set up until the daemon has forked a reception process, error responses
4776 that can happen earlier (such as "too many connections") no longer contain
4779 12. If an expansion in a condition on a "warn" statement fails because a lookup
4780 defers, the "warn" statement is abandoned, and the next ACL statement is
4781 processed. Previously this caused the whole ACL to be aborted.
4783 13. Added the iplsearch lookup type.
4785 14. Added ident_timeout as a log selector.
4787 15. Added tls_certificate_verified as a log selector.
4789 16. Added a global option tls_require_ciphers (compare the smtp transport
4790 option of the same name). This controls incoming TLS connections.
4792 17. I finally figured out how to make tls_require_ciphers do a similar thing
4793 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4794 before starting the TLS session.
4796 18. Tabs are now shown as \t in -bP output.
4798 19. If the log selector return_path_on_delivery was set, Exim crashed when
4799 bouncing a message because it had too many Received: header lines.
4801 20. If two routers both had headers_remove settings, and the first one included
4802 a superfluous trailing colon, the final name in the first list and the
4803 first name in the second list were incorrectly joined into one item (with a
4804 colon in the middle).
4810 1. Added -C and -D options to the exinext utility, mainly to make it easier
4811 to include in the automated testing, but these could be helpful when
4812 multiple configurations are in use.
4814 2. The exinext utility was not formatting the output nicely when there was
4815 an alternate port involved in the retry record key, nor when there was a
4816 message id as well (for retries that were specific to a specific message
4817 and a specific host). It was also confused by IPv6 addresses, because of
4818 the additional colons they contain. I have fixed the IPv4 problem, and
4819 patched it up to do a reasonable job for IPv6.
4821 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4822 delivery, the log line now contains "pipelined" if PIPELINING was used.
4824 4. An SMTP transport process used to panic and die if the bind() call to set
4825 an explicit outgoing interface failed. This has been changed; it is now
4826 treated in the same way as a connect() failure.
4828 5. A reference to $sender_host_name in the part of a conditional expansion
4829 that was being skipped was still causing a DNS lookup. This no longer
4832 6. The def: expansion condition was not recognizing references to header lines
4833 that used bh_ and bheader_.
4835 7. Added the _cache feature to named lists.
4837 8. The code for checking quota_filecount in the appendfile transport was
4838 allowing one more file than it should have been.
4840 9. For compatibility with Sendmail, the command line option
4848 and sets the incoming protocol and host name (for trusted callers). The
4849 host name and its colon can be omitted when only the protocol is to be set.
4850 Note the Exim already has two private options, -pd and -ps, that refer to
4851 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4852 "s", but I don't think that's a major issue.
4854 10. A number of refactoring changes to the code, none of which should affect
4857 (a) The number of logging options was getting close to filling up the
4858 32-bit word that was used as a bit map. I have split them into two classes:
4859 those that are passed in the argument to log_write(), and those that are
4860 only ever tested independently outside of that function. These are now in
4861 separate 32-bit words, so there is plenty of room for expansion again.
4862 There is no change in the user interface or the logging behaviour.
4864 (b) When building, for example, log lines, the code previously used a
4865 macro that called string_cat() twice, in order to add two strings. This is
4866 not really sufficiently general. Furthermore, there was one instance where
4867 it was actually wrong because one of the argument was used twice, and in
4868 one call a function was used. (As it happened, calling the function twice
4869 did not affect the overall behaviour.) The macro has been replaced by a
4870 function that can join an arbitrary number of extra strings onto a growing
4873 (c) The code for expansion conditions now uses a table and a binary chop
4874 instead of a serial search (which was left over from when there were very
4875 few conditions). Also, it now recognizes conditions like "pam" even when
4876 the relevant support is not compiled in: a suitably worded error message is
4877 given if an attempt is made to use such a condition.
4879 11. Added ${time_interval:xxxxx}.
4881 12. A bug was causing one of the ddress fields not to be passed back correctly
4882 from remote delivery subprocesses. The field in question was not being
4883 subsequently used, so this caused to problems in practice.
4885 13. Added new log selectors queue_time and deliver_time.
4887 14. Might have fixed a bug in maildirsizefile handling that threw up
4888 "unexpected character" debug warnings, and recalculated the data
4889 unnecessarily. In any case, I expanded the warning message to give more
4892 15. Added the message "Restricted characters in address" to the statements in
4893 the default ACL that block characters like @ and % in local parts.
4895 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4896 Three changes have been made:
4898 (a) There was a serious bug; a negative response to MAIL caused the whole
4899 recipient domain to be cached as invalid, thereby blocking all messages
4900 to all local parts at the same domain, from all senders. This bug has
4901 been fixed. The domain is no longer cached after a negative response to
4902 MAIL if the sender used is not empty.
4904 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4907 (c) A new callout option, "use_sender" has been added for people who want
4908 the modified behaviour.
4914 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4917 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4918 indeed breaks things for older releases.
4920 3. Added additional logging to the case where there is a problem reading data
4921 from a filter that is running in a subprocess using a pipe, in order to
4922 try to track down a specific problem.
4924 4. Testing facility fudge: when running in the test harness and attempting
4925 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4926 getting "No route to host". Convert this to a timeout.
4928 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4931 6. Some OS don't have socklen_t but use size_t instead. This affects the
4932 fifth argument of getsockopt() amongst other things. This is now
4933 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4934 can be set for individual OS. I have set it for SunOS5, OSF1, and
4935 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4936 some earlier ones do not.
4938 7. Change 4.30/15 was not doing the test caselessly.
4940 8. The standard form for an IPv6 address literal was being rejected by address
4941 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4942 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4943 this, as well as the form without the "IPv6" on the front (but only when
4944 address literals are enabled, of course).
4946 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4948 10. Exim crashed if a message with an empty sender address specified by -f
4949 encountered a router with an errors_to setting. This could be provoked only
4950 by a command such as
4954 where an empty string was supplied; "<>" did not hit this bug.
4956 11. Installed PCRE release 4.5.
4958 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4959 remained set. It is now erased.
4961 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4962 times from message ids (which are base 36 rather than the normal 62).
4964 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4965 were being counted as actual protocol errors, and logged if the log
4966 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4967 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4968 and DATA following a set of rejected RCPTs do not count as protocol errors.
4969 In other words, Exim assumes they were pipelined, though this may not
4970 actually be the case. Of course, in all cases the client gets an
4971 appropriate error code.
4973 15. If a lookup fails in an ACL condition, a message about the failure may
4974 be available; it is used if testing the ACL cannot continue, because most
4975 such messages specify what the cause of the deferral is. However, some
4976 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4977 that caused an old message to be retained and used if a later statement
4978 caused a defer, replacing the real cause of the deferral.
4980 16. If an IP address had so many PTR records that the DNS lookup buffer
4981 was not large enough to hold them, Exim could crash while trying to process
4982 the truncated data. It now detects and logs this case.
4984 17. Further to 4.21/58, another change has been made: if (and only if) the
4985 first line of a message (the first header line) ends with CRLF, a bare LF
4986 in a subsequent header line has a space inserted after it, so as not to
4987 terminate the header.
4989 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4990 unnecessarily, used atoi() instead of strtol(), and didn't check the
4991 termination when getting file sizes from file names by regex.
4993 19. Completely re-implemented the support for maildirsize files, in the light
4994 of a number of problems with the previous contributed implementation
4995 (4.30/29). In particular:
4997 . If the quota is zero, the maildirsize file is maintained, but no quota is
5000 . If the maildir directory does not exist, it is created before any attempt
5001 to write a maildirsize file.
5003 . The quota value in the file is just a cache; if the quota is changed in
5004 the transport, the new value overrides.
5006 . A regular expression is available for excluding directories from the
5009 20. The autoreply transport checks the characters in options that define the
5010 message's headers; it allows continued headers, but it was checking with
5011 isspace() after an embedded newline instead of explicitly looking for a
5014 21. If all the "regular" hosts to which an address was routed had passed their
5015 expiry times, and had not reached their retry times, the address was
5016 bounced, even if fallback hosts were defined. Now Exim should go on to try
5019 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5020 equivalent code in the SMTP transport. Some hosts send humungous responses
5021 to HELO/EHLO, more than 1024 it seems.
5023 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5024 gives compiler warnings in some environments. I've now done it "properly",
5027 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5028 (because of problems with the built-in one) was declared to return uschar *
5029 instead of char *, causing compiler failure.
5031 25. Fixed a file descriptor leak when processing alias/forward files.
5033 26. Fixed a minor format string issue in dbfn.c.
5035 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5037 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5038 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5041 29. When a sender address is verified, it is cached, to save repeating the test
5042 when there is more than one recipient in a message. However, when the
5043 verification involves a callout, it is possible for different callout
5044 options to be set for different recipients. It is too complicated to keep
5045 track of this in the cache, so now Exim always runs a verification when a
5046 callout is required, relying on the callout cache for the optimization.
5047 The overhead is duplication of the address routing, but this should not be
5050 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5051 to be verified with callout=postmaster, and the main callout worked but the
5052 postmaster check failed, the verification correctly failed. However, if a
5053 subsequent RCPT command asked for sender verification *without* the
5054 postmaster check, incorrect caching caused this verification also to fail,
5057 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5058 it was not caching the DNS options (qualify_single, search_parents) that
5059 were used when the lookup failed. A subsequent lookup with different
5060 options therefore always gave the same answer, though there were cases
5061 where it should not have. (Example: a "domains = !$mx_any" option on a
5062 dnslookup router: the "domains" option is always processed without any
5063 widening, but the router might have qualify_single set.) Now Exim uses the
5064 cached value only when the same options are set.
5066 32. Added John Jetmore's "exipick" utility to the distribution.
5068 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5069 than a timeout (e.g. a certificate is required, and is not provided), an
5070 Exim server now closes the connection immediately. Previously it waited for
5071 the client to close - but if the client is SSL, it seems that they each
5072 wait for each other, leading to a delay before one of them times out.
5074 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5075 maintained 0.8.x compatibility because I don't think many are using it, and
5076 it is clearly obsolete.
5078 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5081 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5082 client certificate was expired. A simple patch fixes this, though I don't
5083 understand the full logic of why the verify callback is called multiple
5086 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5087 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5088 which causes problems with some clients (such as the Certicom SSL Plus
5089 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5090 disables the coutermeasure allowing Eudora to connect."
5092 38. Exim was not checking that a write() to a log file succeeded. This could
5093 lead to Bad Things if a log got too big, in particular if it hit a file
5094 size limit. Exim now panics and dies if it cannot write to a log file, just
5095 as it does if it cannot open a log file.
5097 39. Modified OS/Makefile-Linux so that it now contains
5099 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5101 The two -D definitions ensure that Exim is compiled with large file
5102 support, which makes it possible to handle log files that are bigger than
5105 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5106 instance) a domain was checked against a named list that involved a lookup,
5107 causing $domain_data to be set, then another domain was checked against the
5108 same list, then the first domain was re-checked, the value of $domain_data
5109 after the final check could be wrong. In particular, if the second check
5110 failed, it could be set empty. This bug probably also applied to
5113 41. The strip_trailing_dot option was not being applied to the address given
5114 with the -f command-line option.
5116 42. The code for reading a message's header from the spool was incrementing
5117 $received_count, but never initializing it. This meant that the value was
5118 incorrect (doubled) while delivering a message in the same process in which
5119 it was received. In the most common configuration of Exim, this never
5120 happens - a fresh exec is done - but it can happen when
5121 deliver_drop_privilege is set.
5123 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5124 it now includes up to 150 characters of the unexpected data in the log
5127 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5128 and building data strings. The size of both of these buffers was 10 000
5129 bytes - far larger than anybody would *ever* want, thought I. Needless to
5130 say, somebody hit the limit. I have increased the maximum line length to
5131 20 000 and the maximum data length of concatenated lines to 100 000. I have
5132 also fixed two bugs, because there was no checking on these buffers. Tsk,
5133 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5134 buffer is too small.
5136 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5137 lsearch lookups. Now it does.
5139 46. When parsing a route_list item in a manualroute router, a fixed-length
5140 buffer was used for the list of hosts. I made this 1024 bytes long,
5141 thinking that nobody would ever have a list of hosts that long. Wrong.
5142 Somebody had a whole pile of complicated expansion conditions, and the
5143 string was silently truncated, leading to an expansion error. It turns out
5144 that it is easier to change to an unlimited length (owing to other changes
5145 that have happened since this code was originally written) than to build
5146 structure for giving a limitation error. The length of the item that
5147 expands into the list of hosts is now unlimited.
5149 47. The lsearch lookup could not handle data where the length of text line was
5150 more than 4095 characters. Such lines were truncated, leading to shortened
5151 data being returned. It should now handle lines of any length.
5153 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5154 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5157 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5158 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5159 no longer happens in comment lines. A list of replacements is now placed
5160 at the head of all of the source files, except those whose only change is
5161 to replace PERL_COMMAND in the very first #! line.
5163 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5164 messages on the queue, with a bottom-up merge sort, using code contributed
5165 by Michael Haardt. This should make operations like -bp somewhat faster on
5166 large queues. It won't affect queue runners, except when queue_run_in_order
5169 51. Installed eximstats 1.31 in the distribution.
5171 52. Added support for SRV lookups to the dnslookup router.
5173 53. If an ACL referred to $message_body or $message_body_end, the value was not
5174 reset for any messages that followed in the same SMTP session.
5176 54. The store-handling optimization for building very long strings was not
5177 differentiating between the different store pools. I don't think this
5178 actually made any difference in practice, but I've tidied it.
5180 55. While running the routers to verify a sender address, $sender_address
5181 was still set to the sender address. This is wrong, because when routing to
5182 send a bounce to the sender, it would be empty. Therefore, I have changed
5183 it so that, while verifying a sender address, $sender_address is set to <>.
5184 (There is no change to what happens when verifying a recipient address.)
5186 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5187 target A or AAAA records (if not already returned) without resetting the
5188 qualify_single or search_parents options of the DNS resolver. These are
5189 inappropriate in this case because the targets of MX and SRV records must
5190 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5191 target that, when qualified, matched something in the local domain. These
5192 two options are now turned off when doing these lookups.
5194 57. It seems that at least some releases of Reiserfs (which does not have the
5195 concept of a fixed number of inodes) returns zero and not -1 for the
5196 number of available inodes. This interacted badly with check_spool_inodes,
5197 which assumed that -1 was the "no such thing" setting. What I have done is
5198 to check that the total number of inodes is greater than zero before doing
5199 the test of how many are available.
5201 58. When a "warn" ACL statement has a log_message modifier, the message is
5202 remembered, and not repeated. This is to avoid a lot of repetition when a
5203 message has many recipients that cause the same warning to be written.
5204 Howewer, Exim was preserving the list of already written lines for an
5205 entire SMTP session, which doesn't seem right. The memory is now reset if a
5206 new message is started.
5208 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5209 debugging output unless log_rewrite was also set.
5211 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5212 of (int)(handle) when we know that handle contains (void *)(-1).
5214 61. The Exim daemon panic-logs an error return when it closes the incoming
5215 connection. However "connection reset by peer" seems to be common, and
5216 isn't really an error worthy of noting specially, so that particular error
5219 62. When Exim is trying to find all the local interfaces, it used to panic and
5220 die if the ioctl to get the interface flags failed. However, it seems that
5221 on at least one OS (Solaris 9) it is possible to have an interface that is
5222 included in the list of interfaces, but for which you get a failure error
5223 for this call. This happens when the interface is not "plumbed" into a
5224 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5225 failure of the "get flags" call assumes that the interface is down.
5227 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5228 makes life easier for people who are doing arithmetic on fields extracted
5229 from dates, where you often get leading zeros that should not be
5230 interpreted as octal.
5232 64. Added qualify_domain to the redirect router, to override the global
5235 65. If a pathologically long header line contained very many addresses (the
5236 report of this problem mentioned 10 000) and each of them was rewritten,
5237 Exim could use up a very large amount of memory. (It kept on making new
5238 copies of the header line as it rewrote, and never released the old ones.)
5239 At the expense of a bit more processing, the header rewriting function has
5240 been changed so that it no longer eats memory in this way.
5242 66. The generation of the Received: header has been moved from the time that a
5243 message starts to be received, to the time that it finishes. The timestamp
5244 in the Received: header should now be very close to that of the <= log
5245 line. There are two side-effects of this change:
5247 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5248 logged header lines no longer include the local Received: line, because
5249 it has not yet been created. The same applies to a copy of the message
5250 that is returned to a non-SMTP sender when a message is rejected.
5252 (b) When a filter file is tested using -bf, no additional Received: header
5253 is added to the test message. After some thought, I decided that this
5256 This change does not affect the value of $received_for. It is still set
5257 after address rewriting, but before local_scan() is called.
5259 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5261 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5262 gave an unhelpful panic error message, and a defer error. I have managed to
5263 change this behaviour so that it now rejects any supplied certificate,
5264 which seems right, as the list of acceptable certificates is empty.
5266 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5267 gave an unhelpful defer error. I have not managed to make this reject any
5268 supplied certificates, but the error message it gives is "no certificate
5269 supplied", which is not helpful.
5271 70. exigrep's output now also includes lines that are not associated with any
5272 message, but which match the given pattern. Implemented by a patch from
5273 Martin Sluka, which also tidied up the Perl a bit.
5275 71. Recipient callout verification, like sender verification, was using <> in
5276 the MAIL FROM command. This isn't really the right thing, since the actual
5277 sender may affect whether the remote host accepts the recipient or not. I
5278 have changed it to use the actual sender in the callout; this means that
5279 the cache record is now keyed on a recipient/sender pair, not just the
5280 recipient address. There doesn't seem to be a real danger of callout loops,
5281 since a callout by the remote host to check the sender would use <>.
5282 [SEE ABOVE: changed after hitting problems.]
5284 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5285 temporary errors. However, in the case of such a code being given after
5286 the end of a data transmission (i.e. after ".") Exim was failing to write
5287 a retry record for the message. (Yes, there was some broken host that was
5288 actually sending 8xx at this point.)
5290 73. An unknown lookup type in a host list could cause Exim to panic-die when
5291 the list was checked. (An example that provoked this was putting <; in the
5292 middle of a list instead of at the start.) If this happened during a DATA
5293 ACL check, a -D file could be left lying around. This kind of configuration
5294 error no longer causes Exim to die; instead it causes a defer errror. The
5295 incident is still logged to the main and panic logs.
5297 74. Buglet left over from Exim 3 conversion. The message "too many messages
5298 in one connection" was written to the rejectlog but not the mainlog, except
5299 when address rewriting (yes!) was being logged.
5301 75. Added write_rejectlog option.
5303 76. When a system filter was run not as root (that is, when system_filter_user
5304 was set), the values of the $n variables were not being returned to the
5305 main process; thus, they were not subsequently available in the $sn
5308 77. Added +return_path_on_delivery log selector.
5310 78. A connection timeout was being treated differently from recipients deferred
5311 when testing hosts_max_try with a message that was older than the host's
5312 retry timeout. (The host should not be counted, thus allowing all hosts to
5313 be tried at least once before bouncing.) This may have been the cause of an
5314 occasionally reported bug whereby a message would remain on the queue
5315 longer than the retry timeout, but would be bounced if a delivery was
5316 forced. I say "may" because I never totally pinned down the problem;
5317 setting up timeout/retry tests is difficult. See also the next item.
5319 79. The ultimate address timeout was not being applied to errors that involved
5320 a combination of host plus message (for example, a timeout on a MAIL
5321 command). When an address resolved to a number of possible hosts, and they
5322 were not all tried for each delivery (e.g. because of hosts_max_try), a
5323 message could remain on the queue longer than the retry timeout.
5325 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5328 81. Fixed an obscure SMTP outgoing bug which required at least the following
5329 conditions: (a) there was another message waiting for the same server;
5330 (b) the server returned 5xx to all RCPT commands in the first message so
5331 that the message was not completed; (c) the server dropped the connection
5332 or gave a negative response to the RSET that Exim sends to abort the
5333 transaction. The observed case was a dropped connection after DATA that had
5334 been sent in pipelining mode. That is, the server had advertised PIPELINING
5335 but was not implementing it correctly. The effect of the bug was incorrect
5336 behaviour, such as trying another host, and this could lead to a crash.
5342 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5343 and daemon.c were passed as pointers to ints; they should have been
5344 pointers to socklen_t variables (which are typically unsigned ints).
5346 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5349 3. Fixed a really odd bug that affected only the testing scheme; patching a
5350 certain fixed string in the binary changed the value of another string that
5351 happened to be identical to the end of the original first string.
5353 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5354 name", it returns that address as the IP address. On some operating
5355 systems (e.g. Solaris), it also passes back the IP address string as the
5356 "host name". However, on others (e.g. Linux), it passes back an empty
5357 string. Exim wasn't checking for this, and was changing the host name to an
5358 empty string, assuming it had been canonicized.
5360 5. Although rare, it is permitted to have more than one PTR record for a given
5361 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5362 all the names associated with an address, because they do in Solaris.
5363 However, it seems that they do not in Linux for data that comes from the
5364 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5365 I found this out when I moved to a new Linux workstation and tried to run
5366 the Exim test suite.
5368 To get round this problem I have changed the code so that it now does its
5369 own call to the DNS to look up PTR records when searching for a host name.
5370 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5371 addresses that are only in /etc/hosts are still found.
5373 This behaviour is, however, controlled by an option called host_lookup_
5374 order, which defaults to "bydns:byaddr". If people want to use the other
5375 order, or indeed, just use one or the other means of lookup, they can
5376 specify it in this variable.
5378 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5379 some operating systems, this comes back from gethostbyaddr() as an empty
5380 string, and this is what Exim used to test for. However, it seems that in
5381 other systems, "." is yielded. Exim now tests for this case too.
5383 7. The values of check_spool_space and check_log_space are now held internally
5384 as a number of kilobytes instead of an absolute number of bytes. If a
5385 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5386 kilobyte. This means that much larger values can be stored.
5388 8. Exim monitor: an attempt to get the action menu when not actually pointing
5389 at a message produces an empty menu entitled "No message selected". This
5390 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5391 no entries in it ("Shell widget menu has zero width and/or height"). So I
5392 have added a single, blank menu entry in this case.
5394 9. Added ${quote_local_part.
5396 10. MIME decoding is now applied to the contents of Subject: header lines when
5399 11. Now that a reference to $sender_host_address automatically causes a reverse
5400 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5401 host lookup before query-style lookups in lists that might use this
5402 variable. This has therefore been abolished, and the "net-" prefix is no
5403 longer necessary for query-style lookups.
5405 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5406 have been a typo for LFLAGS, so it has been changed.
5408 13. The install script calls Exim with "-C /dev/null" in order to find the
5409 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5410 to be output. Howeve, since Exim outputs its version number before the
5411 error, it didn't break the script. It just looked ugly. I fixed this by
5412 always allowing "-C /dev/null" if the caller is root.
5414 14. Ignore overlarge ACL variable number when reading spool file - insurance
5415 against a later release with more variables having written the file.
5417 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5418 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5419 this, as well as the form without the "IPv6" on the front.
5421 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5422 OS/Makefile-Darwin file.
5424 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5425 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5427 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5428 variable. However, some OS use ints for inodes. I've added cast to long int
5429 to get rid of the compiler warning.
5431 19. I had forgotten to lock out "/../" in configuration file names when
5432 ALT_CONFIG_PREFIX was set.
5434 20. Routers used for verification do not need to specify transports. However,
5435 if such a router generated a host list, and callout was configured, Exim
5436 crashed, because it could not find a port number from the (non-existent)
5437 transport. It now assumes port 25 in this circumstance.
5439 21. Added the -t option to exigrep.
5441 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5442 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5443 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5444 them is set, LOOKUP_LSEARCH is forced.
5446 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5448 24. Added sender and host information to the "rejected by local_scan()" log
5449 line; previously there was no indication of these.
5451 25. Added .include_if_exists.
5453 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5454 Linux. It turns out that not all file systems support this. Apparently some
5455 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5456 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5457 not supported on the file descriptor, is now ignored when Exim is trying to
5458 sync a directory. This applies only to Linux.
5460 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5462 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5463 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5466 29. Added support for maildirsize files from supplied patch (modified a bit).
5468 30. The use of :fail: followed by an empty string could lead Exim to respond to
5469 sender verification failures with (e.g.):
5471 550 Verification failed for <xxx>
5472 550 Sender verify failed
5474 where the first response line was missing the '-' that indicates it is not
5475 the final line of the response.
5477 31. The loop for finding the name of the user that called Exim had a hardwired
5478 limit of 10; it now uses the value of finduser_retries, which is used for
5479 all other user lookups.
5481 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5484 33. Exim was neglecting to zero errno before one call of strtol() when
5485 expanding a string and expecting an integer value. On some systems this
5486 resulted in spurious "integer overflow" errors. Also, it was casting the
5487 result into an int without checking.
5489 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5490 did not work. The code looks as if it has *never* worked, though it appears
5491 to have been documented since at least releast 1.62. I have made it work.
5493 35. The "timeout_DNS" error in retry rules, also documented since at least
5494 1.62, also never worked. As it isn't clear exactly what this means, and
5495 clearly it isn't a major issue, I have abolished the feature by treating it
5496 as "timeout", and writing a warning to the main and panic logs.
5498 36. The display of retry rules for -brt wasn't always showing the error code
5501 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5502 timeout_connect_A, timeout_connect_MX.
5504 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5505 to the empty sender.
5507 39. The daemon was not analysing the content of -oX till after it had closed
5508 stderr and disconnected from the controlling terminal. This meant that any
5509 syntax errors were only noted on the panic log, and the return code from
5510 the command was 0. By re-arranging the code a little, I've made the
5511 decoding happen first, so such errors now appear on stderr, and the return
5512 code is 1. However, the actual setting up of the sockets still happens in
5513 the disconnected process, so errors there are still only recorded on the
5516 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5517 connections (as happens on some IP stacks) was logged at start up time as
5518 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5519 it from "IPv6 and IPv4", which means that two separate sockets are being
5522 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5523 says whether AF_INET or AF_INET6 was passed as an argument.
5525 42. Exiwhat output was messed up when time zones were included in log
5528 43. Exiwhat now gives more information about the daemon's listening ports,
5529 and whether -tls-on-connect was used.
5531 44. The "port" option of the smtp transport is now expanded.
5533 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5534 silently ignored. Now an error message is written to the main and panic
5537 46. There's a new ACL modifier called "logwrite" which writes to a log file
5538 as soon as it is encountered.
5540 47. Added $local_user_uid and $local_user_gid at routing time.
5542 48. Exim crashed when trying to verify a sender address that was being
5545 49. Exim was recognizing only a space character after ".include". It now also
5546 recognizes a tab character.
5548 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5549 extracting the relevant information from the specification. The man page no
5550 longer contains scrambled data for the -d option, and I've added a section
5551 at the front about calling Exim under different names.
5553 51. Added "extra_headers" argument to the "mail" command in filter files.
5555 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5558 53. Installed eximstats 1.29.
5560 54. Added transport_filter_timeout as a generic transport option.
5562 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5563 Cc: header lines. This was required by RFC 822, but it not required by RFC
5566 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5567 incoming messages that did not have them. Now it does so only if the
5568 message originates locally, that is, if there is no associated remote host
5569 address. When Resent- header lines are present, this applies to the Resent-
5570 lines rather than the non-Resent- lines.
5572 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5573 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5575 58. Messages for configuration errors now include the name of the main
5576 configuration file - useful now that there may be more than one file in a
5577 list (.included file names were always shown).
5579 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5580 for those rare installations that do not start the daemon as root or run it
5581 setuid root. I've cut out the call to initgroups() if the daemon is not
5584 60. The Exim user and group can now be bound into the binary as text strings
5585 that are looked up at the start of Exim's processing.
5587 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5589 62. Added $mailstore_basename variable.
5591 63. Installed patch to sieve.c from Michael Haardt.
5593 64. When Exim failed to open the panic log after failing to open the main log,
5594 the original message it was trying to log was written to stderr and debug
5595 output, but if they were not available (the usual case in production), it
5596 was lost. Now it is written to syslog before the two lines that record the
5597 failures to open the logs.
5599 65. Users' Exim filters run in subprocesses under the user's uid. It is
5600 possible for a "deliver" command or an alias in a "personal" command to
5601 provoke an address rewrite. If logging of address rewriting is configured,
5602 this fails because the process is not running as root or exim. There may be
5603 a better way of dealing with this, but for the moment (because 4.30 needs
5604 to be released), I have disabled address rewrite logging when running a
5605 filter in a non-root, non-exim process.
5611 1. The buildconfig auxiliary program wasn't quoting the value set for
5612 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5613 not defined. This bug was masked in 4.22 by the effect that was fixed in
5616 2. Some messages that were rejected after a message id was allocated were
5617 shown as "incomplete" by exigrep. It no longer does this for messages that
5618 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5620 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5621 have allow_domain_literals set, the ID did not get logged in the <= line.
5622 Domain literals are now always recognized in Message-ID: header lines.
5624 4. The first argument for a ${extract expansion item is the key name or field
5625 number. Leading and trailing spaces in this item were not being ignored,
5626 causing some misleading effects.
5628 5. When deliver_drop_privilege was set, single queue runner processes started
5629 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5630 same command in the process it spins off) were not dropping privilege.
5632 6. When the daemon running as "exim" started a queue runner, it always
5633 re-executed Exim in the spun-off process. This is a waste of effort when
5634 deliver_drop_privilege is set. The new process now just calls the
5635 queue-runner function directly.
5641 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5644 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5645 ignored. Though the use of -oP was forcing the writing of a pid file, it
5646 was always written to the default place.
5648 3. If the message "no IP address found for host xxxx" is generated during
5649 incoming verification, it is now followed by identification of the incoming
5650 connection (so you can more easily find what provoked it).
5652 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5654 5. Added some features to "harden" Exim a bit more against certain attacks:
5656 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5657 be put in Local/Makefile. This is like the never_users runtime option,
5658 but it cannot be overridden. The default setting is "root".
5660 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5661 prefix string with which any file named in a -C command line option
5664 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5665 is retained for -C and -D only if the caller of Exim is root. Without
5666 it, the exim user may also use -C and -D and retain privilege.
5668 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5669 command line option is disabled.
5671 6. Macro names set by the -D option must start with an upper case letter, just
5672 like macro names defined in the configuration file.
5674 7. Added "dereference=" facility to LDAP.
5676 8. Two instances of the typo "uknown" in the source files are fixed.
5678 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5679 the Configure-Makefile script screwed up while processing it.
5681 10. Incorporated PCRE 4.4.
5683 11. The SMTP synchronization check was not operating right at the start of an
5684 SMTP session. For example, it could not catch a HELO sent before the client
5685 waited for the greeting. There is now a check for outstanding input at the
5686 point when the greeting is written. Because of the duplex, asynchronous
5687 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5688 way, but not yet received, when the check is performed.
5690 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5691 on TCP/IP sockets, because this apparently causes some broken clients to
5694 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5695 unchanged) from the Cygwin maintainer.
5697 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5698 was supported instead of testing for mbx. Effectively a typo.
5700 15. The spa authenticator server code was not checking that the input it
5701 received was valid base64.
5703 16. The debug output line for the "set" modifier in ACLs was not showing the
5704 name of the variable that was being set.
5706 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5708 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5709 lookup on its own. Something else was needed to trigger the lookup. For
5710 example, a match in host_lookup or the need for a host name in a host list.
5711 Now, if $sender_host_name is referenced and the host name has not yet been
5712 looked up, a lookup is performed. If the lookup fails, the variable remains
5713 empty, and $host_lookup_failed is set to "1".
5715 19. Added "eqi" as a case-independent comparison operator.
5717 20. The saslauthd authentication condition could segfault if neither service
5718 nor realm was specified.
5720 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5721 error message that was logged was misleading, and incoming SMTP
5722 connections were dropped. The message is now more accurate, and temporary
5723 errors are given to SMTP connections.
5725 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5726 (except RSET) with a 421 error, until QUIT is received. However, it was
5727 failing to send a response to QUIT.
5729 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5730 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5731 ACL is now run after the helo_try_verify_hosts code.
5733 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5734 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5735 case-independent, so other case variants are also recognized). Apparently
5736 some systems use these upper case variants.
5738 25. If more than two messages were waiting for the same host, and a transport
5739 filter was specified for the transport, Exim sent two messages over the
5740 same TCP/IP connection, and then failed with "socket operation on non-
5741 socket" when it tried to send the third.
5743 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5745 27. The extern definition of crypt16() in expand.c was not being excluded when
5746 the OS had its own crypt16() function.
5748 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5749 as a preferred synonym for return_size_limit, both as an option and as an
5752 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5754 30. Changed the default configuration ACL to relax the local part checking rule
5755 for addresses that are not in any local domains. For these addresses,
5756 slashes and pipe symbols are allowed within local parts, but the sequence
5757 /../ is explicitly forbidden.
5759 31. SPA server authentication was not clearing the challenge buffer before
5762 32. log_message in a "warn" ACL statement was writing to the reject log as
5763 well as to the main log, which contradicts the documentation and doesn't
5764 seem right (because no rejection is happening). So I have stopped it.
5766 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5767 However, I am unable to do any testing of this.
5769 34. Fixed an infelicity in the appendfile transport. When checking directories
5770 for a mailbox, to see if any needed to be created, it was accidentally
5771 using path names with one or more superfluous leading slashes; tracing
5772 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5774 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5775 contents are added to the log line that is written for every discarded
5776 recipient. (Previously a log_message setting was ignored.)
5778 36. The ${quote: operator now quotes the string if it is empty.
5780 37. The install script runs exim in order to find its version number. If for
5781 some reason other than non-existence or emptiness, which it checks, it
5782 could not run './exim', it was installing it with an empty version number,
5783 i.e. as "exim-". This error state is now caught, and the installation is
5786 38. An argument was missing from the function that creates an error message
5787 when Exim fails to connect to the socket for saslauthd authentication.
5788 This could cause Exim to crash, or give a corrupted message.
5790 39. Added isip, isip4, and isip6 to ${if conditions.
5792 40. The ACL variables $acl_xx are now saved with the message, and can be
5793 accessed later in routers, transports, and filters.
5795 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5796 strings in the file are not string-expanded.
5798 42. If a MAIL command specified a SIZE value that was too large to fit into an
5799 int variable, the check against message_size_limit failed. Such values are
5800 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5801 day this will have to be increased, but I don't think I want to be around
5802 when emails are that large.
5809 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5810 iconv() is not standard in FreeBSD.
5812 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5813 IPv6 enabled. The observed symptom was a segmentation fault on return from
5814 the function os_common_find_running_interfaces() in src/os.c.
5816 3. In the check_special_case() function in daemon.c I had used "errno" as an
5817 argument name, which causes warnings on some systems. This was basically a
5818 typo, since it was named "eno" in the comments!
5820 4. The code that waits for the clock to tick (at a resolution of some fraction
5821 of a second) so as to ensure message-id uniqueness was always waiting for
5822 at least one whole tick, when it could have waited for less. [This is
5823 almost certainly not relevant at current processor speeds, where it is
5824 unlikely to ever wait at all. But we try to future-proof.]
5826 5. The function that sleeps for a time interval that includes fractions of a
5827 second contained a race. It did not block SIGALRM between setting the
5828 timer, and suspending (a couple of lines later). If the interval was short
5829 and the sigsuspend() was delayed until after it had expired, the suspension
5830 never ended. On busy systems this could lead to processes getting stuck for
5833 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5834 process, before it forks any delivery processes. The open lookup caching
5835 mechanism meant that the open file or database connection was passed into
5836 the delivery process. The problem was that delivery processes always tidy
5837 up cached lookup data. This could cause a problem for the next delivery
5838 process started by the queue runner, because the external queue runner
5839 process does not know about the closure. So the next delivery process
5840 still has data in the lookup cache. In the case of a file lookup, there was
5841 no problem because closing a file descriptor in a subprocess doesn't affect
5842 the parent. However, if the lookup was caching a connection to a database,
5843 the connection was closed, and the second delivery process was likely to
5844 see errors such as "PGSQL: query failed: server closed the connection
5845 unexpectedly". The problem has been fixed by closing all cached lookups
5846 in a queue runner before running a delivery process.
5848 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5849 seem to have the "const" qualifier which it has on other OS. I've
5852 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5853 *of the same type* (client or server) with the same public name that an
5854 error should be diagnosed.
5856 9. When Exim looked up a host name for an IP address, but failed to find the
5857 original IP address when looking up the host name (a safety check), it
5858 output the message "<ip address> does not match any IP for NULL", which was
5859 confusing, to say the least. The bug was that the host name should have
5860 appeared instead of "NULL".
5862 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5863 user that is built into the binary, and the -C or -D options is used, root
5864 privilege is dropped before the configuration file is read. In addition,
5865 logging is switched to stderr instead of the normal log files. If the
5866 configuration then re-defines the Exim user, the unprivileged environment
5867 is probably not what is expected, so Exim logs a panic warning message (but
5870 However, if deliver_drop_privilege is set, the unprivileged state may well
5871 be exactly what is intended, so the warning has been cut out in that case,
5872 and Exim is allowed to try to write to its normal log files.
5878 1. smtp_return_error_details was not giving details for temporary sender
5879 or receiver verification errors.
5881 2. Diagnose a configuration error if two authenticators have the same public
5884 3. Exim used not to create the message log file for a message until the first
5885 delivery attempt. This could be confusing when incoming messages were held
5886 for policy or load reasons. The message log file is now created at the time
5887 the message is received, and an initial "Received" line is written to it.
5889 4. The automatically generated man page for command line options had a minor
5890 bug that caused no ill effects; however, a more serious problem was that
5891 the procedure for building the man page automatically didn't always
5892 operate. Consequently, release 4.20 contains an out-of-date version. This
5893 shouldn't happen again.
5895 5. When building Exim with embedded Perl support, the script that builds the
5896 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5897 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5899 6. The freeze_tell option was not being used for messages that were frozen on
5900 arrival, either by an ACL or by local_scan().
5902 7. Added the smtp_incomplete_transaction log selector.
5904 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5905 was accepting AUTH without a new EHLO.
5907 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5908 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5911 10. Logging of TCP/IP connections (when configured) now happens in the main
5912 daemon process instead of the child process, so that the TCP/IP connection
5913 count is more accurate (but it can never be perfect).
5915 11. The use of "drop" in a nested ACL was not being handled correctly in the
5916 outer ACL. Now, if condition failure induced by the nested "drop" causes
5917 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5918 or "require"), the connection is dropped.
5920 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5921 that yield "discard" can now be used with an "accept" or a "discard" verb,
5922 but an error is generated for any others (because I can't see a useful way
5923 to define what should happen).
5925 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5926 are now processed in the same way as lines in the Exim configuration file.
5927 In particular, continuation lines are supported.
5929 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5931 15. Added -ti meaning -t -i.
5933 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5934 domains, and warn by logging if others are found.
5936 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5937 returned from GIFCONF when Exim is trying to find the list of interfaces on
5938 a host. The code in os.c has been modified to copy each ifreq to an aligned
5939 structure in all cases.
5941 Also, in some cases, the returned ifreq's were being copied to a 'struct
5942 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5943 means the last couple of bytes of an IPv6 address could be chopped if the
5944 ifreq contained only a normal sockaddr (14 bytes storage).
5946 18. Named domain lists were not supported in the hosts_treat_as_local option.
5947 An entry such as +xxxx was not recognized, and was treated as a literal
5950 19. Ensure that header lines added by a DATA ACL are included in the reject log
5951 if the ACL subsequently rejects the message.
5953 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5954 MD5 (which is deprecated).
5956 21. When testing a filter file using -bf, Exim was writing a message when it
5957 took the sender from a "From " line in the message, but it was not doing so
5958 when it took $return_path from a Return-Path: header line. It now does.
5960 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5961 with a valid header line field name (a series of printing characters
5962 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5964 23. Changed "disc" in the source to "disk" to conform to the documentation and
5965 the book and for uniformity.
5967 24. Ignore Sendmail's -Ooption=value command line item.
5969 25. When execve() failed while trying to run a command in a pipe transport,
5970 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5971 could be confused with a return value of 69 from the command itself. This
5972 has been changed to 127, the value the shell returns if it is asked to run
5973 a non-existent command. The wording for the related log line suggests a
5974 non-existent command as the problem.
5976 26. If received_header_text expands to an empty string, do not add a Received:
5977 header line to the message. (Well, it adds a token one on the spool, but
5978 marks it "old" so that it doesn't get used or transmitted.)
5980 27. Installed eximstats 1.28 (addition of -nt option).
5982 28. There was no check for failure on the call to getsockname() in the daemon
5983 code. This can fail if there is a shortage of resources on the system, with
5984 ENOMEM, for example. A temporary error is now given on failure.
5986 29. Contrary to the C standard, it seems that in some environments, the
5987 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5988 program. Exim now does this explicitly; it affects the formatting of
5989 timestamps using strftime().
5991 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5992 complaints. I've now initialized all the variables, to avoid this.
5994 32. Header lines added by a system filter were not being "seen" during
5995 transport-time rewrites.
5997 33. The info_callback() function passed to OpenSSL is set up with type void
5998 (*)(SSL *, int, int), as described somewhere. However, when calling the
5999 function (actually a macro) that sets it up, the type void(*)() is
6000 expected. I've put in a cast to prevent warnings from picky compilers.
6002 34. If a DNS black list lookup found a CNAME record, but there were no A
6003 records associated with the domain it pointed at, Exim crashed.
6005 35. If a DNS black list lookup returned more than one A record, Exim ignored
6006 all but the first. It now scans all returned addresses if a particular IP
6007 value is being sought. In this situation, the contents of the
6008 $dnslist_value variable are a list of all the addresses, separated by a
6011 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6012 to accept a host name and all its aliases if the forward lookup for any of
6013 them yielded the IP address of the incoming connection. Now it accepts only
6014 those names whose forward lookup yields the correct IP address. Any other
6015 names are discarded. This closes a loophole whereby a rogue DNS
6016 administrator could create reverse DNS records to break through a
6017 wildcarded host restriction in an ACL.
6019 37. If a user filter or a system filter that ran in a subprocess used any of
6020 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6021 the wrong values were passed to the pipe command ($thisaddress had the
6022 value of $0, $0 had the value of $1, etc). This bug was introduced by
6023 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6025 38. Improved the line breaking for long SMTP error messages from ACLs.
6026 Previously, if there was no break point between 40 and 75 characters, Exim
6027 left the rest of the message alone. Two changes have been made: (a) I've
6028 reduced the minimum length to 35 characters; (b) if it can't find a break
6029 point between 35 and 75 characters, it looks ahead and uses the first one
6030 that it finds. This may give the occasional overlong line, but at least the
6031 remaining text gets split now.
6033 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6034 file descriptors might be low, and that setting 1000 would always raise it.
6035 It turns out that in some environments, the limit is already over 1000 and
6036 that lowering it causes trouble. So now Exim takes care not to decrease it.
6038 40. When delivering a message, the value of $return_path is set to $sender_
6039 address at the start of routing (routers may change the value). By an
6040 oversight, this default was not being set up when an address was tested by
6041 -bt or -bv, which affected the outcome if any router or filter referred to
6044 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6045 log only when "message" or "log_message" are set. However, if one of the
6046 conditions was an address verification, or a call to a nested ACL, the
6047 messages generated by the underlying test were being passed through. This
6048 no longer happens. The underlying message is available in $acl_verify_
6049 message for both "message" and "log_message" expansions, so it can be
6050 passed through if needed.
6052 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6053 new expansion $bh_ to give the encoded byte string without charset
6054 translation. Translation happens only if iconv() is available; HAVE_ICONV
6055 indicates this at build time. HEADERS_CHARSET gives the charset to
6056 translate to; headers_charset can change it in the configuration, and
6057 "headers charset" can change it in an individual filter file.
6059 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6060 that creates RFC 2047 encoded "words" labels them as that charset instead
6061 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6062 expansion operator; (ii) when Exim creates a From: line for a local
6063 message; (iii) when a header line is rewritten to include a "phrase" part.
6065 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6066 buggy, causing it to skip the first lines of messages whose message ID
6067 ended in 'D'. This would not have bitten before Exim release 4.14, because
6068 message IDs were unlikely to end in 'D' before then. The effect was to have
6069 incorrect size information for certain domains.
6071 45. #include "config.h" was missing at the start of the crypt16.c module. This
6072 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6075 46. If there was a timeout during a "random" callout check, Exim treated it as
6076 a failure of the random address, and carried on sending RSET and the real
6077 address. If the delay was just some slowness somewhere, the response to the
6078 original RCPT would be taken as a response to RSET and so on, causing
6079 mayhem of various kinds.
6081 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6082 when I implemented it. It didn't allow for the fact that some option values
6083 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6084 the right test for positive values.
6086 48. Domain names in DNS records are case-independent. Exim always looks them up
6087 in lower case. Some resolvers return domain names in exactly the case they
6088 appear in the zone file, that is, they may contain uppercase letters. Not
6089 all resolvers do this - some return always lower case. Exim was treating a
6090 change of case by a resolver as a change of domain, similar to a widening
6091 of a domain abbreviation. This triggered its re-routing code and so it was
6092 trying to route what was effectively the same domain again. This normally
6093 caused routing to fail (because the router wouldn't handle the domain
6094 twice). Now Exim checks for this case specially, and just changes the
6095 casing of the domain that it ultimately uses when it transmits the message
6098 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6101 50. If a filter generated a file delivery with a non-absolute name (possible if
6102 no home directory exists for the router), the forbid_file option was not
6105 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6106 the existing equality matching.
6108 52. Exim was using ints instead of ino_t variables in some places where it was
6109 dealing with inode numbers.
6111 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6112 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6113 called TMPDIR, and if it finds it is different, it changes its value.
6115 54. The smtp_printf() function is now made available to local_scan() so
6116 additional output lines can be written before returning. There is also an
6117 smtp_fflush() function to enable the detection of a dropped connection.
6118 The variables smtp_input and smtp_batched_input are exported to
6121 55. Changed the default runtime configuration: the message "Unknown user"
6122 has been removed from the ACL, and instead placed on the localuser router,
6123 using the cannot_route_message feature. This means that any verification
6124 failures that generate their own messages won't get overridden. Similarly,
6125 the "Unrouteable address" message that was in the ACL for unverifiable
6126 relay addresses has also been removed.
6128 56. Added hosts_avoid_esmtp to the smtp transport.
6130 57. The exicyclog script was not checking for the esoteric option
6131 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6132 will work only if exicyclog is run under the appropriate euid.
6134 58. Following a discussion on the list, the rules by which Exim recognises line
6135 endings on incoming messages have been changed. The -dropcr and drop_cr
6136 options are now no-ops, retained only for backwards compatibility. The
6137 following line terminators are recognized: LF CRLF CR. However, special
6138 processing applies to CR:
6140 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6141 nor a local message in the state where . is a terminator.
6143 (ii) If a bare CR is encountered in a header line, an extra space is added
6144 after the line terminator so as not to end the header. The reasoning
6145 behind this is that bare CRs in header lines are most likely either
6146 to be mistakes, or people trying to play silly games.
6148 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6149 was being incorrectly given as 18 bytes larger than it should have been.
6150 This is a VOB (very old bug).
6152 60. This may never have affected anything current, but just in case it has:
6153 When the local host is found other than at the start of a list of hosts,
6154 the local host, those with the same MX, and any that follow, are discarded.
6155 When the list in question was part of a longer list of hosts, the following
6156 hosts (not currently being processed) were also being discarded. This no
6157 longer happens. I'm not sure if this situation could ever has previously
6160 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6163 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6164 header. This is something that is recommended in a new Internet Draft, and
6165 is something that is documented as being done by Sendmail. There are two
6166 possible values. For messages generated by the autoreply transport, Exim
6169 Auto-Submitted: auto-replied
6171 whereas for all other generated messages (e.g. bounces) it adds
6173 Auto-Submitted: auto-generated
6175 63. The "personal" condition in filters now includes a test for the
6176 Auto-Submitted: header. If it contains the string "auto-" the message it
6177 not considered personal.
6179 64. Added rcpt_include_affixes as a generic transport option.
6181 65. Added queue_only_override (default true).
6183 66. Added the syslog_duplication option.
6185 67. If what should have been the first header line of a message consisted of
6186 a space followed by a colon, Exim was mis-interpreting it as a header line.
6187 It isn't of course - it is syntactically invalid and should therefore be
6188 treated as the start of the message body. The misbehaviour could have
6189 caused a number of strange effects, including loss of data in subsequent
6190 header lines, and spool format errors.
6192 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6193 client host had authenticated. This control can now be exercised by an ACL
6194 for more flexibility.
6196 69. By default, callouts do not happen when testing with -bh. There is now a
6197 variant, -bhc, which does actually run the callout code, including
6198 consulting and updating the callout cache.
6200 70. Added support for saslauthd authentication, courtesy of Alexander
6203 71. If statvfs() failed on the spool or log directories while checking their
6204 size for availability, Exim confusingly gave the error "space shortage".
6205 Furthermore, in debugging mode it crashed with a floating point exception.
6206 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6207 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6208 serious problem, Exim now writes to the main and panic logs when this
6209 happens, with details of the failure. It then refuses to accept the
6210 incoming message, giving the message "spool directory problem" or "log
6211 directory problem" with a 421 code for SMTP messages.
6213 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6214 0, 1, and 2 exist, because some OS complain for execs without them (see
6215 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6216 descriptors. However, the code omitted to check that the open succeeded,
6217 causing mysterious errors if for some reason the permissions on /dev/null
6218 got screwed. Now Exim writes a message to the main and panic logs, and
6219 bombs out if it can't open /dev/null.
6221 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6222 interact so that it is all more flexible. It is supposed to remain
6223 backwards compatible. Also added extra_local_interfaces.
6225 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6226 to bomb out with an assertion failure - to the client this appears as a
6227 connection drop. This problem occurs in the part of the code that was taken
6228 from the Samba project. Fortunately, the assertion is in a very simple
6229 function, so I have fixed this by reproducing the function inline in the
6230 one place where it is called, and arranging for authentication to fail
6231 instead of killing the process with assert().
6233 75. The SPA client code was not working when the server requested OEM rather
6234 than Unicode encoding.
6236 76. Added code to make require_files with a specific uid setting more usable in
6237 the case where statting the file as root fails - usually a non-root-mounted
6238 NFS file system. When this happens and the failure is EACCES, Exim now
6239 forks a subprocess and does the per-uid checking as the relevant uid.
6241 77. Added process_log_path.
6243 78. If log_file_path was not explicitly set, a setting of check_log_space or
6244 check_log_inodes was ignored.
6246 79. If a space check for the spool or log partitions fails, the incident is now
6247 logged. Of course, in the latter case the data may get lost...
6249 80. Added the %p formatting code to string_format() so that it can be used to
6250 print addresses in debug_print(). Adjusted all the address printing in the
6251 debugging in store.c to use %p rather than %d.
6253 81. There was a concern that a line of code in smtp_in.c could overflow a
6254 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6255 initially expressed, the concern was not well-founded, because trailing
6256 spaces are removed early. However, if the trailing spaces were followed by
6257 a NULL, they did not get removed, so the overflow was possible. Two fixes
6260 (a) I re-wrote the offending code in a cleaner fashion.
6261 (b) If an incoming SMTP command contains a NULL character, it is rejected
6264 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6265 runs initgroups(), so that if the Exim user is in any additional groups,
6266 they will be used during message reception.
6272 The change log for 4.20 and earlier releases has been archived.