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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
85 TL/01 When running the test suite, the README says that variables such as
86 no_msglog_check are global and can be placed anywhere in a specific
87 test's script, however it was observed that placement needed to be near
88 the beginning for it to behave that way. Changed the runtest perl
89 script to read through the entire script once to detect and set these
90 variables, reset to the beginning of the script, and then run through
91 the script parsing/test process like normal.
93 TL/02 The BSD's have an arc4random API. One of the functions to induce
94 adding randomness was arc4random_stir(), but it has been removed in
95 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
96 function when detected.
98 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
99 cause callback expansion.
101 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
102 syntax errors in an expansion can be treated as a string instead of
103 logging or causing an error, due to the internal use of bool_lax
104 instead of bool when processing it.
106 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
107 server certificates when making smtp deliveries.
109 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
111 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
113 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
115 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
116 Merged patch from Sebastian Wiedenroth.
118 JH/05 Fix results-pipe from transport process. Several recipients, combined
119 with certificate use, exposed issues where response data items split
120 over buffer boundaries were not parsed properly. This eventually
121 resulted in duplicates being sent. This issue only became common enough
122 to notice due to the introduction of conection certificate information,
123 the item size being so much larger. Found and fixed by Wolfgang Breyha.
125 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
126 size buffer was used, resulting in syntax errors when an expansion
129 JH/07 Add support for directories of certificates when compiled with a GnuTLS
130 version 3.3.6 or later.
132 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
133 is EXPERIMENTAL_EVENT, the main-configuration and transport options
134 both become "event_action", the variables become $event_name, $event_data
135 and $event_defer_errno. There is a new variable $verify_mode, usable in
136 routers, transports and related events. The tls:cert event is now also
137 raised for inbound connections, if the main configuration event_action
140 TL/06 In test suite, disable OCSP for old versions of openssl which contained
141 early OCSP support, but no stapling (appears to be less than 1.0.0).
143 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
144 server certificate names available under the smtp transport option
145 "tls_verify_cert_hostname" now do not permit multi-component wildcard
148 JH/10 Time-related extraction expansions from certificates now use the main
149 option "timezone" setting for output formatting, and are consistent
150 between OpenSSL and GnuTLS compilations. Bug 1541.
152 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
153 encoded parameter in the incoming message. Bug 1558.
155 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
156 include certificate info, eximon was claiming there were spoolfile
159 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
161 JH/14 Log delivery-related information more consistently, using the sequence
162 "H=<name> [<ip>]" wherever possible.
164 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
165 are problematic for Debian distribution, omit them from the release
168 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
170 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
173 JH/17 Fix a null-indirection in certextract expansions when a nondefault
174 output list separator was used.
179 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
180 checkers that were complaining about end of non-void function with no
183 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
184 This was a regression intruduced in 4.83 by another bugfix.
186 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
188 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
189 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
195 TF/01 Correctly close the server side of TLS when forking for delivery.
197 When a message was received over SMTP with TLS, Exim failed to clear up
198 the incoming connection properly after forking off the child process to
199 deliver the message. In some situations the subsequent outgoing
200 delivery connection happened to have the same fd number as the incoming
201 connection previously had. Exim would try to use TLS and fail, logging
202 a "Bad file descriptor" error.
204 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
205 utilities have not been installed.
207 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
208 temporary space as the ACL may create new global variables.
210 TL/01 LDAP support uses per connection or global context settings, depending
211 upon the detected version of the libraries at build time.
213 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
214 to extract and use the src ip:port in logging and expansions as if it
215 were a direct connection from the outside internet. PPv2 support was
216 updated based on HAProxy spec change in May 2014.
218 JH/02 Add ${listextract {number}{list}{success}{fail}}.
220 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
221 Properly escape header and check for NULL return.
223 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
226 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
228 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
229 characters in header names, implemented as a verify condition.
230 Contributed by Michael Fischer v. Mollard.
232 TL/05 Rename SPF condition results err_perm and err_temp to standardized
233 results permerror and temperror. Previous values are deprecated but
234 still accepted. In a future release, err_perm and err_temp will be
235 completely removed, which will be a backward incompatibility if the
236 ACL tests for either of these two old results. Patch contributed by
237 user bes-internal on the mailing list.
239 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
241 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
242 selectors, in both main and reject logs.
244 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
247 JH/07 Add malware type "sock" for talking to simple daemon.
249 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
251 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
252 routers/transports under cutthrough routing.
254 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
255 numbers. Touch up "bool" conditional to keep the same definition.
257 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
259 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
261 JH/12 Expand items in router/transport headers_add or headers_remove lists
262 individually rather than the list as a whole. Bug 1452.
264 Required for reasonable handling of multiple headers_ options when
265 they may be empty; requires that headers_remove items with embedded
266 colons must have them doubled (or the list-separator changed).
268 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
269 view the policy declared in the DMARC record. Currently, $dmarc_status
270 is a combined value of both the record presence and the result of the
273 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
275 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
276 dnslookup router and the smtp transport (applying to the forward
279 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
280 of ldap servers used for a specific lookup. Patch provided by Heiko
283 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
284 New variable $lookup_dnssec_authenticated for observability.
286 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
287 Patch submitted by Lars Timman.
289 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
291 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
292 Requires trusted mode and valid format message id, aborts otherwise.
293 Patch contributed by Heiko Schlichting.
295 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
296 certextract with support for various fields. Bug 1358.
298 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
299 is requested by default, modifiable by smtp transport option
302 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
303 operate on certificate variables to give certificate fingerprints
304 Also new ${sha256:cert_variable}.
306 JH/23 The PRDR feature is moved from being Experimental into the mainline.
308 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
309 Christian Aistleitner.
311 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
313 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
314 file. Patch from Wolfgang Breyha.
316 JH/25 Expand the coverage of the delivery $host and $host_address to
317 client authenticators run in verify callout. Bug 1476.
319 JH/26 Port service names are now accepted for tls_on_connect_ports, to
320 align with daemon_smtp_ports. Bug 72.
322 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
323 support and error reporting did not work properly.
325 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
326 and is readable. Patch from Andrew Colin Kissa.
328 TL/14 Enhance documentation of ${run expansion and how it parses the
329 commandline after expansion, particularly in the case when an
330 unquoted variable expansion results in an empty value.
332 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
334 PP/02 Fix internal collision of T_APL on systems which support RFC3123
335 by renaming away from it. Addresses GH issue 15, reported by
338 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
340 TL/15 SECURITY: prevent double expansion in math comparison functions
341 (can expand unsanitized data). Not remotely exploitable.
348 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
350 PP/02 Make -n do something, by making it not do something.
351 When combined with -bP, the name of an option is not output.
353 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
356 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
357 $sender_host_name and config options to manage this, and basic check
360 PP/05 DSCP support for outbound connections and control modifier for inbound.
362 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
363 (Only plugin which currently uses this is kerberos4, which nobody should
364 be using, but we should make it available and other future plugins might
365 conceivably use it, even though it would break NAT; stuff *should* be
366 using channel bindings instead).
368 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
369 name; added for Sendmail compatibility; requires admin caller.
370 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
371 just ignore it); requires trusted caller.
372 Also parse but ignore: -Ac -Am -X<logfile>
375 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
377 TL/02 Add +smtp_confirmation as a default logging option.
379 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
380 Patch by Magnus Holmgren from 2007-02-20.
382 TL/04 Bugzilla 1281 - Spec typo.
383 Bugzilla 1283 - Spec typo.
384 Bugzilla 1290 - Spec grammar fixes.
386 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
388 TL/06 Add Experimental DMARC support using libopendmarc libraries.
390 TL/07 Fix an out of order global option causing a segfault. Reported to dev
391 mailing list by by Dmitry Isaikin.
393 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
395 JH/02 Support "G" suffix to numbers in ${if comparisons.
397 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
399 NM/01 Bugzilla 1197 - Spec typo
400 Bugzilla 1196 - Spec examples corrections
402 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
404 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
405 gnutls_enable_pkcs11, but renamed to more accurately indicate its
408 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
409 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
411 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
412 "acl {{name}{arg}...}", and optional args on acl condition
415 JH/05 Permit multiple router/transport headers_add/remove lines.
417 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
419 JH/07 Avoid using a waiting database for a single-message-only transport.
420 Performance patch from Paul Fisher. Bugzilla 1262.
422 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
425 JH/09 Add $headers_added variable, with content from use of ACL modifier
426 add_header (but not yet added to the message). Bugzilla 199.
428 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
429 Pulled from Bugzilla 817 by Wolfgang Breyha.
431 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
433 (nb: this is the same fix as in Exim 4.80.1)
435 JH/11 Add A= logging on delivery lines, and a client_set_id option on
438 JH/12 Add optional authenticated_sender logging to A= and a log_selector
441 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
443 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
444 advertise SMTP AUTH mechanism to us, instead of a generic
445 protocol violation error. Also, make Exim more robust to bad
446 data from the Dovecot auth socket.
448 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
450 When a queue runner is handling a message, Exim first routes the
451 recipient addresses, during which it prunes them based on the retry
452 hints database. After that it attempts to deliver the message to
453 any remaining recipients. It then updates the hints database using
456 So if a recipient address works intermittently, it can get repeatedly
457 deferred at routing time. The retry hints record remains fresh so the
458 address never reaches the final cutoff time.
460 This is a fairly common occurrence when a user is bumping up against
461 their storage quota. Exim had some logic in its local delivery code
462 to deal with this. However it did not apply to per-recipient defers
463 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
465 This change adds a proper retry rule check during routing so that the
466 final cutoff time is checked against the message's age. We only do
467 this check if there is an address retry record and there is not a
468 domain retry record; this implies that previous attempts to handle
469 the address had the retry_use_local_parts option turned on. We use
470 this as an approximation for the destination being like a local
471 delivery, as in LMTP.
473 I suspect this new check makes the old local delivery cutoff check
474 redundant, but I have not verified this so I left the code in place.
476 TF/02 Correct gecos expansion when From: is a prefix of the username.
478 Test 0254 submits a message to Exim with the header
482 When I ran the test suite under the user fanf2, Exim expanded
483 the header to contain my full name, whereas it should have added
484 a Resent-Sender: header. It erroneously treats any prefix of the
485 username as equal to the username.
487 This change corrects that bug.
489 GF/01 DCC debug and logging tidyup
490 Error conditions log to paniclog rather than rejectlog.
491 Debug lines prefixed by "DCC: " to remove any ambiguity.
493 TF/03 Avoid unnecessary rebuilds of lookup-related code.
495 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
496 Bug spotted by Jeremy Harris; was flawed since initial commit.
497 Would have resulted in OCSP responses post-SNI triggering an Exim
498 NULL dereference and crash.
500 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
502 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
503 Bug detection, analysis and fix by Samuel Thibault.
504 Bugzilla 1331, Debian bug #698092.
506 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
508 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
509 Server implementation by Todd Lyons, client by JH.
510 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
511 config variable "prdr_enable" controls whether the server
512 advertises the facility. If the client requests PRDR a new
513 acl_data_smtp_prdr ACL is called once for each recipient, after
514 the body content is received and before the acl_smtp_data ACL.
515 The client is controlled by bolth of: a hosts_try_prdr option
516 on the smtp transport, and the server advertisement.
517 Default client logging of deliveries and rejections involving
518 PRDR are flagged with the string "PRDR".
520 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
521 fclose(). Diagnosis by Todd Lyons.
523 PP/17 Update configure.default to handle IPv6 localhost better.
524 Patch by Alain Williams (plus minor tweaks).
527 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
528 This is now consistent with GnuTLS, and is now documented: the
529 previous undocumented portable approach to treating the option as
530 unset was to force an expansion failure. That still works, and
531 an empty string is now equivalent.
533 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
534 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
535 not performing validation itself.
537 PP/20 Added force_command boolean option to pipe transport.
538 Patch from Nick Koston, of cPanel Inc.
540 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
543 TF/04 Added udpsend ACL modifer and hexquote expansion operator
545 PP/21 Fix eximon continuous updating with timestamped log-files.
546 Broken in a format-string cleanup in 4.80, missed when I repaired the
547 other false fix of the same issue.
548 Report and fix from Heiko Schlichting.
551 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
552 Report from Prashanth Katuri.
554 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
555 It's SecureTransport, so affects any MacOS clients which use the
556 system-integrated TLS libraries, including email clients.
558 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
559 using a MIME ACL for non-SMTP local injection.
560 Report and assistance in diagnosis by Warren Baker.
562 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
564 JH/16 Fix comparisons for 64b. Bugzilla 1385.
566 TL/09 Add expansion variable $authenticated_fail_id to keep track of
567 last id that failed so it may be referenced in subsequent ACL's.
569 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
572 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
573 ldap library initialization, allowing self-signed CA's to be
574 used. Also properly sets require_cert option later in code by
575 using NULL (global ldap config) instead of ldap handle (per
576 session). Bug diagnosis and testing by alxgomz.
578 TL/12 Enhanced documentation in the ratelimit.pl script provided in
579 the src/util/ subdirectory.
581 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
582 renamed to Transport Post Delivery Action by Jeremy Harris, as
585 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
586 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
587 redis_servers = needs to be configured which will be used by the redis
588 lookup. Patch from Warren Baker, of The Packet Hub.
590 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
592 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
593 hostname or reverse DNS when processing a host list. Used suggestions
594 from multiple comments on this bug.
596 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
598 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
599 Missed a few lines, added it to make the runtest require no keyboard
602 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
603 contains upper case chars. Make router use caseful_local_part.
605 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
606 support when GnuTLS has been built with p11-kit.
612 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
614 This, or similar/improved, will also be change PP/11 of 4.82.
620 PP/01 Handle short writes when writing local log-files.
621 In practice, only affects FreeBSD (8 onwards).
622 Bugzilla 1053, with thanks to Dmitry Isaikin.
624 NM/01 Bugzilla 949 - Documentation tweak
626 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
629 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
631 PP/02 Implemented gsasl authenticator.
633 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
635 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
636 `pkg-config foo` for cflags/libs.
638 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
639 with rest of GSASL and with heimdal_gssapi.
641 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
642 `pkg-config foo` for cflags/libs for the TLS implementation.
644 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
645 properties get this fed in as external SSF. A number of robustness
646 and debugging improvements to the cyrus_sasl authenticator.
648 PP/08 cyrus_sasl server now expands the server_realm option.
650 PP/09 Bugzilla 1214 - Log authentication information in reject log.
651 Patch by Jeremy Harris.
653 PP/10 Added dbmjz lookup type.
655 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
657 PP/12 MAIL args handles TAB as well as SP, for better interop with
658 non-compliant senders.
659 Analysis and variant patch by Todd Lyons.
661 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
662 Bug report from Lars Müller <lars@samba.org> (via SUSE),
663 Patch from Dirk Mueller <dmueller@suse.com>
665 PP/13 tls_peerdn now print-escaped for spool files.
666 Observed some $tls_peerdn in wild which contained \n, which resulted
667 in spool file corruption.
669 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
670 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
671 or write after TLS renegotiation, which otherwise led to messages
674 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
675 as a tracking header (ie: a signed header comes before the signature).
676 Patch from Wolfgang Breyha.
678 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
679 comma-sep list; embedded commas doubled.
681 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
683 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
685 Report and patch from Dmitry Banschikov.
687 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
688 Removed SSL_clear() after SSL_new() which led to protocol negotiation
689 failures. We appear to now support TLS1.1+ with Exim.
691 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
692 lets Exim select keys and certificates based upon TLS SNI from client.
693 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
694 before an outbound SMTP session. New log_selector, +tls_sni.
696 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
697 NULL dereference. Report and patch from Alun Jones.
699 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
700 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
701 Not seeing resolver debug output on NetBSD, but suspect this is a
702 resolver implementation change.
704 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
705 Left warnings. Added "eximon gdb" invocation mode.
707 PP/21 Defaulting "accept_8bitmime" to true, not false.
709 PP/22 Added -bw for inetd wait mode support.
711 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
712 locate the relevant includes and libraries. Made this the default.
714 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
715 Bugzilla 1246, report and most of solution from Tomasz Kusy.
717 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
718 This may cause build issues on older platforms.
720 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
721 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
722 gnutls_require_mac & gnutls_require_protocols (no longer supported).
723 Added SNI support via GnuTLS too.
724 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
726 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
728 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
729 Applied second patch from Janne, implementing suggestion to default
730 multiple-strings-in-record handling to match SPF spec.
732 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
734 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
735 read-only, out of scope).
736 Patch from Wolfgang Breyha, report from Stuart Northfield.
738 PP/29 Fix three issues highlighted by clang analyser static analysis.
739 Only crash-plausible issue would require the Cambridge-specific
740 iplookup router and a misconfiguration.
741 Report from Marcin Mirosław.
743 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
745 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
746 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
747 As part of this, removing so much warning spew let me fix some minor
748 real issues in debug logging.
750 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
751 assignment on my part. Fixed.
753 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
754 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
755 Janne Snabb (who went above and beyond: thank you).
757 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
758 string otherwise requires a connection and a bunch more work and it's
759 relatively easy to get wrong. Should also expose TLS library linkage
762 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
763 64-bit ${eval} (JH/03).
765 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
766 GNU libc to support some of the 64-bit stuff, should not lead to
767 conflicts. Defined before os.h is pulled in, so if a given platform
768 needs to override this, it can.
770 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
771 protection layer was required, which is not implemented.
772 Bugzilla 1254, patch from Wolfgang Breyha.
774 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
775 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
776 tls_dhparam take prime identifiers. Also unbreak combination of
777 OpenSSL+DH_params+TLSSNI.
779 PP/39 Disable SSLv2 by default in OpenSSL support.
785 PP/01 Solaris build fix for Oracle's LDAP libraries.
786 Bugzilla 1109, patch from Stephen Usher.
788 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
790 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
793 TF/02 Fix a couple more cases where we did not log the error message
794 when unlink() failed. See also change 4.74-TF/03.
796 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
797 lock up or crash if it happened to be inside a call to libc when it
798 got a SIGUSR1 from exiwhat.
800 The SIGUSR1 handler appends the current process status to the process
801 log which is later printed by exiwhat. It used to use the general
802 purpose logging code to do this, but several functions it calls are
803 not safe for signals.
805 The new output code in the SIGUSR1 handler is specific to the process
806 log, and simple enough that it's easy to inspect for signal safety.
807 Removing some special cases also simplifies the general logging code.
808 Removing the spurious timestamps from the process log simplifies
811 TF/04 Improved ratelimit ACL condition.
813 The /noupdate option has been deprecated in favour of /readonly which
814 has clearer semantics. The /leaky, /strict, and /readonly update modes
815 are mutually exclusive. The update mode is no longer included in the
816 database key; it just determines when the database is updated. (This
817 means that when you upgrde Exim will forget old rate measurements.)
819 Exim now checks that the per_* options are used with an update mode that
820 makes sense for the current ACL. For example, when Exim is processing a
821 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
822 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
823 must specify per_mail/readonly. If you omit the update mode it defaults to
824 /leaky where that makes sense (as before) or /readonly where required.
826 The /noupdate option is now undocumented but still supported for
827 backwards compatibility. It is equivalent to /readonly except that in
828 ACLs where /readonly is required you may specify /leaky/noupdate or
829 /strict/noupdate which are treated the same as /readonly.
831 A useful new feature is the /count= option. This is a generalization
832 of the per_byte option, so that you can measure the throughput of other
833 aggregate values. For example, the per_byte option is now equivalent
834 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
836 The per_rcpt option has been generalized using the /count= mechanism
837 (though it's more complicated than the per_byte equivalence). When it is
838 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
839 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
840 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
841 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
842 Note that using per_rcpt with a non-readonly update mode in more than
843 one ACL will cause the recipients to be double-counted. (The per_mail
844 and per_byte options don't have this problem.)
846 The handling of very low rates has changed slightly. If the computed rate
847 is less than the event's count (usually one) then this event is the first
848 after a long gap. In this case the rate is set to the same as this event's
849 count, so that the first message of a spam run is counted properly.
851 The major new feature is a mechanism for counting the rate of unique
852 events. The new per_addr option counts the number of different
853 recipients that someone has sent messages to in the last time period. It
854 behaves like per_rcpt if all the recipient addresses are different, but
855 duplicate recipient addresses do not increase the measured rate. Like
856 the /count= option this is a general mechanism, so the per_addr option
857 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
858 example, measure the rate that a client uses different sender addresses
859 with the options per_mail/unique=$sender_address. There are further
860 details in the main documentation.
862 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
864 TF/06 Removed a few PCRE remnants.
866 TF/07 Automatically extract Exim's version number from tags in the git
867 repository when doing development or release builds.
869 PP/02 Raise smtp_cmd_buffer_size to 16kB.
870 Bugzilla 879. Patch from Paul Fisher.
872 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
873 Heavily based on revision 40f9a89a from Simon Arlott's tree.
876 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
878 PP/05 Variable $av_failed, true if the AV scanner deferred.
879 Bugzilla 1078. Patch from John Horne.
881 PP/06 Stop make process more reliably on build failure.
882 Bugzilla 1087. Patch from Heiko Schlittermann.
884 PP/07 Make maildir_use_size_file an _expandable_ boolean.
885 Bugzilla 1089. Patch from Heiko Schlittermann.
887 PP/08 Handle ${run} returning more data than OS pipe buffer size.
888 Bugzilla 1131. Patch from Holger Weiß.
890 PP/09 Handle IPv6 addresses with SPF.
891 Bugzilla 860. Patch from Wolfgang Breyha.
893 PP/10 GnuTLS: support TLS 1.2 & 1.1.
895 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
898 PP/11 match_* no longer expand right-hand-side by default.
899 New compile-time build option, EXPAND_LISTMATCH_RHS.
900 New expansion conditions, "inlist", "inlisti".
902 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
904 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
906 PP/14 fix log_write() format string regression from TF/03.
907 Bugzilla 1152. Patch from Dmitry Isaikin.
913 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
915 PP/02 Harmonised TLS library version reporting; only show if debugging.
916 Layout now matches that introduced for other libraries in 4.74 PP/03.
918 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
920 PP/04 New "dns_use_edns0" global option.
922 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
925 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
926 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
928 TK/01 Updated PolarSSL code to 0.14.2.
929 Bugzilla 1097. Patch from Andreas Metzler.
931 PP/07 Catch divide-by-zero in ${eval:...}.
934 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
937 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
938 format-string attack -- SECURITY: remote arbitrary code execution.
940 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
941 time unintentionally subject to list matching rules, letting the header
942 cause arbitrary Exim lookups (of items which can occur in lists, *not*
943 arbitrary string expansion). This allowed for information disclosure.
945 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
946 INT_MIN/-1 -- value coerced to INT_MAX.
952 NM/01 Workround for PCRE version dependency in version reporting
955 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
956 This fixes portability to compilers other than gcc, notably
957 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
959 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
960 makefiles for portability to HP-UX and POSIX correctness.
962 PP/01 Permit LOOKUP_foo enabling on the make command-line.
963 Also via indented variable definition in the Makefile.
964 (Debugging by Oliver Heesakkers).
966 PP/02 Restore caching of spamd results with expanded spamd_address.
967 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
969 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
970 Improves build reliability. Fix from: Frank Elsner
972 NM/02 Fix wide character breakage in the rfc2047 coding
973 Fixes bug 1064. Patch from Andrey N. Oktyabrski
975 NM/03 Allow underscore in dnslist lookups
976 Fixes bug 1026. Patch from Graeme Fowler
978 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
979 Code patches from Adam Ciarcinski of NetBSD.
981 NM/04 Fixed exiqgrep to cope with mailq missing size issue
984 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
985 is logged, to avoid truncation. Patch from John Horne.
987 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
988 Patch from Jakob Hirsch.
990 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
991 SQL string expansion failure details.
992 Patch from Andrey Oktyabrski.
994 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
995 Patch from Simon Arlott.
997 PP/09 New lookups functionality failed to compile on old gcc which rejects
998 extern declarations in function scope.
999 Patch from Oliver Fleischmann
1001 PP/10 Use sig_atomic_t for flags set from signal handlers.
1002 Check getgroups() return and improve debugging.
1003 Fixed developed for diagnosis in bug 927 (which turned out to be
1006 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1007 Patch from Mark Zealey.
1009 PP/12 Bugzilla 1056: Improved spamd server selection.
1010 Patch from Mark Zealey.
1012 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1013 Based on patch from Heiko Schlittermann.
1015 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1016 Patch from Uwe Doering, sign-off by Michael Haardt.
1018 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1019 variable declaration deep within a block. Bug and patch from
1022 PP/15 lookups-Makefile IRIX compatibilty coercion.
1024 PP/16 Make DISABLE_DKIM build knob functional.
1026 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1027 Patch by Simon Arlott
1029 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1030 variable argument macros. Our copy now differs from upstream.
1036 TF/01 Failure to get a lock on a hints database can have serious
1037 consequences so log it to the panic log.
1039 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1040 controlled using the smtp_confirmation log selector.
1042 TF/03 Include the error message when we fail to unlink a spool file.
1044 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1045 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1046 for maintaining out-of-tree patches for some time.
1048 PP/01 Bugzilla 139: Documentation and portability issues.
1049 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1050 Handle per-OS dynamic-module compilation flags.
1052 PP/02 Let /dev/null have normal permissions.
1053 The 4.73 fixes were a little too stringent and complained about the
1054 permissions on /dev/null. Exempt it from some checks.
1055 Reported by Andreas M. Kirchwitz.
1057 PP/03 Report version information for many libraries, including
1058 Exim version information for dynamically loaded libraries. Created
1059 version.h, now support a version extension string for distributors
1060 who patch heavily. Dynamic module ABI change.
1062 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1063 privilege escalation vulnerability whereby the Exim run-time user
1064 can cause root to append content of the attacker's choosing to
1067 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1070 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1071 If dropping privileges for untrusted macros, we disabled normal logging
1072 on the basis that it would fail; for the Exim run-time user, this is not
1073 the case, and it resulted in successful deliveries going unlogged.
1074 Fixed. Reported by Andreas Metzler.
1080 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1081 only prepend for the Resent-* case. Fixes regression introduced in
1082 Exim 4.70 by NM/22 for Bugzilla 607.
1084 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1085 increasing numbers of administrators be bitten by this.
1087 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1089 PP/03 Bugzilla 994: added openssl_options main configuration option.
1091 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1093 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1095 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1097 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1098 without a peer certificate, leading to a segfault because of an
1099 assumption that peers always have certificates. Be a little more
1100 paranoid. Problem reported by Martin Tscholak.
1102 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1103 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1104 NB: ClamAV planning to remove STREAM in "middle of 2010".
1105 CL also introduces -bmalware, various -d+acl logging additions and
1106 more caution in buffer sizes.
1108 PP/09 Implemented reverse_ip expansion operator.
1110 PP/10 Bugzilla 937: provide a "debug" ACL control.
1112 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1114 PP/12 Bugzilla 973: Implement --version.
1116 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1118 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1120 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1122 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1123 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1124 ignore trailing whitespace.
1126 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1128 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1131 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1132 Notification from Dr Andrew Aitchison.
1134 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1135 ExtendedDetectionInfo response format.
1136 Notification from John Horne.
1138 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1141 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1142 XSL and documented dependency on system catalogs, with examples of how
1145 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1148 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1149 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1150 configuration file which is writeable by the Exim user or group.
1152 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1153 of configuration files to cover files specified with the -C option if
1154 they are going to be used with root privileges, not just the default
1157 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1158 option (effectively making it always true).
1160 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1161 files to be used while preserving root privileges.
1163 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1164 that rogue child processes cannot use them.
1166 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1167 run-time user, instead of root.
1169 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1170 Exim run-time user without dropping privileges.
1172 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1173 result string, instead of calling string_vformat() twice with the same
1176 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1177 for other users. Others should always drop root privileges if they use
1178 -C on the command line, even for a whitelisted configure file.
1180 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1182 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1188 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1189 $data_path, and $header_path variables; fixed documentation bugs and
1192 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1193 exipick to access non-standard spools, including the "frozen" queue
1196 NM/01 Bugzilla 965: Support mysql stored procedures.
1197 Patch from Alain Williams
1199 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1201 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1202 Patch from Andreas Metzler
1204 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1205 Patch from Kirill Miazine
1207 NM/05 Bugzilla 671: Added umask to procmail example.
1209 JJ/03 installed exipick 20100323.0, fixing doc bug
1211 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1212 directory. Notification and patch from Dan Rosenberg.
1214 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1216 TK/02 Improve log output when DKIM signing operation fails.
1218 MH/01 Treat the transport option dkim_domain as a colon separated
1219 list, not as a single string, and sign the message with each element,
1220 omitting multiple occurences of the same signer.
1222 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1223 Bugzilla 985, 986. Patch by Simon Arlott
1225 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1226 Patch by Simon Arlott
1228 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1229 MBX locking. Notification from Dan Rosenberg.
1235 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1237 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1239 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1241 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1243 NM/04 Bugzilla 915: Flag broken perl installation during build.
1249 TK/01 Added patch by Johannes Berg that expands the main option
1250 "spamd_address" if it starts with a dollar sign.
1252 TK/02 Write list of recipients to X-Envelope-Sender header when building
1253 the mbox-format spool file for content scanning (suggested by Jakob
1256 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1257 (http://www.dcc-servers.net/) support via dccifd. Activated by
1258 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1260 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1261 by Mark Daniel Reidel <mr@df.eu>.
1263 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1264 When building exim an external PCRE library is now needed -
1265 PCRE is a system library on the majority of modern systems.
1266 See entry on PCRE_LIBS in EDITME file.
1268 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1269 conversation. Added nologin parameter to request.
1270 Patch contributed by Kirill Miazine.
1272 TF/01 Do not log submission mode rewrites if they do not change the address.
1274 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1276 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1277 log files in place. Contributed by Roberto Lima.
1279 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1281 TF/03 Bugzilla 615: When checking the local_parts router precondition
1282 after a local_part_suffix or local_part_prefix option, Exim now
1283 does not use the address's named list lookup cache, since this
1284 contains cached lookups for the whole local part.
1286 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1287 Robert Millan. Documentation is in experimental-spec.txt.
1289 TF/04 Bugzilla 668: Fix parallel build (make -j).
1291 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1293 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1294 Patch provided by Jan Srzednicki.
1296 TF/05 Leading white space used to be stripped from $spam_report which
1297 wrecked the formatting. Now it is preserved.
1299 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1300 that they are available at delivery time.
1302 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1304 TF/08 TLS error reporting now respects the incoming_interface and
1305 incoming_port log selectors.
1307 TF/09 Produce a more useful error message if an SMTP transport's hosts
1308 setting expands to an empty string.
1310 NM/06 Bugzilla 744: EXPN did not work under TLS.
1311 Patch provided by Phil Pennock.
1313 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1314 Patch provided by Richard Godbee.
1316 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1317 acl_smtp_notquit, added index entry.
1319 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1320 Patch provided by Eugene Bujak.
1322 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1323 accept(). Patch provided by Maxim Dounin.
1325 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1326 Patch provided by Phil Pennock.
1328 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1330 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1331 Patch provided by Brad "anomie" Jorsch.
1333 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1334 Patch provided by Dean Brooks.
1336 TK/05 Add native DKIM support (does not depend on external libraries).
1338 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1339 Patch provided by Graeme Fowler.
1341 NM/16 Bugzilla 851: Documentation example syntax fix.
1343 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1345 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1348 NM/19 Bugzilla 745: TLS version reporting.
1349 Patch provided by Phil Pennock.
1351 NM/20 Bugzilla 167: bool: condition support.
1352 Patch provided by Phil Pennock.
1354 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1355 clients. Patch provided by Phil Pennock.
1357 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1358 Patch provided by Brad "anomie" Jorsch.
1360 NM/23 Bugzilla 687: Fix misparses in eximstats.
1361 Patch provided by Heiko Schlittermann.
1363 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1364 Patch provided by Heiko Schlittermann.
1366 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1367 plus update to original patch.
1369 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1371 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1372 Patch provided by David Brownlee.
1374 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1376 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1378 NM/30 Bugzilla 888: TLS documentation bugfixes.
1380 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1382 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1383 Unlike the original bugzilla I have changed all shell scripts in src tree.
1385 NM/33 Bugzilla 898: Transport filter timeout fix.
1386 Patch by Todd Rinaldo.
1388 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1389 Patch by Serge Demonchaux.
1391 NM/35 Bugzilla 39: Base64 decode bug fixes.
1392 Patch by Jakob Hirsch.
1394 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1396 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1398 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1400 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1406 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1407 ALT-N's libdkim that I have put here:
1408 http://duncanthrax.net/exim-experimental/
1410 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1411 were called 'true' and it seems that C99 defines that as a reserved
1412 keyword to be used with 'bool' variable types. That means you could
1413 not include C99-style headers which use bools without triggering
1414 build errors in sieve.c.
1416 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1417 as mailq or other aliases. Changed the --help handling significantly
1418 to do whats expected. exim_usage() emits usage/help information.
1420 SC/01 Added the -bylocaldomain option to eximstats.
1422 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1424 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1426 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1432 PH/01 Another patch from the Sieve maintainer.
1434 PH/02 When an IPv6 address is converted to a string for single-key lookup
1435 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1436 dots are used instead of colons so that keys in lsearch files need not
1437 contain colons. This was done some time before quoting was made available
1438 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1439 (notated using the quote facility) so as to distinguish them from IPv4
1440 keys. This meant that lookups for IP addresses in host lists did not work
1441 for iplsearch lookups.
1443 This has been fixed by arranging for IPv6 addresses to be expressed with
1444 colons if the lookup type is iplsearch. This is not incompatible, because
1445 previously such lookups could never work.
1447 The situation is now rather anomolous, since one *can* have colons in
1448 ordinary lsearch keys. However, making the change in all cases is
1449 incompatible and would probably break a number of configurations.
1451 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1454 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1455 conversion specification without a maximum field width, thereby enabling
1456 a rogue spamd server to cause a buffer overflow. While nobody in their
1457 right mind would setup Exim to query an untrusted spamd server, an
1458 attacker that gains access to a server running spamd could potentially
1459 exploit this vulnerability to run arbitrary code as the Exim user.
1461 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1462 $primary_hostname instead of what libspf2 thinks the hosts name is.
1464 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1465 a directory entry by the name of the lookup key. Previously, if a
1466 symlink pointed to a non-existing file or a file in a directory that
1467 Exim lacked permissions to read, a lookup for a key matching that
1468 symlink would fail. Now it is enough that a matching directory entry
1469 exists, symlink or not. (Bugzilla 503.)
1471 PH/03 The body_linecount and body_zerocount variables are now exported in the
1474 PH/04 Added the $dnslist_matched variable.
1476 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1477 This means they are set thereafter only if the connection becomes
1480 PH/06 Added the client_condition to authenticators so that some can be skipped
1481 by clients under certain conditions.
1483 PH/07 The error message for a badly-placed control=no_multiline_responses left
1484 "_responses" off the end of the name.
1486 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1488 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1489 (without spaces) instead of just copying the configuration text.
1491 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1493 PH/11 Added $max_received_linelength.
1495 PH/12 Added +ignore_defer and +include_defer to host lists.
1497 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1498 way in which PCRE > 7.0 is built.
1500 PH/14 Implemented queue_only_load_latch.
1502 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1503 MAIL command. The effect was to mangle the value on 64-bit systems.
1505 PH/16 Another patch from the Sieve maintainer.
1507 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1509 PH/18 If a system quota error occurred while trying to create the file for
1510 a maildir delivery, the message "Mailbox is full" was not appended to the
1511 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1512 only to a quota excession during the actual writing of the file.
1514 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1515 characters?) which causes problems in log lines. The DN values are now
1516 passed through string_printing() before being added to log lines.
1518 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1519 and InterBase are left for another time.)
1521 PH/21 Added message_body_newlines option.
1523 PH/22 Guard against possible overflow in moan_check_errorcopy().
1525 PH/23 POSIX allows open() to be a macro; guard against that.
1527 PH/24 If the recipient of an error message contained an @ in the local part
1528 (suitably quoted, of course), incorrect values were put in $domain and
1529 $local_part during the evaluation of errors_copy.
1535 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1536 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1539 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1540 issue a MAIL command.
1542 PH/02 In an ACL statement such as
1544 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1546 if a client was not listed at all, or was listed with a value other than
1547 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1548 the condition was not true (as it should be), so access was not denied.
1549 The bug was that the ! inversion was incorrectly passed on to the second
1550 item. This has been fixed.
1552 PH/03 Added additional dnslists conditions == and =& which are different from
1553 = and & when the dns lookup returns more than one IP address.
1555 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1556 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1558 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1559 FSYNC, which compiles an option called disable_fsync that allows for
1560 bypassing fsync(). The documentation is heavily laced with warnings.
1562 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1564 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1565 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1566 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1567 including adding "make clean"; (3) Added -fPIC when compiling the test
1568 dynamically loaded module, to get rid of a warning.
1570 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1571 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1572 = 0s. The bug is otherwise harmless.
1574 PH/07 There was a bug in the dovecot authenticator such that the value of
1575 $auth1 could be overwritten, and so not correctly preserved, after a
1576 successful authentication. This usually meant that the value preserved by
1577 the server_setid option was incorrect.
1579 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1581 PH/09 Installed PCRE release 7.0.
1583 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1584 run for batched SMTP input. It is now run at the start of every message
1585 in the batch. While fixing this I discovered that the process information
1586 (output by running exiwhat) was not always getting set for -bs and -bS
1587 input. This is fixed, and it now also says "batched" for BSMTP.
1589 PH/11 Added control=no_pipelining.
1591 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1592 patch, slightly modified), and move the expansion of helo_data till after
1593 the connection is made in the smtp transport (so it can use these
1596 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1598 PH/14 Added log_selector = +pid.
1600 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1602 PH/16 Add ${if forany and ${if forall.
1604 PH/17 Added dsn_from option to vary the From: line in DSNs.
1606 PH/18 Flush SMTP output before performing a callout, unless control =
1607 no_callout_flush is set.
1609 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1610 was true (the default) a successful delivery failed to delete the retry
1611 item, thus causing premature timeout of the address. The bug is now
1614 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1616 PH/21 Long custom messages for fakedefer and fakereject are now split up
1617 into multiline reponses in the same way that messages for "deny" and
1618 other ACL rejections are.
1620 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1621 with slight modification.
1623 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1624 draft, changing the syntax and factoring some duplicate code".
1626 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1627 for deliveries of the second and subsequent messages over the same SMTP
1630 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1631 ${reduce, with only minor "tidies".
1633 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1635 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1636 expansion side effects.
1638 PH/27 When a message times out after an over-quota error from an Exim-imposed
1639 quota, the bounce message says "mailbox is full". This message was not
1640 being given when it was a system quota that was exceeded. It now should
1643 MH/03 Made $recipients available in local_scan(). local_scan() already has
1644 better access to the recipient list through recipients_list[], but
1645 $recipients can be useful in postmaster-provided expansion strings.
1647 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1648 in the case of a MAIL command with additional options following the
1649 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1650 were accidentally chopped off.
1652 PH/29 SMTP synchronization checks are implemented when a command is read -
1653 there is a check that no more input is waiting when there shouldn't be
1654 any. However, for some commands, a delay in an ACL can mean that it is
1655 some time before the response is written. In this time, more input might
1656 arrive, invalidly. So now there are extra checks after an ACL has run for
1657 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1658 pipelining has not been advertised.
1660 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1662 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1663 returned in the SMTP response when smtp_return_error_details was false.
1664 This has been fixed.
1666 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1667 instead of the C I/O that was originally supplied, because problems were
1668 reported on Solaris.
1670 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1671 Exim which did not show up earlier: it was assuming that a call to
1672 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1673 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1674 was a macro that became an assignment, so it seemed to work. This has
1675 changed to a proper function call with a void return, hence the compile
1676 error. Exim's code has been fixed.
1678 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1681 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1683 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1685 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1686 because it turns out to be too big or there is a timeout) there is no
1687 "Completed" line in the log. When some messages of this type were
1688 selected by exigrep, they were listed as "not completed". Others were
1689 picked up by some special patterns. I have improved the selection
1690 criteria to be more general.
1692 PH/38 The host_find_failed option in the manualroute router can now be set
1693 to "ignore", to completely ignore a host whose IP address cannot be
1694 found. If all hosts are ignored, the behaviour is controlled by the new
1695 host_all_ignored option.
1697 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1698 homing or because of multiple MX records with /mx) generated more than
1699 one IP address, and the following item turned out to be the local host,
1700 all the secondary addresses of the first item were incorrectly removed
1701 from the list, along with the local host and any following hosts (which
1702 is what is supposed to happen).
1704 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1705 whoever called Exim into the -H file. In the case of the daemon it was
1706 behaving confusingly. When first started, it used values for whoever
1707 started the daemon, but after a SIGHUP it used the Exim user (because it
1708 calls itself on a restart). I have changed the code so that it now always
1711 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1712 message are rejected with the same error (e.g. no authentication or bad
1713 sender address), and a DATA command is nevertheless sent (as can happen
1714 with PIPELINING or a stupid MUA), the error message that was given to the
1715 RCPT commands is included in the rejection of the DATA command. This is
1716 intended to be helpful for MUAs that show only the final error to their
1719 PH/42 Another patch from the Sieve maintainer.
1721 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1722 Eximstats - Fixed some broken HTML links and added missing column headers
1724 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1725 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1727 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1729 PH/43 Yet another patch from the Sieve maintainer.
1731 PH/44 I found a way to check for a TCP/IP connection going away before sending
1732 the response to the final '.' that terminates a message, but only in the
1733 case where the client has not sent further data following the '.'
1734 (unfortunately, this is allowed). However, in many cases there won't be
1735 any further data because there won't be any more messages to send. A call
1736 to select() can be used: if it shows that the input is "ready", there is
1737 either input waiting, or the socket has been closed. An attempt to read
1738 the next input character can distinguish the two cases. Previously, Exim
1739 would have sent an OK response which the client would never have see.
1740 This could lead to message repetition. This fix should cure that, at
1741 least in a lot of common cases.
1743 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1744 advertised in response to EHLO.
1750 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1751 fixed by 4.65/MH/01 (is this a record?) are fixed:
1753 (i) An empty string was always treated as zero by the numeric comparison
1754 operators. This behaviour has been restored.
1756 (ii) It is documented that the numeric comparison operators always treat
1757 their arguments as decimal numbers. This was broken in that numbers
1758 starting with 0 were being interpreted as octal.
1760 While fixing these problems I realized that there was another issue that
1761 hadn't been noticed. Values of message_size_limit (both the global option
1762 and the transport option) were treated as octal if they started with 0.
1763 The documentation was vague. These values are now always treated as
1764 decimal, and I will make that clear in the documentation.
1770 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1771 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1774 MH/01 Don't check that the operands of numeric comparison operators are
1775 integers when their expansion is in "skipping" mode (fixes bug
1776 introduced by 4.64-PH/07).
1778 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1779 child addresses, Exim now panics and dies. Previously, because the count
1780 is held in a short int, deliveries were likely to be lost. As such a
1781 large number of recipients for a single message is ridiculous
1782 (performance will be very, very poor), I have chosen to impose a limit
1783 rather than extend the field.
1789 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1790 leftover -K file (the existence of which was triggered by #402).
1791 While we were at it, introduced process PID as part of the -K
1792 filename. This should rule out race conditions when creating
1795 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1796 processing considerably. Previous code took too long for large mails,
1797 triggering a timeout which in turn triggers #401.
1799 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1800 in the DK code in transports.c. sendfile() is not really portable,
1801 hence the _LINUX specificness.
1803 TF/01 In the add_headers option to the mail command in an Exim filter,
1804 there was a bug that Exim would claim a syntax error in any
1805 header after the first one which had an odd number of characters
1808 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1809 callout verification, Exim cached a "reject" for the entire domain. This
1810 is correct for most verifications, but it is not correct for a recipient
1811 verification with use_sender or use_postmaster set, because in that case
1812 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1813 case of MAIL FROM:<> rejection from other early rejections (e.g.
1814 rejection of HELO). When verifying a recipient using a non-null MAIL
1815 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1816 Whatever the result of the callout, the value of the domain cache is
1817 left unchanged (for any other kind of callout, getting as far as trying
1818 RCPT means that the domain itself is ok).
1820 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1823 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1824 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1825 session, the daemon ignores ECONNECT errors and logs others; it now
1826 ignores EPIPE as well.
1828 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1829 (quoted-printable decoding).
1831 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1832 later the small subsequent patch to fix an introduced bug.
1834 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1836 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1838 PH/08 An error is now given if message_size_limit is specified negative.
1840 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1841 to be given (somewhat) arbitrary names.
1843 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1846 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1847 miscellaneous code fixes
1849 PH/10 Added the log_reject_target ACL modifier to specify where to log
1852 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1853 hostname. This is wrong, because it relates to the incoming message (and
1854 probably the interface on which it is arriving) and not to the outgoing
1855 callout (which could be using a different interface). This has been
1856 changed to use the value of the helo_data option from the smtp transport
1857 instead - this is what is used when a message is actually being sent. If
1858 there is no remote transport (possible with a router that sets up host
1859 addresses), $smtp_active_hostname is used.
1861 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1862 tweaks were necessary in order to get it to work (see also 21 below):
1863 (a) The code assumed that strncpy() returns a negative number on buffer
1864 overflow, which isn't the case. Replaced with Exim's string_format()
1866 (b) There were several signed/unsigned issues. I just did the minimum
1867 hacking in of casts. There is scope for a larger refactoring.
1868 (c) The code used strcasecmp() which is not a standard C function.
1869 Replaced with Exim's strcmpic() function.
1870 (d) The code set only $1; it now sets $auth1 as well.
1871 (e) A simple test gave the error "authentication client didn't specify
1872 service in request". It would seem that Dovecot has changed its
1873 interface. Fortunately there's a specification; I followed it and
1874 changed what the client sends and it appears to be working now.
1876 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1879 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1880 address A is aliased to B and C, where B exists and C does not. Without
1881 -v the output is "A verified" because verification stops after a
1882 successful redirection if more than one address is generated. However,
1883 with -v the child addresses are also verified. Exim was outputting "A
1884 failed to verify" and then showing the successful verification for C,
1885 with its parentage. It now outputs "B failed to verify", showing B's
1886 parentage before showing the successful verification of C.
1888 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1889 look up a TXT record in a specific list after matching in a combined
1892 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1893 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1894 they consult the DNS. I had assumed they would set it the way they
1895 wanted; and indeed my experiments on Linux seem to show that in some
1896 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1897 To be on the safe side, however, I have now made the interface to
1898 host_find_byname() similar to host_find_bydns(), with an argument
1899 containing the DNS resolver options. The host_find_byname() function now
1900 sets these options at its start, just as host_find_bydns() does. The smtp
1901 transport options dns_qualify_single and dns_search_parents are passed to
1902 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1903 of host_find_byname() use the default settings of RES_DEFNAMES
1904 (qualify_single) but not RES_DNSRCH (search_parents).
1906 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1907 spool_read_header() do less string testing, by means of a preliminary
1908 switch on the second character of optional "-foo" lines. (This is
1909 overdue, caused by the large number of possibilities that now exist.
1910 Originally there were few.) While I was there, I also converted the
1911 str(n)cmp tests so they don't re-test the leading "-" and the first
1912 character, in the hope this might squeeze out yet more improvement.
1914 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1915 flag allowing group syntax was set by the header_syntax check but not
1916 turned off, possible causing trouble later; (2) The flag was not being
1917 set at all for the header_verify test, causing "group"-style headers to
1918 be rejected. I have now set it in this case, and also caused header_
1919 verify to ignore an empty address taken from a group. While doing this, I
1920 came across some other cases where the code for allowing group syntax
1921 while scanning a header line wasn't quite right (mostly, not resetting
1922 the flag correctly in the right place). These bugs could have caused
1923 trouble for malformed header lines. I hope it is now all correct.
1925 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1926 with the "reply" argument non-NULL. The code, however (which originally
1927 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1928 but it didn't always do it. This confused somebody who was copying the
1929 code for some other use. I have removed all the tests.
1931 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1932 feature that was used to support insecure browsers during the U.S. crypto
1933 embargo. It requires special client support, and Exim is probably the
1934 only MTA that supported it -- and would never use it because real RSA is
1935 always available. This code has been removed, because it had the bad
1936 effect of slowing Exim down by computing (never used) parameters for the
1937 RSA_EXPORT functionality.
1939 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1940 authenticator to fail if there's a tab character in the incoming data
1941 (there should never be unless someone is messing about, as it's supposed
1942 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1943 if the connection is using TLS or if the remote IP is the same as the
1944 local IP, and the "valid-client-cert option" if a client certificate has
1947 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1948 authenticators. This can be used for authorization after authentication
1949 succeeds. (In the case of plaintext, it servers for both authentication
1952 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1953 if any retry times were supplied.
1955 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1956 connection, where there is no client IP address to check. In this
1957 situation, the verify now always succeeds.
1959 PH/25 Applied John Jetmore's -Mset patch.
1961 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1963 PH/27 In a string expansion for a processed (not raw) header when multiple
1964 headers of the same name were present, leading whitespace was being
1965 removed from all of them, but trailing whitespace was being removed only
1966 from the last one. Now trailing whitespace is removed from each header
1967 before concatenation. Completely empty headers in a concatenation (as
1968 before) are ignored.
1970 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1971 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1973 PH/29 [Removed. This was a change that I later backed out, and forgot to
1974 correct the ChangeLog entry (that I had efficiently created) before
1975 committing the later change.]
1977 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1978 address errors (4xx response to RCPT) over the same connection as other
1979 messages routed to the same hosts. Such deliveries are always "forced",
1980 so retry times are not inspected. This resulted in far too many retries
1981 for the affected addresses. The effect occurred only when there were more
1982 hosts than the hosts_max_try setting in the smtp transport when it had
1983 the 4xx errors. Those hosts that it had tried were not added to the list
1984 of hosts for which the message was waiting, so if all were tried, there
1985 was no problem. Two fixes have been applied:
1987 (i) If there are any address or message errors in an SMTP delivery, none
1988 of the hosts (tried or untried) are now added to the list of hosts
1989 for which the message is waiting, so the message should not be a
1990 candidate for sending over the same connection that was used for a
1991 successful delivery of some other message. This seems entirely
1992 reasonable: after all the message is NOT "waiting for some host".
1993 This is so "obvious" that I'm not sure why it wasn't done
1994 previously. Hope I haven't missed anything, but it can't do any
1995 harm, as the worst effect is to miss an optimization.
1997 (ii) If, despite (i), such a delivery is accidentally attempted, the
1998 routing retry time is respected, so at least it doesn't keep
1999 hammering the server.
2001 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2002 in ${readsocket because some servers need this prod.
2004 PH/32 Added some extra debug output when updating a wait-xxx database.
2006 PH/33 The hint "could be header name not terminated by colon", which has been
2007 given for certain expansion errors for a long time, was not being given
2008 for the ${if def:h_colon_omitted{... case.
2010 PH/34 The spec says: "With one important exception, whenever a domain list is
2011 being scanned, $domain contains the subject domain." There was at least
2012 one case where this was not true.
2014 PH/35 The error "getsockname() failed: connection reset by peer" was being
2015 written to the panic log as well as the main log, but it isn't really
2016 panic-worthy as it just means the connection died rather early on. I have
2017 removed the panic log writing for the ECONNRESET error when getsockname()
2020 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2021 runs only) independently of the message's sender address. This meant
2022 that, if the 4xx error was in fact related to the sender, a different
2023 message to the same recipient with a different sender could confuse
2024 things. In particualar, this can happen when sending to a greylisting
2025 server, but other circumstances could also provoke similar problems.
2026 I have changed the default so that the retry time for these errors is now
2027 based a combination of the sender and recipient addresses. This change
2028 can be overridden by setting address_retry_include_sender=false in the
2031 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2032 remote server are returned as part of bounce messages. This was not
2033 happening for LMTP over a pipe (the lmtp transport), but now it is the
2034 same for both kinds of LMTP.
2036 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2037 in header lines that were in fact CNAMEs. This is no longer the case.
2039 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2040 and queue runs started by the daemon processed all messages. This has
2041 been fixed so that -R and -S can now usefully be given with -q<time>.
2043 PH/40 Import PCRE release 6.7 (fixes some bugs).
2045 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2047 PH/42 Give an error if -q is specified more than once.
2049 PH/43 Renamed the variables $interface_address and $interface_port as
2050 $received_ip_address and $received_port, to make it clear that these
2051 values apply to message reception, and not to the outgoing interface when
2052 a message is delivered. (The old names remain recognized, of course.)
2054 PH/44 There was no timeout on the connect() call when using a Unix domain
2055 socket in the ${readsocket expansion. There now is.
2057 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2058 be meaningful with "accept".
2060 SC/01 Eximstats V1.43
2061 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2063 SC/02 Eximstats V1.44
2064 Use a glob alias rather than an array ref in the generated
2065 parser. This improves both readability and performance.
2067 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2068 Collect SpamAssassin and rejection statistics.
2069 Don't display local sender or destination tables unless
2070 there is data to show.
2071 Added average volumes into the top table text output.
2073 SC/04 Eximstats V1.46
2074 Collect data on the number of addresses (recipients)
2075 as well as the number of messages.
2077 SC/05 Eximstats V1.47
2078 Added 'Message too big' to the list of mail rejection
2079 reasons (thanks to Marco Gaiarin).
2081 SC/06 Eximstats V1.48
2082 Mainlog lines which have GMT offsets and are too short to
2083 have a flag are now skipped.
2085 SC/07 Eximstats V1.49 (Alain Williams)
2086 Added the -emptyok flag.
2088 SC/08 Eximstats V1.50
2089 Fixes for obtaining the IP address from reject messages.
2091 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2092 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2093 whitesspace changes from 4.64-PH/27
2095 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2098 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2099 are found, allow negative numbers in numeric criteria)
2101 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2103 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2104 to match changes made in 4.64-PH/43
2106 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2108 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2109 "domain" is actually an IP address, and force a failure. This locks out
2110 those revolvers/nameservers that support "A-for-A" lookups, in
2111 contravention of the specifications.
2113 PH/48 When a host name was looked up from an IP address, and the subsequent
2114 forward lookup of the name timed out, the host name was left in
2115 $sender_host_name, contrary to the specification.
2117 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2118 restricted to single-key lookups, Exim was not diagnosing an error if
2119 * or *@ was used with a query-style lookup.
2121 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2123 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2124 long ago, but noone interested enough thought of it. Let's just say that
2125 the "1.1" means that there are some new functions that weren't there at
2126 some point in the past.
2128 PH/51 Error processing for expansion failure of helo_data from an smtp
2129 transport during callout processing was broken.
2131 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2132 tested/used via the -bh/-bhc/-bs options.
2134 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2135 bug, fixed in subsequent PCRE releases).
2137 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2138 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2140 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2146 SC/01 Use a glob alias rather than an array ref in eximstats generated
2147 parser. This improves both readability and performance.
2149 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2150 Don't display local sender or destination tables in eximstats unless
2151 there is data to show.
2152 Added average volumes into the eximstats top table text output.
2154 SC/03 Collect data on the number of addresses (recipients) as well
2155 as the number of messages in eximstats.
2157 TF/01 Correct an error in the documentation for the redirect router. Exim
2158 does (usually) call initgroups() when daemonizing.
2160 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2161 with consistent privilege compared to when running as a daemon.
2163 TF/03 Note in the spec that $authenticated_id is not set for local
2164 submissions from trusted users.
2166 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2167 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2169 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2170 by adding some example configuration directives to the default
2171 configuration file. A little bit of work is required to uncomment the
2172 directives and define how usernames and passwords are checked, but
2173 there is now a framework to start from.
2175 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2176 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2177 without this. I don't know how relevant this is to other LDAP libraries.
2179 PH/02 Add the verb name to the "unknown ACL verb" error.
2181 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2183 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2185 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2186 directory not expanded when it should be if an expanded home directory
2187 was set for the address (which is overridden by the transport).
2189 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2192 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2193 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2194 because it is too late at that time, and has no effect.
2196 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2197 security issue with \' (bugzilla #107). I could not use the
2198 PQescapeStringConn() function, because it needs a PGconn value as one of
2201 PH/08 When testing addresses using -bt, indicate those final addresses that
2202 are duplicates that would not cause an additional delivery. At least one
2203 person was confused, thinking that -bt output corresponded to deliveries.
2204 (Suppressing duplicates isn't a good idea as you lose the information
2205 about possibly different redirections that led to the duplicates.)
2207 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2208 systems where poll() doesn't work, in particular OS X.
2210 PH/10 Added more information to debugging output for retry time not reached.
2212 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2213 operations in malware.c.
2215 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2218 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2219 syslog_duplication set false, log lines that would normally be written
2220 both the the main log and to the reject log were not written to syslog at
2223 PH/14 In the default configuration, change the use of "message" in ACL warn
2224 statements to "add_header".
2226 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2227 not followed by a command (e.g. "seen endif").
2229 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2230 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2233 PH/17 Added extra conditions to the default value of delay_warning_condition
2237 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2238 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2239 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2242 The Auto-Submitted: and various List- headers are standardised, whereas I
2243 don't think Precedence: ever was.
2245 PH/18 Refactored debugging code in route_finduser() to show more information,
2246 in particular, the error code if getpwnam() issues one.
2248 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2249 This is apparently needed in addition to the PH/07 change above to avoid
2250 any possible encoding problems.
2252 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2253 but not after initializing Perl.
2255 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2256 output them only if debugging. By default they are written stderr,
2257 apparently, which is not desirable.
2259 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2262 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2265 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2267 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2268 authenticated or an ident call has been made. Suppress the default
2269 values for $authenticated_id and $authenticated_sender (but permit -oMai
2270 and -oMas) when testing with -bh.
2272 PH/24 Re-jigged the order of the tests in the default configuration so that the
2273 tests for valid domains and recipients precede the DNS black list and CSA
2274 tests, on the grounds that those ones are more expensive.
2276 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2277 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2278 This bug exists in every version of Exim that I still have, right back to
2281 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2282 However, an attempt to turn on case-sensitivity in a regex key by
2283 including (?-i) didn't work because the subject string was already
2284 lowercased, and the effects were non-intuitive. It turns out that a
2285 one-line patch can be used to allow (?-i) to work as expected.
2291 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2292 other effects) broke the use of negated acl sub-conditions.
2294 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2297 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2298 "Deny" causes Exim to reject the incoming connection with a 554 error.
2299 Unfortunately, if there is a major crisis, such as a disk failure,
2300 tcp-wrappers gives "deny", whereas what one would like would be some
2301 kind of temporary error. A kludge has been added to help with this.
2302 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2303 554 error is used if errno is still zero or contains ENOENT (which occurs
2304 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2307 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2309 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2310 errors. Otherwise a message that provokes a temporary error (when other
2311 messages do not) can cause a whole host to time out.
2313 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2314 addresses were routed directly to files or pipes from a redirect router.
2315 File deliveries just didn't batch; pipe deliveries might have suffered
2318 PH/06 A failure to get a lock for a hints database would erroneously always say
2319 "Failed to get write lock", even when it was really a read lock.
2321 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2322 of 0600. This has been changed to use the value of the lockfile_mode
2323 option (which defaults to 0600).
2325 PH/08 Applied small patch from the Sieve maintainer.
2327 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2328 folder from quota calculations, a direct delivery into this folder messed
2329 up the contents of the maildirsize file. This was because the regex was
2330 used only to exclude .Trash (or whatever) when the size of the mailbox
2331 was calculated. There was no check that a delivery was happening into an
2332 excluded directory. This bug has been fixed by ignoring all quota
2333 processing for deliveries into excluded directories.
2335 PH/10 Added the maildirfolder_create_regex option to appendfile.
2341 PH/01 The code for finding all the local interface addresses on a FreeBSD
2342 system running IPv6 was broken. This may well have applied to all BSD
2343 systems, as well as to others that have similar system calls. The broken
2344 code found IPv4 interfaces correctly, but gave incorrect values for the
2345 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2346 that it would not match correctly against @[] and not recognize the IPv6
2349 PH/02 The ipliteral router was not recognizing addresses of the form user@
2350 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2352 PH/03 Added disable_ipv6.
2354 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2355 decoded form, because it is most often used to construct To: headers
2356 lines in autoreplies, and the decoded form may well be syntactically
2357 invalid. However, $reply_address has leading white space removed, and all
2358 newlines turned into spaces so that the autoreply transport does not
2361 PH/05 If group was specified without a user on a router, and no group or user
2362 was specified on a transport, the group from the router was ignored.
2364 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2365 for visible compile-time settings that can be used to change these
2366 numbers, for those that want even more. Backwards compatibility with old
2367 spool files has been maintained. However, going back to a previous Exim
2368 release will lost any variables that are in spool files.
2370 PH/07 Two small changes when running in the test harness: increase delay when
2371 passing a TCP/IP connection to a new process, in case the original
2372 process has to generate a bounce, and remove special handling of
2373 127.0.0.2 (sic), which is no longer necessary.
2375 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2376 be the same on different OS.
2378 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2381 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2382 whether --show-vars was specified or not
2384 JJ/02 exipick: Added support for new ACL variable spool format introduced
2387 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2388 syntactically invalid From: or Reply-to: line, and a filter used this to
2389 generate an autoreply, and therefore failed to obtain an address for the
2390 autoreply, Exim could try to deliver to a non-existent relative file
2391 name, causing unrelated and misleading errors. What now happens is that
2392 it logs this as a hard delivery error, but does not attempt to create a
2395 PH/11 The exinext utility has a -C option for testing purposes, but although
2396 the given file was scanned by exinext itself; it wasn't being passed on
2397 when Exim was called.
2399 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2400 an end-of-file indication when reading a command response.
2402 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2403 compiled. In many other places in Exim, IPv6 addresses are always
2404 recognized, so I have changed this. It also means that IPv4 domain
2405 literals of the form [IPV4:n.n.n.n] are now always recognized.
2407 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2408 used if the router is not running as root, for example, when verifying at
2409 ACL time, or when using -bh. The debugging output from this situation was
2410 non-existent - all you got was a failure to exec. I have made two
2413 (a) Failures to set uid/gid, the current directory, or a process leader
2414 in a subprocess such as that created by queryprogram now generate
2415 suitable debugging ouput when -d is set.
2417 (b) The queryprogram router detects when it is not running as root,
2418 outputs suitable debugging information if -d is set, and then runs
2419 the subprocess without attempting to change uid/gid.
2421 PH/15 Minor change to Makefile for building test_host (undocumented testing
2424 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2425 additional section of a DNS packet that returns MX or SRV records.
2426 Instead, it always explicitly searches for A/AAAA records. This avoids
2427 major problems that occur when a DNS server includes only records of one
2428 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2429 fixed another bug: if SRV records were looked up and the corresponding
2430 address records were *not* found in the additional section, the port
2431 values from the SRV records were lost.
2433 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2434 using the correct key (the original address) when searching the retry
2435 rules in order to find which one to use for generating the retry hint.
2437 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2438 adding the default one. Similarly, if it contains a Reply-To: header, the
2439 errors_reply_to option, if set, is not used.
2441 PH/19 When calculating a retry time, Exim used to measure the "time since
2442 failure" by looking at the "first failed" field in the retry record. Now
2443 it does not use this if it is later than than the arrival time of the
2444 message. Instead it uses the arrival time. This makes for better
2445 behaviour in cases where some deliveries succeed, thus re-setting the
2446 "first failed" field. An example is a quota failure for a huge message
2447 when small messages continue to be delivered. Without this change, the
2448 "time since failure" will always be short, possible causing more frequent
2449 delivery attempts for the huge message than are intended.
2450 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2452 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2453 $1, $2, $3) because the numerical variables can be reset during some
2454 expansion items (e.g. "match"), thereby losing the authentication data.
2456 PH/21 Make -bV show the size of off_t variables so that the test suite can
2457 decide whether to run tests for quotas > 2G.
2459 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2460 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2461 filecount value is greater than 2G or if a quota value is greater than 2G
2462 on a system where the size of off_t is not greater than 4, a panic error
2465 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2466 never match. The debug and -bh output now contains an explicit error
2467 message indicating a malformed IPv4 address or mask.
2469 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2470 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2471 PH/23 above applies.
2473 PH/25 Do not write to syslog when running in the test harness. The only
2474 occasion when this arises is a failure to open the main or panic logs
2475 (for which there is an explicit test).
2477 PH/26 Added the /no_tell option to "control=freeze".
2479 PH/27 If a host name lookup failed very early in a connection, for example, if
2480 the IP address matched host_lookup and the reverse lookup yielded a name
2481 that did not have a forward lookup, an error message of the form "no IP
2482 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2483 could be logged. Now it outputs the IP address instead of "NULL".
2485 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2486 allows the sender and the authenticated sender to be set when
2487 submitting a message from within Exim. Since child_open_exim() is
2488 documented for local_scan(), the new function should be too.
2490 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2491 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2492 results in an empty string is now treated as unset.
2494 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2496 PH/31 Added sender_verify_fail logging option.
2498 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2499 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2500 tidied the source and removed it altogether.
2502 PH/33 When a queue run was abandoned because the load average was too high, a
2503 log line was always written; now it is written only if the queue_run log
2504 selector is set. In addition, the log line for abandonment now contains
2505 information about the queue run such as the pid. This is always present
2506 in "start" and "stop" lines but was omitted from the "abandon" line.
2508 PH/34 Omit spaces between a header name and the colon in the error message that
2509 is given when verify = headers_syntax fails (if there are lots of them,
2510 the message gets confusing).
2512 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2513 names, as there are now some PTR records that contain slashes. This check
2514 is only to protect against broken name servers that fall over on strange
2515 characters, so the fact that it applies to all lookups doesn't matter.
2517 PH/36 Now that the new test suite is complete, we can remove some of the
2518 special code in Exim that was needed for the old test suite. For example,
2519 sorting DNS records because real resolvers return them in an arbitrary
2520 order. The new test suite's fake resolver always returns records in the
2523 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2524 bounces) except when queue_only is set, to avoid logging races between
2525 the different processes.
2527 PH/38 Panic-die if .include specifies a non-absolute path.
2529 PH/39 A tweak to the "H" retry rule from its user.
2531 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2532 a label. They prevented compilation on older perls.
2534 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2535 a warning to be raised on newish perls.
2537 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2538 on queue. Changes to match documented behaviour of showing count of
2539 messages matching specified criteria.
2541 PH/40 Changed the default ident timeout from 30s to 5s.
2543 PH/41 Added support for the use of login_cap features, on those BSD systems
2544 that have them, for controlling the resources used by pipe deliveries.
2546 PH/42 The content-scanning code uses fopen() to create files in which to put
2547 message data. Previously it was not paying any attention to the mode of
2548 the files. Exim runs with umask(0) because the rest of the code creates
2549 files with open(), and sets the required mode explicitly. Thus, these
2550 files were ending up world-writeable. This was not a big issue, because,
2551 being within the spool directory, they were not world-accessible. I have
2552 created a function called modefopen, which takes an additional mode
2553 argument. It sets umask(777), creates the file, chmods it to the required
2554 mode, then resets the umask. All the relevant calls to fopen() in the
2555 content scanning code have been changed to use this function.
2557 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2558 to 24 hours. This avoids potential overflow problems when processing G
2559 and H retry rules. I suspect nobody ever tinkers with this value.
2561 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2563 PH/45 When the plaintext authenticator is running as a client, the server's
2564 challenges are checked to ensure they are valid base64 strings. By
2565 default, the authentication attempt is cancelled if an invalid string is
2566 received. Setting client_ignore_invalid_base64 true ignores these errors.
2567 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2568 they are received. Thus, the responses can be made to depend on the
2569 challenges. If an invalid string is ignored, an empty string is placed in
2572 PH/46 Messages that are created by the autoreply transport now contains a
2573 References: header, in accordance with RFCs 2822 and 3834.
2575 PH/47 Added authenticated_sender_force to the smtp transport.
2577 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2579 PH/49 Installed latest patch from the Sieve maintainer.
2581 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2582 was also set, the appendfile transport was unnecessarily scanning a
2583 directory of message files (e.g. for maildir delivery) to find the count
2584 of files (along with the size), even though it did not need this
2585 information. It now does the scan only if it needs to find either the
2586 size of the count of files.
2588 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2590 PH/52 Two bugs concerned with error handling when the smtp transport is
2593 (i) Exim was not creating retry information for temporary errors given
2594 for individual recipients after the DATA command when the smtp transport
2595 was used in LMTP mode. This meant that they could be retried too
2596 frequently, and not timed out correctly.
2598 (ii) Exim was setting the flag that allows error details to be returned
2599 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2600 recipients that were returned after the DATA command.
2602 PH/53 This is related to PH/52, but is more general: for any failing address,
2603 when detailed error information was permitted to be returned to the
2604 sender, but the error was temporary, then after the final timeout, only
2605 "retry timeout exceeded" was returned. Now it returns the full error as
2606 well as "retry timeout exceeded".
2608 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2609 do this, and (what is worse) MTAs that accept it.
2611 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2612 will now be deprecated.
2614 PH/56 New os.c-cygwin from the Cygwin maintainer.
2616 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2617 formats (previously only available in exim formats via -bpr, -bpru,
2618 and -bpra. Now also available in native and exiqgrep formats)
2620 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2621 with very large, slow to parse queues
2623 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2625 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2627 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2628 responses to authentication challenges, though it was showing the
2629 challenges; (ii) I've removed the CR characters from the debug output for
2632 PH/58 Allow for the insertion of a newline as well as a space when a string
2633 is turned into more than one encoded-word during RFC 2047 encoding. The
2634 Sieve code now uses this.
2636 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2637 data_4xx, lost_connection, tls_required.
2639 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2640 message was being sent as an SMTP response.
2642 PH/61 Add -l and -k options to exicyclog.
2644 PH/62 When verifying, if an address was redirected to one new address, so that
2645 verification continued, and the new address failed or deferred after
2646 having set something in $address_data, the value of $address_data was not
2647 passed back to the ACL. This was different to the case when no
2648 redirection occurred. The value is now passed back in both cases.
2650 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2651 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2652 use login_cap.h, so on its own it isn't the distinguishing feature. The
2653 new name refers directly to the setclassresources() function.
2655 PH/65 Added configuration files for NetBSD3.
2657 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2659 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2660 is preferred over IPv4.
2662 PH/68 The bounce_return_message and bounce_return_body options were not being
2663 honoured for bounces generated during the reception of non-SMTP messages.
2664 In particular, this applied to messages rejected by the ACL. This bug has
2665 been fixed. However, if bounce_return_message is true and bounce_return_
2666 body is false, the headers that are returned for a non-SMTP message
2667 include only those that have been read before the error was detected.
2668 (In the case of an ACL rejection, they have all been read.)
2670 PH/69 The HTML version of the specification is now built in a directory called
2671 spec_html instead of spec.html, because the latter looks like a path with
2672 a MIME-type, and this confuses some software.
2674 PH/70 Catch two compiler warnings in sieve.c.
2676 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2677 function verify_get_ident() calls ip_connect() to connect a socket, but
2678 if the "connect()" function timed out, ip_connect() used to close the
2679 socket. However, verify_get_ident() also closes the socket later, and in
2680 between Exim writes to the log, which may get opened at this point. When
2681 the socket was closed in ip_connect(), the log could get the same file
2682 descriptor number as the socket. This naturally causes chaos. The fix is
2683 not to close the socket in ip_connect(); the socket should be closed by
2684 the function that creates it. There was only one place in the code where
2685 this was missing, in the iplookup router, which I don't think anybody now
2686 uses, but I've fixed it anyway.
2688 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2689 well as to direct DNS lookups. Otherwise the handling of names in host
2690 lists is inconsistent and therefore confusing.
2696 PH/01 Two changes to the default runtime configuration:
2698 (1) Move the checks for relay_from_hosts and authenticated clients from
2699 after to before the (commented out) DNS black list checks.
2701 (2) Add control=submission to the relay_from_hosts and authenticated
2702 clients checks, on the grounds that messages accepted by these
2703 statements are most likely to be submissions.
2705 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2707 (1) Generate an error if the third argument for the ${prvs expansion is
2710 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2713 (3) Reset the variables that are obtained from the first argument of
2714 ${prvscheck and used in the second argument before leaving the code,
2715 because their memory is reclaimed, so using them afterwards may do
2718 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2719 one (it's much easier than Tom thought :-).
2721 (5) Because of (4), we can now allow for the use of $prvscheck_result
2722 inside the third argument.
2724 PH/03 For some reason, the default setting of PATH when running a command from
2725 a pipe transport was just "/usr/bin". I have changed it to
2728 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2729 anything to be listed in the output from -bV.
2731 PH/05 When a filter generated an autoreply, the entire To: header line was
2732 quoted in the delivery log line, like this:
2734 => >A.N.Other <ano@some.domain> <original@ddress> ...
2736 This has been changed so that it extracts the operative address. There
2737 may be more than one such address. If so, they are comma-separated, like
2740 => >ano@some.domain,ona@other.domain <original@ddress> ...
2742 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2743 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2744 not being looked up in the rDNS to get a host name, Exim was showing the
2745 IP address twice in Received: lines, even though the IP addresses were
2746 identical. For example:
2748 Received: from [1.2.3.4] (helo=[1.2.3.4])
2750 However, if the real host name was known, it was omitting the HELO data
2751 if it matched the actual IP address. This has been tidied up so that it
2752 doesn't show the same IP address twice.
2754 PH/07 When both +timestamp and +memory debugging was on, the value given by
2755 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2756 called by the debug printing, thereby overwriting the timestamp buffer.
2757 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2759 PH/08 When the original message was included in an autoreply transport, it
2760 always said "this is a copy of the message, including all the headers",
2761 even if body_only or headers_only was set. It now gives an appropriate
2764 PH/09 Applied a patch from the Sieve maintainer which:
2766 o fixes some comments
2767 o adds the (disabled) notify extension core
2768 o adds some debug output for the result of if/elsif tests
2769 o points to the current vacation draft in the documentation
2770 and documents the missing references header update
2774 o fixes a bug in processing the envelope test (when testing
2775 multiple envelope elements, the last element determinted the
2778 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2779 Electronic Mail") by including:
2781 Auto-submitted: auto-generated
2783 in the messages that it generates (bounce messages and others, such as
2784 warnings). In the case of bounce messages for non-SMTP mesages, there was
2785 also a typo: it was using "Auto_submitted" (underscore instead of
2786 hyphen). Since every message generated by Exim is necessarily in response
2787 to another message, thes have all been changed to:
2789 Auto-Submitted: auto-replied
2791 in accordance with these statements in the RFC:
2793 The auto-replied keyword:
2795 - SHOULD be used on messages sent in direct response to another
2796 message by an automatic process,
2798 - MUST NOT be used on manually-generated messages,
2800 - MAY be used on Delivery Status Notifications (DSNs) and Message
2801 Disposition Notifications (MDNs),
2803 - MUST NOT be used on messages generated by automatic or periodic
2804 processes, except for messages which are automatic responses to
2807 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2808 to the default Received: header definition.
2810 PH/12 Added log selector acl_warn_skipped (default on).
2812 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2813 variables because (a) they are in the wrong storage pool and (b) even if
2814 they were copied, it wouldn't work properly because of the caching.
2816 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2817 checking when decoding. Apparently there are clients that generate
2818 overlong encoded strings. Why am I not surprised?
2820 PH/15 If the first argument of "${if match_address" was not empty, but did not
2821 contain an "@" character, Exim crashed. Now it writes a panic log message
2822 and treats the condition as false.
2824 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2826 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2827 extension "envelope-auth". The code is finished and in agreement with
2828 other implementations, but there is no documentation so far and in fact,
2829 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2830 not changing the active code.
2832 Print executed "if" and "elsif" statements when debugging is used. This
2833 helps a great deal to understand what a filter does.
2835 Document more things not specified clearly in RFC3028. I had all this
2836 sorted out, when out of a sudden new issues came to my mind. Oops."
2838 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2841 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2842 canonical form (as documented). However, after a host name lookup from
2843 the IP address, check_host() was doing a simple string comparison with
2844 addresses acquired from the DNS when checking that the found name did
2845 have the original IP as one of its addresses. Since any found IPv6
2846 addresses are likely to be in abbreviated form, the comparison could
2847 fail. Luckily, there already exists a function for doing the comparison
2848 by converting both addresses to binary, so now that is used instead of
2849 the text comparison.
2851 PH/20 There was another similar case to PH/19, when a complete host name was
2852 given in a host list; looking up its IP address could give an abbreviated
2853 form, whereas the current host's name might or might not be abbreviated.
2854 The same fix has been applied.
2860 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2861 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2864 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2865 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2867 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2869 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2870 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2871 Also, for "Auto-Submitted", treat anything other than "no" as
2872 non-personal, in accordance with RFC 3834. (Previously it treated
2873 anything starting "auto-" as non-personal.)
2875 TF/01 The control=submission/name=... option had a problem with syntax
2876 errors if the name included a slash character. The /name= option
2877 now slurps the rest of the string, so it can include any characters
2878 but it must come last in the list of options (after /sender_retain
2881 PH/05 Some modifications to the interface to the fake nameserver for the new
2889 TK/01 Added the "success_on_redirect" address verification option. See
2890 NewStuff for rationale and an example.
2892 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2894 PH/02 Patch to exigrep to allow it to work on syslog lines.
2896 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2897 fread() to skip over the body file's header line, because in Cygwin the
2898 header line is locked and is inaccessible.
2900 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2901 co-exist for some time) to make it clear that it is the Exim ID that is
2902 referenced, not the Message-ID: header line.
2904 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2905 string_format() function, because snprintf() does not exist on all
2908 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2909 ${stat: expansion item.
2911 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2912 protocol synchronization error", to keep the pedants happy.
2914 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2915 well as for IRIX systems, when gcc is being used. See the host.c source
2918 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2920 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2923 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2924 transport and to the smtp transport in LMTP mode.
2926 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2928 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2929 run a filter in a subprocess. This could lead to confusion in subsequent
2930 lookups in the parent process. There should also be a search_tidyup() at
2931 the end of the subprocess.
2933 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2934 only if the host matched helo_try_verify_hosts, which caused the
2935 verification to occur when the EHLO/HELO command was issued. The ACL just
2936 tested the remembered result. Now, if a previous verification attempt has
2937 not happened, "verify = helo" does it there and then.
2939 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2941 TK/03 Fix log output including CR from clamd.
2943 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2944 exist provoked a memory error which could cause a segfault.
2946 PH/15 Installed PCRE 6.2
2948 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2950 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2951 of the problem. Specifically, suggested +O2 rather than +O1 for the
2954 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2956 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2957 record, the port was not correctly logged when the outgoing_port log
2958 selector was set (it logged the transort's default port).
2960 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2961 fallback_hosts, and "hosts" in the smtp transport.
2963 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2964 host errors such as "Connection refused".
2966 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2967 authentication with radiusclient 0.4.9:
2969 - Error returned from rc_read_config was caught wrongly
2970 - Username/password not passed on to radius server due to wrong length.
2972 The presumption is that some radiusclient API changes for 4.51/PH/17
2973 were not taken care of correctly. The code is still untested by me (my
2974 Linux distribution still has 0.3.2 of radiusclient), but it was
2975 contributed by a Radius user.
2977 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2978 expanding the "port" option of the smtp transport.
2980 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2981 while reading a MIME header. Thanks to Tom Hughes for a patch.
2983 PH/24 Include config.h inside local_scan.h so that configuration settings are
2986 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2987 that in an ACL for RCPT (for example), you can examine exactly what was
2990 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2991 commands, but it was not correctly comparing the address with the actual
2992 client host address. Thus, it would show the EHLO address in Received:
2993 header lines when this was not necessary.
2995 PH/27 Added the % operator to ${eval:}.
2997 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2998 it should be ignoring failures (because with -C, for example, it has lost
2999 privilege). It wasn't ignoring creation failures other than "already
3002 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3005 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3006 because an input line was too long, either on its own, or by virtue of
3007 too many continuations, the temporary file was not being removed, and the
3008 return code was incorrect.
3010 PH/31 Missing "BOOL" in function definition in filtertest.c.
3012 PH/32 Applied Sieve patches from the maintainer.
3014 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3016 PH/33 Added "verify = not_blind".
3018 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3019 Local/Makefile (with some defaults set). These are used in built scripts
3020 such as exicyclog, but they have never been used in the exim_install
3021 script (though there are many overriding facilities there). I have
3022 arranged that the exim_install script now takes note of these two
3025 PH/35 Installed configuration files for Dragonfly.
3027 PH/36 When a locally submitted message by a trusted user did not contain a
3028 From: header, and the sender address was obtained from -f or from an SMTP
3029 MAIL command, and the trusted user did not use -F to supply a sender
3030 name, $originator_name was incorrectly used when constructing a From:
3031 header. Furthermore, $originator_name was used for submission mode
3032 messages from external hosts without From: headers in a similar way,
3033 which is clearly wrong.
3035 PH/37 Added control=suppress_local_fixups.
3037 PH/38 When log_selector = +received_sender was set, and the addition of the
3038 sender made the log line's construction buffer exactly full, or one byte
3039 less than full, an overflow happened when the terminating "\n" was
3042 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3043 when the result of a list match is failure because a DNS lookup failed.
3045 PH/40 RM_COMMAND is now used in the building process.
3047 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3048 the "build-* directories that it finds.
3050 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3051 address in a domain literal was a prefix of an interface address.
3053 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3054 when verifying a sender address, unless rewrite_headers is false.
3056 PH/44 Wrote a long comment about why errors_to addresses are verified as
3057 recipients, not senders.
3059 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3060 the ratelimit ACL was added.
3062 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3064 PH/46 Added extra information about PostgreSQL errors to the error string.
3066 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3067 avoiding the need to install special zones in a real server. This is
3068 backwards compatible; if it can't find the fake resolver, it drops back.
3069 Thus, both old and new test suites can be run.
3071 TF/02 Added util/ratelimit.pl
3073 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3074 clock is set back in time.
3076 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3077 Candler <B.Candler@pobox.com>.
3079 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3080 OK for addresses that are the result of redirections.
3082 PH/48 A number of further additions for the benefit of the new test suite,
3083 including a fake gethostbyname() that interfaces to the fake DNS resolver
3086 TF/06 The fix for widen_domains has also been applied to qualify_single and
3087 search_parents which are the other dnslookup options that can cause
3090 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3093 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3095 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3096 service" if it encountered an MX record with an empty target hostname.
3097 The message is now "an MX or SRV record indicated no SMTP service".
3099 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3100 if the DNS of the sending site is misconfigured. This is quite a
3101 common situation. This change restores the behaviour of treating a
3102 helo verification defer as a failure.
3104 PH/51 If self=fail was set on a router, the bounce message did not include the
3105 actual error message.
3111 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3113 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3114 command itself ended in error, the underlying message about the transport
3115 filter timeout was being overwritten with the pipe command error. Now the
3116 underlying error message should be appended to the second error message.
3118 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3120 PH/02 Reduce the amount of output that "make" produces by default. Full output
3121 can still be requested.
3123 PH/03 The warning log line about a condition test deferring for a "warn" verb
3124 was being output only once per connection, rather than after each
3125 occurrence (because it was using the same function as for successful
3126 "warn" verbs). This seems wrong, so I have changed it.
3128 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3129 it should not have, which might have caused a crash in the right
3130 circumstances, but probably never did.
3132 PH/04 Installed a modified version of Tony Finch's patch to make submission
3133 mode fix the return path as well as the Sender: header line, and to
3134 add a /name= option so that you can make the user's friendly name appear
3137 TF/03 Added the control = fakedefer ACL modifier.
3139 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3140 Mark Lowes for thorough testing.
3142 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3144 TK/03 Merged latest SRS patch from Miles Wilton.
3146 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3147 with the definition in sysexits.h (which is #included earlier).
3148 Fortunately, Exim does not actually use EX_OK. The code used to try to
3149 preserve the sysexits.h value, by assumimg that macro definitions were
3150 scanned for macro replacements. I have been disabused of this notion,
3151 so now the code just undefines EX_OK before #including unistd.h.
3153 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3154 in the smtp transport. When a block could not be written in a single
3155 write() function, the timeout was being re-applied to each part-write.
3156 This seems wrong - if the receiver was accepting one byte at a time it
3157 would take for ever. The timeout is now adjusted when this happens. It
3158 doesn't have to be particularly precise.
3160 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3161 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3163 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3164 without a domain if the check to <postmaster@domain> fails.
3166 SC/01 Eximstats: added -xls and the ability to specify output files
3167 (patch written by Frank Heydlauf).
3169 SC/02 Eximstats: use FileHandles for outputing results.
3171 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3173 SC/04 Eximstats: fixed display of large numbers with -nvr option
3175 SC/05 Eximstats: fixed merging of reports with empty tables.
3177 SC/06 Eximstats: added the -include_original_destination flag
3179 SC/07 Eximstats: removed tabs and trailing whitespace.
3181 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3183 TK/06 MBOX spool code: Add real "From " MBOX separator line
3184 so the .eml file is really in mbox format (even though
3185 most programs do not really care). Patch from Alex Miller.
3187 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3188 The latter is generated from $received_to and is only set if the
3189 message has one envelope recipient. SA can use these headers,
3190 obviously out-of-the-box. Patch from Alex Miller.
3192 PH/08 The ${def test on a variable was returning false if the variable's
3193 value was "0", contrary to what the specification has always said!
3194 The result should be true unless the variable is empty.
3196 PH/09 The syntax error of a character other than { following "${if
3197 def:variable_name" (after optional whitespace) was not being diagnosed.
3198 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3199 accidental colon was present, for example, could give incorrect results.
3201 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3202 result is used (not including appendfile, where other changes are about
3205 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3206 This involved changing a lot of size variables from int to off_t. It
3207 should work with maildirs and everything.
3209 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3210 spamd dying while we are connected to it.
3212 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3215 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3216 function for BDB 4.3.
3218 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3220 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3221 the sources for PCRE have been re-arranged and more files are now
3224 PH/15 The code I had for printing potentially long long variables in PH/11
3225 above was not the best (it lost precision). The length of off_t variables
3226 is now inspected at build time, and an appropriate printing format (%ld
3227 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3228 to be "long long int" or "long int". This is needed for the internal
3229 formatting function string_vformat().
3231 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3232 the configuration file to be ":syslog", then the script "guesses" where
3233 the logs files are, rather than using the compiled in default. In our
3234 case the guess is not the same as the compiled default, so the script
3235 suddenly stopped working when I started to use syslog. The patch checks
3236 to see if log_file_path is "". If so, it attempts to read it from exim
3237 with no configuration file to get the compiled in version, before it
3238 falls back to the previous guessing code."
3240 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3241 implementing BATV in an Exim configuration. See NewStuff for the gory
3244 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3245 Makefile that are specific to HP-UX.
3247 PH/18 If the "use_postmaster" option was set for a recipient callout together
3248 with the "random" option, the postmaster address was used as the MAIL
3249 FROM address for the random test, but not for the subsequent recipient
3250 test. It is now used for both.
3252 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3253 patch removes a few documentation additions to RFC 3028, because the
3254 latest draft now contains them. It adds the new en;ascii-case comparator
3255 and a new error check for 8bit text in MIME parts. Comparator and
3256 require names are now matched exactly. I enabled the subaddress
3257 extension, but it is not well tested yet (read: it works for me)."
3259 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3260 rework some of the code of TK/09 above to avoid the hardwired use of
3261 "%lld" and "long long". Replaced the call to snprintf() with a call to
3264 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3265 records point to non-existent hosts", "retry timeout exceeded", and
3266 "retry time not reached for any host after a long failure period".
3268 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3269 experimental DomainKeys support:
3271 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3272 (2) On an error such as an illegally used "control", the wrong name for
3273 the control was given.
3275 These problems did NOT occur unless DomainKeys support was compiled.
3277 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3279 PH/24 Added ${if match_ip condition.
3281 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3282 databases so that it will be absolutely obvious if a crash occurs in the
3283 DB library. This is a regular occurrence (often caused by mis-matched
3286 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3287 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3288 picked up on a user's system that detects such things. There doesn't seem
3289 to be a gcc warning option for this - only an attribute that has to be
3290 put on the function's prototype. It seems that in Fedora Core 4 they have
3291 set this on a number of new functions. No doubt there will be more in due
3294 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3295 specify a transport. However, if an address that was verified by such a
3296 router was the subject of a callout, Exim crashed because it tried to
3297 read the rcpt_include_affixes from the non-existent transport. Now it
3298 just assumes that the setting of that option is false. This bug was
3299 introduced by 4.51/PH/31.
3301 PH/28 Changed -d+all to exclude +memory, because that information is very
3302 rarely of interest, but it makes the output a lot bigger. People tend to
3303 do -d+all out of habit.
3305 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3306 code in os-type was giving problems when libc.so lives in lib64, like on
3309 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3310 aren't the modern standard, and it seems that some systems' include files
3311 don't always have them. Exim was already checking for some of the newer
3312 ones like T_AAAA, and defining it itself. I've added checks for all the
3313 record types that Exim uses.
3315 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3316 not automatically generating a new one, as it is supposed to. This
3317 prevented TLS from working. If the file did exist, but contained invalid
3318 data, a new version was generated, as expected. It was only the case of a
3319 non-existent file that was broken.
3321 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3322 with a change in libdomainkeys > 0.64.
3324 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3325 from DNS. If the selector record carries the flag, it now has
3326 precedence over the domain-wide flag.
3328 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3330 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3331 the use of a port name that isn't defined in /etc/services) occurred, the
3332 message was deferred as in a normal delivery, and thus remained on the
3333 spool, instead of being failed because of the mua_wrapper setting. This
3334 is now fixed, and I tidied up some of the mua_wrapper messages at the
3337 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3338 the messages in a hash of arrays rather than using individual hashes.
3339 This is a bit cleaner and results in dramatic memory savings, albeit
3340 at a slight CPU cost.
3342 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3343 as requested by Marc Sherman.
3345 SC/10 Eximstats: added histograms for user specified patterns as requested
3348 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3350 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3351 fopen() in the content-scanning modules that did not already have it.
3357 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3358 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3360 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3362 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3364 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3365 to test Sieve filters that use "vacation".
3367 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3368 that changes the way the GnuTLS parameters are stored in the cache file.
3369 The new format can be generated externally. For backward compatibility,
3370 if the data in the cache doesn't make sense, Exim assumes it has read an
3371 old-format file, and it generates new data and writes a new file. This
3372 means that you can't go back to an older release without removing the
3375 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3376 work if there are any delivery delays because "one_time" forces the
3377 parent to be marked "delivered", so its unseen clone is never tried
3378 again. For this reason, Exim now forbids the simultaneous setting of
3381 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3382 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3383 entry if you want to know the details. The fix, however, neglected to
3384 consider the case when local delivery batching is involved. The test for
3385 "previously delivered" was not happening when checking to see if an
3386 address could be batched with a previous (undelivered) one; under
3387 certain circumstances this could lead to multiple deliveries to the same
3390 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3391 in its include files, and this causes problems building Exim.
3393 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3394 header_syntax) but Exim was just ignoring anything given after a slash.
3395 In particular, this caused confusion with an attempt to use "verify =
3396 reverse_host_lookup/defer_ok". An error is now given when options are
3397 supplied for verify items that do not have them. (Maybe reverse_host_
3398 lookup should have a defer_ok option, but that's a different point.)
3400 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3401 defined by RFC 821) to 2048, because there were problems with some AUTH
3402 commands, and RFC 1869 says the size should be increased for extended
3403 SMTP commands that take arguments.
3405 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3408 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3409 "unknown" error; now it says that the functionality isn't in the binary.
3411 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3412 an address' error message when a string expansion fails (syntax or
3413 whatever). Otherwise the password may appear in the log. Following change
3414 PH/42 below, there is no longer a chance of it appearing in a bounce
3417 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3419 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3420 of its addresses was ever tried. (Bugzilla bug #2.)
3422 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3423 the result incorrectly in the debug output. (It correctly added a newline
3424 to what was transported.)
3426 TF/01 Added $received_time.
3428 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3429 commented out examples of how to interface to a virus scanner and to
3430 SpamAssassin. Also added commented examples of av_scanner and
3431 spamd_address settings.
3433 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3434 and controls are allowed in which ACLs. There were a couple of minor
3435 errors. Some of the entries in the conditions table (which is a table of
3436 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3437 negation of where the condition IS allowed.
3439 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3441 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3442 header file does not have a version number, so I've had to invent a new
3443 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3444 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3445 radiusclient), but it was contributed by a Radius user.
3447 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3448 files or directories, for OpenSSL.
3450 PH/19 When an Exim process that is running as root has to create an Exim log
3451 file, it does so in a subprocess that runs as exim:exim so as to get the
3452 ownership right at creation (otherwise, other Exim processes might see
3453 the file with the wrong ownership). There was no test for failure of this
3454 fork() call, which would lead to the process getting stuck as it waited
3455 for a non-existent subprocess. Forks do occasionally fail when resources
3456 run out. I reviewed all the other calls to fork(); they all seem to check
3459 PH/20 When checking for unexpected SMTP input at connect time (before writing
3460 the banner), Exim was not dealing correctly with a non-positive return
3461 from the read() function. If the client had disconnected by this time,
3462 the result was a log entry for a synchronization error with an empty
3463 string after "input=" when read() returned zero. If read() returned -1
3464 (an event I could not check), uninitialized data bytes were printed.
3465 There were reports of junk text (parts of files, etc) appearing after
3468 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3470 PH/22 Added support for macro redefinition, and (re)definition in between
3471 driver and ACL definitions.
3473 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3474 forgetting to use the resulting value; it was using the unexpanded value.
3476 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3477 hadn't been configured. The fix is from Juergen Kreileder, who
3478 understands it better than I do:
3480 "Here's what I see happening with three configured cyrus_sasl
3481 authenticators configured (plain, login, cram-md5):
3483 On startup auth_cyrus_sasl_init() gets called for each of these.
3484 This means three calls to sasl_listmech() without a specified mech_list.
3485 => SASL tests which mechs of all available mechs actually work
3486 => three warnings about OTP not working
3487 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3489 With the patch, sasl_listmech() also gets called three times. But now
3490 SASL's mech_list option is set to the server_mech specified in the the
3491 authenticator. Or in other words, the answer from sasl_listmech()
3492 gets limited to just the mech you're testing for (which is different
3494 => the return list contains just 'plain' or 'login', 'cram-md5' or
3495 nothing depending on the value of ob->server_mech.
3497 I've just tested the patch: Authentication still works fine,
3498 unavailable mechs specified in the exim configuration are still
3499 caught, and the auth.log warnings about OTP are gone."
3501 PH/25 When debugging is enabled, the contents of the command line are added
3502 to the debugging output, even when log_selector=+arguments is not
3505 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3506 answer is "GNU", and only if the return is "GNU/something" is the answer
3509 PH/27 $acl_verify_message is now set immediately after the failure of a
3510 verification in an ACL, and so is available in subsequent modifiers. In
3511 particular, the message can be preserved by coding like this:
3513 warn !verify = sender
3514 set acl_m0 = $acl_verify_message
3516 Previously, $acl_verify_message was set only while expanding "message"
3517 and "log_message" when a very denied access.
3519 PH/28 Modified OS/os.c-Linux with
3521 -#ifndef OS_LOAD_AVERAGE
3522 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3524 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3525 nomenclature these days.)
3527 PH/29 Installed patch from the Sieve maintainer that adds the options
3528 sieve_useraddress and sieve_subaddress to the redirect router.
3530 PH/30 In these circumstances:
3531 . Two addresses routed to the same list of hosts;
3532 . First host does not offer TLS;
3533 . First host accepts first address;
3534 . First host gives temporary error to second address;
3535 . Second host offers TLS and a TLS session is established;
3536 . Second host accepts second address.
3537 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3538 and peerdn, if requested) that were in fact used only for the second
3541 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3542 attention to any local part prefix or suffix that was matched by the
3543 router that accepted the address. It now behaves in the same way as it
3544 does for delivery: the affixes are removed from the local part unless
3545 rcpt_include_affixes is set on the transport.
3547 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3548 timeout during the DATA phase of an incoming message.
3550 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3551 applied a patch sanctioned by the Sieve maintainer.
3553 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3554 the uid or gid is negative. A case of a negative gid caused this to be
3555 noticed. The fix allows for either to be negative.
3557 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3558 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3561 PH/36 The change PH/12 above was broken. Fixed it.
3563 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3564 the grounds that routing the same address twice would always produce the
3565 same answer. This might have been true once, but it is certainly no
3566 longer true now. Routing a child address may depend on the previous
3567 routing that produced that child. Some complicated redirection strategies
3568 went wrong when messages had multiple recipients, and made Exim's
3569 behaviour dependent on the order in which the addresses were given.
3571 I have moved the duplicate checking until after the routing is complete.
3572 Exim scans the addresses that are assigned to local and remote
3573 transports, and removes any duplicates. This means that more work will be
3574 done, as duplicates will always all be routed, but duplicates are
3575 presumably rare, so I don't expect this is of any significance.
3577 For deliveries to pipes, files, and autoreplies, the duplicate checking
3578 still happens during the routing process, since they are not going to be
3581 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3582 It corrects a timeout issue with spamd. This is Ian's comment: "The
3583 background is that sometimes spamd either never reads data from a
3584 connection it has accepted, or it never writes response data. The exiscan
3585 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3586 blindly assumes that writes won't block so it may never time out."
3588 PH/39 Allow G after quota size as well as K and M.
3590 PH/40 The value set for $authenticated_id in an authenticator may not contain
3591 binary zeroes or newlines because the value is written to log lines and
3592 to spool files. There was no check on this. Now the value is run through
3593 the string_printing() function so that such characters are converted to
3594 printable escape sequences.
3596 PH/41 $message_linecount is a new variable that contains the total number of
3597 lines in the message. Compare $body_linecount, which is the count for the
3600 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3601 bounce and delay warning messages, except in certain special cases, which
3604 (a) An SMTP error message from a remote host;
3605 (b) A message specified in a :fail: redirection;
3606 (c) A message specified in a "fail" command in a system filter;
3607 (d) A message specified in a FAIL return from the queryprogram router;
3608 (e) A message specified by the cannot_route_message router option.
3610 In these cases only, Exim does include the error details in bounce and
3611 warning messages. There are also a few cases where bland messages such
3612 as "unrouteable address" or "local delivery error" are given.
3614 PH/43 $value is now also set for the "else" part of a ${run expansion.
3616 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3617 being worked on, but at least Exim now implements the latest version to
3620 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3621 process to complete was treated as a delivery failure, a timeout while
3622 writing the message to the pipe was logged, but erroneously treated as a
3623 successful delivery. Such timeouts include transport filter timeouts. For
3624 consistency with the overall process timeout, these timeouts are now
3625 treated as errors, giving rise to delivery failures by default. However,
3626 there is now a new Boolean option for the pipe transport called
3627 timeout_defer, which, if set TRUE, converts the failures into defers for
3628 both kinds of timeout. A transport filter timeout is now identified in
3631 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3632 systems where "make" and "gmake" are different, calling "gmake" at top
3633 level broke things. I've arranged for the value of $(MAKE) to be passed
3634 from the Makefile to this script so that it can call the same version of
3638 A note about Exim versions 4.44 and 4.50
3639 ----------------------------------------
3641 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3642 changes of various kinds. As a consequence, a big documentation update was
3643 needed. This delayed the release for rather longer than seemed good, especially
3644 in the light of a couple of (minor) security issues. Therefore, the changes
3645 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3646 release. So 4.44 and 4.50 are in effect two different branches that both start
3649 I have left the 4.50 change log unchanged; it contains all the changes since
3650 4.43. The change log for 4.44 is below; many of its items are identical to
3651 those for 4.50. This seems to be the most sensible way to preserve the
3652 historical information.
3658 1. Minor wording change to the doc/README.SIEVE file.
3660 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3661 computation of the current number of files was incorrect.
3663 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3664 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3667 4. Give more explanation in the error message when the command for a transport
3668 filter fails to execute.
3670 5. There are several places where Exim runs a non-Exim command in a
3671 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3672 was being done only for the command run by the queryprogram router. It is
3673 now done for all such subprocesses. The other cases are: ${run, transport
3674 filters, and the commands run by the lmtp and pipe transports.
3676 6. Added CONFIGURE_GROUP build-time option.
3678 7. Some older OS have a limit of 256 on the maximum number of file
3679 descriptors. Exim was using setrlimit() to set 1000 as a large value
3680 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3681 systems. I've change it so that if it can't get 1000, it tries for 256.
3683 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3684 was an oversight, and furthermore, ever since the addition of extra
3685 controls (e.g. 4.43/32), the checks on when to allow different forms of
3686 "control" were broken. There should now be diagnostics for all cases when a
3687 control that does not make sense is encountered.
3689 9. Added the /retain_sender option to "control=submission".
3691 10. $recipients is now available in the predata ACL (oversight).
3693 11. Tidy the search cache before the fork to do a delivery from a message
3694 received from the command line. Otherwise the child will trigger a lookup
3695 failure and thereby defer the delivery if it tries to use (for example) a
3696 cached ldap connection that the parent has called unbind on.
3698 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3699 of $address_data from the recipient verification was clobbered by the
3700 sender verification.
3702 13. The value of address_data from a sender verification is now available in
3703 $sender_address_data in subsequent conditions in the ACL statement.
3705 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3707 15. Added a new option "connect=<time>" to callout options, to set a different
3710 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3711 was its contents. (It was OK if the option was not defined at all.)
3713 17. A "Completed" log line is now written for messages that are removed from
3714 the spool by the -Mrm option.
3716 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3717 information about exactly what failed.
3719 19. Added -dd to debug only the daemon process.
3721 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3722 handles timeouts, both on the server side and network timeouts. Renamed the
3723 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3725 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3726 It is now set to "smtps".
3728 22. $host_address is now set to the target address during the checking of
3729 ignore_target_hosts.
3731 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3732 being passed; this would have caused $sender_host_name to have been used if
3733 matching the list had actually called for a host name (not very likely,
3734 since this list is usually IP addresses). A host name is now passed as
3737 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3738 code that specifies a non-restarting handler (typically sigaction() in
3739 modern systems) in an attempt to fix a rare and obscure crash bug.
3741 25. Narrowed the window for a race in the daemon that could cause it to ignore
3742 SIGCHLD signals. This is not a major problem, because they are used only to
3743 wake it up if nothing else does.
3745 26. A malformed maildirsize file could cause Exim to calculate negative values
3746 for the mailbox size or file count. Odd effects could occur as a result.
3747 The maildirsize information is now recalculated if the size or filecount
3750 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3751 support for a long time. Removed HAVE_SYS_VFS_H.
3753 28. Installed the latest version of exipick from John Jetmore.
3755 29. In an address list, if the pattern was not a regular expression, an empty
3756 subject address (from a bounce message) matched only if the pattern was an
3757 empty string. Non-empty patterns were not even tested. This was the wrong
3758 because it is perfectly reasonable to use an empty address as part of a
3759 database query. An empty address is now tested by patterns that are
3760 lookups. However, all the other forms of pattern expect the subject to
3761 contain a local part and a domain, and therefore, for them, an empty
3762 address still always fails if the pattern is not itself empty.
3764 30. Exim went into a mad DNS loop when attempting to do a callout where the
3765 host was specified on an smtp transport, and looking it up yielded more
3766 than one IP address.
3768 31. Re-factored the code for checking spool and log partition space into a
3769 function that finds that data and another that does the check. The former
3770 is then used to implement four new variables: $spool_space, $log_space,
3771 $spool_inodes, and $log_inodes.
3773 32. The RFC2047 encoding function was originally intended for short strings
3774 such as real names; it was not keeping to the 75-character limit for
3775 encoded words that the RFC imposes. It now respects the limit, and
3776 generates multiple encoded words if necessary. To be on the safe side, I
3777 have increased the buffer size for the ${rfc2047: expansion operator from
3780 33. It is now permitted to omit both strings after an "if" condition; if the
3781 condition is true, the result is "true". As before, when the second string
3782 is omitted, a false condition yields an empty string. This makes it less
3783 cumbersome to write custom ACL and router conditions.
3785 34. Failure to deliver a bounce message always caused it to be frozen, even if
3786 there was an errors_to setting on the router. The errors_to setting is now
3789 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3790 canonical form (fully expanded) before being placed in
3791 $sender_host_address.
3793 36. The table in the code that translates DNS record types into text (T_A to
3794 "A" for instance) was missing entries for NS and CNAME. It is just possible
3795 that this could have caused confusion if both these types were looked up
3796 for the same domain, because the text type is used as part of Exim's
3797 per-process caching. But the chance of anyone hitting this buglet seems
3800 37. The dnsdb lookup has been extended in a number of ways.
3802 (1) There is a new type, "zns", which walks up the domain tree until it
3803 finds some nameserver records. It should be used with care.
3805 (2) There is a new type, "mxh", which is like "mx" except that it returns
3806 just the host names, not the priorities.
3808 (3) It is now possible to give a list of domains (or IP addresses) to be
3809 looked up. The behaviour when one of the lookups defers can be
3810 controlled by a keyword.
3812 (4) It is now possible to specify the separator character for use when
3813 multiple records are returned.
3815 38. The dnslists ACL condition has been extended: it is now possible to supply
3816 a list of IP addresses and/or domains to be looked up in a particular DNS
3819 39. Added log_selector=+queue_time_overall.
3821 40. When running the queue in the test harness, wait just a tad after forking a
3822 delivery process, to get repeatability of debugging output.
3824 41. Include certificate and key file names in error message when GnuTLS fails
3825 to set them up, because the GnuTLS error message doesn't include the name
3826 of the failing file when there is a problem reading it.
3828 42. Allow both -bf and -bF in the same test run.
3830 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3832 44. The "Exiscan patch" is now merged into the mainline Exim source.
3834 45. Sometimes the final signoff response after QUIT could fail to get
3835 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3836 < 0 before doing a fflush(). This bug looks as though it goes back to the
3837 introduction of TLS in release 3.20, but "sometimes" must have been rare
3838 because the tests only now provoked it.
3840 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3841 (this can affect the format of dates).
3843 47. exim_tidydb, when checking for the continued existence of a message for
3844 which it has found a message-specific retry record, was not finding
3845 messages that were in split spool directories. Consequently, it was
3846 deleting retry records that should have stayed in existence.
3848 48. Steve fixed some bugs in eximstats.
3850 49. The SPA authentication driver was not abandoning authentication and moving
3851 on to the next authenticator when an expansion was forced to fail,
3852 contradicting the general specification for all authenticators. Instead it
3853 was generating a temporary error. It now behaves as specified.
3855 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3856 (the order specifies the preference for clients). The order is now AES256,
3857 AES128, 3DES, ARCFOUR128.
3859 51. Small patch to Sieve code - explicitly set From: when generating an
3862 52. Exim crashed if a remote delivery caused a very long error message to be
3863 recorded - for instance if somebody sent an entire SpamAssassin report back
3864 as a large number of 550 error lines. This bug was coincidentally fixed by
3865 increasing the size of one of Exim's internal buffers (big_buffer) that
3866 happened as part of the Exiscan merge. However, to be on the safe side, I
3867 have made the code more robust (and fixed the comments that describe what
3870 53. Now that there can be additional text after "Completed" in log lines (if
3871 the queue_time_overall log selector is set), a one-byte patch to exigrep
3872 was needed to allow it to recognize "Completed" as not the last thing in
3875 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3876 patch that reportedly fixes this has been added. I am not expert enough to
3877 create a test for it. This is what the patch creator wrote:
3879 "I found a little strange behaviour of ldap code when working with
3880 Windows 2003 AD Domain, where users was placed in more than one
3881 Organization Units. When I tried to give exim partial DN, the exit code
3882 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3883 But simultaneously result of request was absolutely normal ldap result,
3884 so I produce this patch..."
3886 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3887 REFERENCE, so I have modified the code to exclude the patch when that macro
3890 55. Some experimental protocols are using DNS PTR records for new purposes. The
3891 keys for these records are domain names, not reversed IP addresses. The
3892 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3893 leaves it alone. Component reversal etc. now happens only for IP addresses.
3896 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3898 57. Double the size of the debug message buffer (to 2048) so that more of very
3899 long debug lines gets shown.
3901 58. The exicyclog utility now does better if the number of log files to keep
3902 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3904 59. Two changes related to the smtp_active_hostname option:
3906 (1) $smtp_active_hostname is now available as a variable.
3907 (2) The default for smtp_banner uses $smtp_active_hostname instead
3908 of $primary_hostname.
3910 60. The host_aton() function is supposed to be passed a string that is known
3911 to be a valid IP address. However, in the case of IPv6 addresses, it was
3912 not checking this. This is a hostage to fortune. Exim now panics and dies
3913 if the condition is not met. A case was found where this could be provoked
3914 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3915 components; fortuitously, this particular loophole had already been fixed
3916 by change 4.50/55 above.
3918 If there are any other similar loopholes, the new check in host_aton()
3919 itself should stop them being exploited. The report I received stated that
3920 data on the command line could provoke the exploit when Exim was running as
3921 exim, but did not say which command line option was involved. All I could
3922 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3923 running as the user.
3926 61. There was a buffer overflow vulnerability in the SPA authentication code
3927 (which came originally from the Samba project). I have added a test to the
3928 spa_base64_to_bits() function which I hope fixes it.
3931 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3932 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3934 63. The daemon start-up calls getloadavg() while still root for those OS that
3935 need the first call to be done as root, but it missed one case: when
3936 deliver_queue_load_max is set with deliver_drop_privilege. This is
3937 necessary for the benefit of the queue runner, because there is no re-exec
3938 when deliver_drop_privilege is set.
3940 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3941 This has been fixed.
3943 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3944 was in use, was not putting the data itself into the right store pool;
3945 consequently, it could be overwritten for a subsequent message in the same
3946 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3949 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3951 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3952 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3953 the function were treating the return as a boolean value, which happened to
3954 work because 0=false and not-0=true, but is not correct code.
3956 68. The host_aton() function was not handling scoped IPv6 addresses (those
3957 with, for example, "%eth0" on the end) correctly.
3959 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3960 negated items (that is, ~something) in unsigned ints. Some compilers
3961 apparently mutter when there is no cast.
3963 70. If an address verification called from an ACL failed, and did not produce a
3964 user-specific message (i.e. there was only a "system" message), nothing was
3965 put in $acl_verify_message. In this situation, it now puts the system
3968 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3969 session; change 4.31/43 added the unwanted input to the log line - except
3970 that it did not do this in the start of session case. It now does.
3972 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3973 This is wrong and can cause the other end to generate a synchronization
3974 error if it is another Exim or anything else that does the synchronization
3975 check. A QUIT command is no longer sent after a timeout.
3977 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3978 during host lookups.
3980 74. The defer_ok option of callout verification was not working if it was used
3981 when verifying addresses in header lines, that is, for this case:
3983 verify = header_sender/callout=defer_ok
3985 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3986 those file descriptors could be used for SMTP connections. If anything
3987 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3988 could be sent to the SMTP client, causing chaos. The daemon now opens
3989 stdin, stdout, and stderr to /dev/null when it puts itself into the
3992 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3993 log by default. The user can override this with suitable Perl magic.
3995 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3996 the log message when discard triggers, was not working for the DATA ACL or
3997 for the non-SMTP ACL.
3999 78. Error message wording change in sieve.c.
4001 79. If smtp_accept_max_per_host was set, the number of connections could be
4002 restricted to fewer than expected, because the daemon was trying to set up
4003 a new connection before checking whether the processes handling previous
4004 connections had finished. The check for completed processes is now done
4005 earlier. On busy systems, this bug wouldn't be noticed because something
4006 else would have woken the daemon, and it would have reaped the completed
4009 80. If a message was submitted locally by a user whose login name contained one
4010 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4011 It caused a spool format error. I have fixed the spool reading code. A
4012 related problem was that the "from" clause in the Received: line became
4013 illegal because of the space(s). It is now covered by ${quote_local_part.
4015 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4018 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4020 83. Patch from Sieve maintainer to fix unterminated string problem in
4021 "vacation" handling.
4023 84. Some minor changes to the Linux configuration files to help with other
4024 OS variants using glibc.
4026 85. One more patch for Sieve to update vacation handling to latest spec.
4029 ----------------------------------------------------
4030 See the note above about the 4.44 and 4.50 releases.
4031 ----------------------------------------------------
4037 1. Change 4.43/35 introduced a bug that caused file counts to be
4038 incorrectly computed when quota_filecount was set in an appendfile
4041 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4042 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4045 3. Give more explanation in the error message when the command for a transport
4046 filter fails to execute.
4048 4. There are several places where Exim runs a non-Exim command in a
4049 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4050 was being done only for the command run by the queryprogram router. It is
4051 now done for all such subprocesses. The other cases are: ${run, transport
4052 filters, and the commands run by the lmtp and pipe transports.
4054 5. Some older OS have a limit of 256 on the maximum number of file
4055 descriptors. Exim was using setrlimit() to set 1000 as a large value
4056 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4057 systems. I've change it so that if it can't get 1000, it tries for 256.
4059 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4060 was an oversight, and furthermore, ever since the addition of extra
4061 controls (e.g. 4.43/32), the checks on when to allow different forms of
4062 "control" were broken. There should now be diagnostics for all cases when a
4063 control that does not make sense is encountered.
4065 7. $recipients is now available in the predata ACL (oversight).
4067 8. Tidy the search cache before the fork to do a delivery from a message
4068 received from the command line. Otherwise the child will trigger a lookup
4069 failure and thereby defer the delivery if it tries to use (for example) a
4070 cached ldap connection that the parent has called unbind on.
4072 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4073 of $address_data from the recipient verification was clobbered by the
4074 sender verification.
4076 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4077 was its contents. (It was OK if the option was not defined at all.)
4079 11. A "Completed" log line is now written for messages that are removed from
4080 the spool by the -Mrm option.
4082 12. $host_address is now set to the target address during the checking of
4083 ignore_target_hosts.
4085 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4086 being passed; this would have caused $sender_host_name to have been used if
4087 matching the list had actually called for a host name (not very likely,
4088 since this list is usually IP addresses). A host name is now passed as
4091 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4092 code that specifies a non-restarting handler (typically sigaction() in
4093 modern systems) in an attempt to fix a rare and obscure crash bug.
4095 15. Narrowed the window for a race in the daemon that could cause it to ignore
4096 SIGCHLD signals. This is not a major problem, because they are used only to
4097 wake it up if nothing else does.
4099 16. A malformed maildirsize file could cause Exim to calculate negative values
4100 for the mailbox size or file count. Odd effects could occur as a result.
4101 The maildirsize information is now recalculated if the size or filecount
4104 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4105 support for a long time. Removed HAVE_SYS_VFS_H.
4107 18. Updated exipick to current release from John Jetmore.
4109 19. Allow an empty sender to be matched against a lookup in an address list.
4110 Previously the only cases considered were a regular expression, or an
4113 20. Exim went into a mad DNS lookup loop when doing a callout where the
4114 host was specified on the transport, if the DNS lookup yielded more than
4117 21. The RFC2047 encoding function was originally intended for short strings
4118 such as real names; it was not keeping to the 75-character limit for
4119 encoded words that the RFC imposes. It now respects the limit, and
4120 generates multiple encoded words if necessary. To be on the safe side, I
4121 have increased the buffer size for the ${rfc2047: expansion operator from
4124 22. Failure to deliver a bounce message always caused it to be frozen, even if
4125 there was an errors_to setting on the router. The errors_to setting is now
4128 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4129 canonical form (fully expanded) before being placed in
4130 $sender_host_address.
4132 24. Updated eximstats to version 1.33
4134 25. Include certificate and key file names in error message when GnuTLS fails
4135 to set them up, because the GnuTLS error message doesn't include the name
4136 of the failing file when there is a problem reading it.
4138 26. Expand error message when OpenSSL has problems setting up cert/key files.
4141 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4142 (this can affect the format of dates).
4144 28. exim_tidydb, when checking for the continued existence of a message for
4145 which it has found a message-specific retry record, was not finding
4146 messages that were in split spool directories. Consequently, it was
4147 deleting retry records that should have stayed in existence.
4149 29. eximstats updated to version 1.35
4150 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4151 1.35 - bugfix such that pie charts by volume are generated correctly
4153 30. The SPA authentication driver was not abandoning authentication and moving
4154 on to the next authenticator when an expansion was forced to fail,
4155 contradicting the general specification for all authenticators. Instead it
4156 was generating a temporary error. It now behaves as specified.
4158 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4159 (the order specifies the preference for clients). The order is now AES256,
4160 AES128, 3DES, ARCFOUR128.
4162 31. Small patch to Sieve code - explicitly set From: when generating an
4165 32. Exim crashed if a remote delivery caused a very long error message to be
4166 recorded - for instance if somebody sent an entire SpamAssassin report back
4167 as a large number of 550 error lines. This bug was coincidentally fixed by
4168 increasing the size of one of Exim's internal buffers (big_buffer) that
4169 happened as part of the Exiscan merge. However, to be on the safe side, I
4170 have made the code more robust (and fixed the comments that describe what
4173 33. Some experimental protocols are using DNS PTR records for new purposes. The
4174 keys for these records are domain names, not reversed IP addresses. The
4175 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4176 leaves it alone. Component reversal etc. now happens only for IP addresses.
4179 34. The host_aton() function is supposed to be passed a string that is known
4180 to be a valid IP address. However, in the case of IPv6 addresses, it was
4181 not checking this. This is a hostage to fortune. Exim now panics and dies
4182 if the condition is not met. A case was found where this could be provoked
4183 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4184 components; fortuitously, this particular loophole had already been fixed
4185 by change 4.50/55 or 4.44/33 above.
4187 If there are any other similar loopholes, the new check in host_aton()
4188 itself should stop them being exploited. The report I received stated that
4189 data on the command line could provoke the exploit when Exim was running as
4190 exim, but did not say which command line option was involved. All I could
4191 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4192 running as the user.
4195 35. There was a buffer overflow vulnerability in the SPA authentication code
4196 (which came originally from the Samba project). I have added a test to the
4197 spa_base64_to_bits() function which I hope fixes it.
4200 36. The daemon start-up calls getloadavg() while still root for those OS that
4201 need the first call to be done as root, but it missed one case: when
4202 deliver_queue_load_max is set with deliver_drop_privilege. This is
4203 necessary for the benefit of the queue runner, because there is no re-exec
4204 when deliver_drop_privilege is set.
4206 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4207 was in use, was not putting the data itself into the right store pool;
4208 consequently, it could be overwritten for a subsequent message in the same
4209 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4212 38. Sometimes the final signoff response after QUIT could fail to get
4213 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4214 < 0 before doing a fflush(). This bug looks as though it goes back to the
4215 introduction of TLS in release 3.20, but "sometimes" must have been rare
4216 because the tests only now provoked it.
4222 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4223 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4224 Now it returns an integer. A number of calls were still expecting a T/F
4225 return. Fortuitously, in all cases, the tests worked in OK situations,
4226 which is the norm. However, things would have gone wrong on any write
4227 failures on the smtp file descriptor. This function is used when sending
4228 messages over SMTP and also when doing verify callouts.
4230 2. When Exim is called to do synchronous delivery of a locally submitted
4231 message (the -odf or -odi options), it no longer closes stderr before doing
4234 3. Implemented the mua_wrapper option.
4236 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4238 5. Implemented the functions header_remove(), header_testname(),
4239 header_add_at_position(), and receive_remove_recipient(), and exported them
4242 6. If an ACL "warn" statement specified the addition of headers, Exim already
4243 inserted X-ACL-Warn: at the start if there was no header name. However, it
4244 was not making this test for the second and subsequent header lines if
4245 there were newlines in the string. This meant that an invalid header could
4246 be inserted if Exim was badly configured.
4248 7. Allow an ACL "warn" statement to add header lines at the start or after all
4249 the Received: headers, as well as at the end.
4251 8. Added the rcpt_4xx retry error code.
4253 9. Added postmaster_mailfrom=xxx to callout verification option.
4255 10. Added mailfrom=xxxx to the callout verification option, for verify=
4258 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4259 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4261 12. Inserted some casts to stop certain compilers warning when using pointer
4262 differences as field lengths or precisions in printf-type calls (mostly
4263 affecting debugging statements).
4265 13. Added optional readline() support for -be (dynamically loaded).
4267 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4268 same clock tick as a message's arrival, so that its received time was the
4269 same as the "first fail" time on the retry record, and that message
4270 remained on the queue past the ultimate address timeout, every queue runner
4271 would try a delivery (because it was past the ultimate address timeout) but
4272 after another failure, the ultimate address timeout, which should have then
4273 bounced the address, did not kick in. This was a "< instead of <=" error;
4274 in most cases the first failure would have been in the next clock tick
4275 after the received time, and all would be well.
4277 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4278 being recognized when the domain list was tested by the match_domain
4279 condition in an expansion string.
4281 16. Added the ${str2b64: operator.
4283 17. Exim was always calling setrlimit() to set a large limit for the number of
4284 processes, without checking whether the existing limit was already
4285 adequate. (It did check for the limit on file descriptors.) Furthermore,
4286 errors from getrlimit() and setrlimit() were being ignored. Now they are
4287 logged to the main and panic logs, but Exim does carry on, to try to do its
4288 job under whatever limits there are.
4290 18. Imported PCRE 5.0.
4292 19. Trivial typo in log message " temporarily refused connection" (the leading
4295 20. If the log selector return_path_on_delivery was set and an address was
4296 redirected to /dev/null, the delivery process crashed because it assumed
4297 that a return path would always be set for a "successful" delivery. In this
4298 case, the whole delivery is bypassed as an optimization, and therefore no
4301 21. Internal re-arrangement: the function for sending a challenge and reading
4302 a response while authentication was assuming a zero-terminated challenge
4303 string. It's now changed to take a pointer and a length, to allow for
4304 binary data in such strings.
4306 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4308 23. Exim was not respecting finduser_retries when seeking the login of the
4309 uid under which it was called; it was always trying 10 times. (The default
4310 setting of finduser_retries is zero.) Also, it was sleeping after the final
4311 failure, which is pointless.
4313 24. Implemented tls_on_connect_ports.
4315 25. Implemented acl_smtp_predata.
4317 26. If the domain in control=submission is set empty, Exim assumes that the
4318 authenticated id is a complete email address when it generates From: or
4319 Sender: header lines.
4321 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4322 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4323 chown and chgrp in /bin and hostname in /usr/bin.
4325 28. Exim was keeping the "process log" file open after each use, just as it
4326 does for the main log. This opens the possibility of it remaining open for
4327 long periods when the USR1 signal hits a daemon. Occasional processlog
4328 errors were reported, that could have been caused by this. Anyway, it seems
4329 much more sensible not to leave this file open at all, so that is what now
4332 29. The long-running daemon process does not normally write to the log once it
4333 has entered its main loop, and it closes the log before doing so. This is
4334 so that log files can straightforwardly be renamed and moved. However,
4335 there are a couple of unusual error situations where the daemon does write
4336 log entries, and I had neglected to close the log afterwards.
4338 30. The text of an SMTP error response that was received during a remote
4339 delivery was being truncated at 512 bytes. This is too short for some of
4340 the long messages that one sometimes sees. I've increased the limit to
4343 31. It is now possible to make retry rules that apply only when a message has a
4344 specific sender, in particular, an empty sender.
4346 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4347 it possible to be selective about when SMTP synchronization is enforced.
4349 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4351 32. Implemented hosts_connection_nolog.
4353 33. Added an ACL for QUIT.
4355 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4358 35. Added mailbox_size and mailbox_filecount to appendfile.
4360 36. Added control = no_multiline_responses to ACLs.
4362 37. There was a bug in the logic of the code that waits for the clock to tick
4363 in the case where the clock went backwards by a substantial amount such
4364 that the microsecond fraction of "now" was more than the microsecond
4365 fraction of "then" (but the whole seconds number was less).
4367 38. Added support for the libradius Radius client library this is found on
4368 FreeBSD (previously only the radiusclient library was supported).
4374 1. When certain lookups returned multiple values in the form name=value, the
4375 quoting of the values was not always being done properly. Specifically:
4376 (a) If the value started with a double quote, but contained no whitespace,
4378 (b) If the value contained whitespace other than a space character (i.e.
4379 tabs or newlines or carriage returns) it was not quoted.
4380 This fix has been applied to the mysql and pgsql lookups by writing a
4381 separate quoting function and calling it from the lookup code. The fix
4382 should probably also be applied to nisplus, ibase and oracle lookups, but
4383 since I cannot test any of those, I have not disturbed their existing code.
4385 2. A hit in the callout cache for a specific address caused a log line with no
4386 reason for rejecting RCPT. Now it says "Previous (cached) callout
4387 verification failure".
4389 3. There was an off-by-one bug in the queryprogram router. An over-long
4390 return line was truncated at 256 instead of 255 characters, thereby
4391 overflowing its buffer with the terminating zero. As well as fixing this, I
4392 have increased the buffer size to 1024 (and made a note to document this).
4394 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4395 when Exim is waiting for an SMTP response from a remote server, Exim
4396 restarts its select() call on the socket, thereby resetting its timeout.
4397 This is not a problem when such interrupts are rare. Somebody set up a cron
4398 job to run exiwhat every 2 minutes, which is less than the normal select()
4399 timeout (5 or 10 minutes). This meant that the select() timeout never
4400 kicked in because it was always reset. I have fixed this by comparing the
4401 time when an interrupt arrives with the time at the start of the first call
4402 to select(). If more time than the timeout has elapsed, the interrupt is
4403 treated as a timeout.
4405 5. Some internal re-factoring in preparation for the addition of Sieve
4406 extensions (by MH). In particular, the "personal" test is moved to a
4407 separate function, and given an option for scanning Cc: and Bcc: (which is
4408 not set for Exim filters).
4410 6. When Exim created an email address using the login of the caller as the
4411 local part (e.g. when creating a From: or Sender: header line), it was not
4412 quoting the local part when it contained special characters such as @.
4414 7. Installed new OpenBSD configuration files.
4416 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4417 try to make them clearer.
4419 9. Callout options, other than the timeout value, were being ignored when
4420 verifying sender addresses in header lines. For example, when using
4422 verify = header_sender/callout=no_cache
4424 the cache was (incorrectly) being used.
4426 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4427 only the Cygwin environment.
4429 11. When return_path_on_delivery was set as a log selector, if different remote
4430 addresses in the same message used different return paths and parallel
4431 remote delivery occurred, the wrong values would sometimes be logged.
4432 (Whenever a remote delivery process finished, the return path value from
4433 the most recently started remote delivery process was logged.)
4435 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4436 lines when AUTH and/or TLS are in use. This is the "received protocol"
4437 field. Exim used to use "asmtp" for authenticated SMTP, without any
4438 indication (in the protocol name) for TLS use. Now it follows the RFC and
4439 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4440 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4441 names appear in log lines as well as in Received: header lines.
4443 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4444 extensions, and comparison tests, and to fix some bugs.
4446 14. Changes to the "personal" filter test:
4448 (1) The test was buggy in that it was just doing the equivalent of
4449 "contains" tests on header lines. For example, if a user's address was
4450 anne@some.where, the "personal" test would incorrectly be true for
4452 To: susanne@some.where
4454 This test is now done by extracting each address from the header in turn,
4455 and checking the entire address. Other tests that are part of "personal"
4456 are now done using regular expressions (for example, to check local parts
4457 of addresses in From: header lines).
4459 (2) The list of non-personal local parts in From: addresses has been
4460 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4461 taken from the Sieve specification recommendations.
4463 (3) If the message contains any header line starting with "List-" it is
4464 treated as non-personal.
4466 (4) The test for "circular" in the Subject: header line has been removed
4467 because it now seems ill-conceived.
4469 15. Minor typos in src/EDITME comments corrected.
4471 16. Installed latest exipick from John Jetmore.
4473 17. If headers_add on a router specified a text string that was too long for
4474 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4475 of string_sprintf() is now avoided.
4477 18. $message_body_size was not set (it was always zero) when running the DATA
4478 ACL and the local_scan() function.
4480 19. For the "mail" command in an Exim filter, no default was being set for
4481 the once_repeat time, causing a random time value to be used if "once" was
4482 specified. (If the value happened to be <= 0, no repeat happened.) The
4483 default is now 0s, meaning "never repeat". The "vacation" command was OK
4484 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4485 (I found it when inspecting the code).
4487 20. There is now an overall timeout for performing a callout verification. It
4488 defaults to 4 times the callout timeout, which applies to individual SMTP
4489 commands during the callout. The overall timeout applies when there is more
4490 than one host that can be tried. The timeout is checked before trying the
4491 next host. This prevents very long delays if there are a large number of
4492 hosts and all are timing out (e.g. when the network connections are timing
4493 out). The value of the overall timeout can be changed by specifying an
4494 additional sub-option for "callout", called "maxwait". For example:
4496 verify = sender/callout=5s,maxwait=20s
4498 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4499 to the end before writing, but this should make it even safer).
4501 22. Exim was forgetting that it had advertised PIPELINING for the second and
4502 subsequent messages on an SMTP connection. It was also not resetting its
4503 memory on STARTTLS and an internal HELO.
4505 23. When Exim logs an SMTP synchronization error within a session, it now
4506 records whether PIPELINING has been advertised or not.
4508 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4509 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4510 rather than long int.
4512 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4514 26. Added the never_mail option to autoreply.
4520 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4521 crash if the getsockname() call failed; this can happen if a connection is
4522 closed very soon after it is established. The problem was simply in the
4523 order in which certain operations were done, causing Exim to try to write
4524 to the SMTP stream before it had set up the file descriptor. The bug has
4525 been fixed by making things happen in the correct order.
4531 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4532 before the connection was closed, thus losing the rejection response.
4534 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4535 some early Solaris releases, but causes trouble in current releases where
4536 socklen_t is defined.
4538 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4541 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4544 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4545 that releasing the top of it at the end releases what was used for sub-
4546 expansions (unless the block got too big). However, discard this block if
4547 the first thing is a variable or header, so that we can use its block when
4548 it is dynamic (useful for very large $message_headers, for example).
4550 6. Lookups now cache *every* query, not just the most recent. A new, separate
4551 store pool is used for this. It can be recovered when all lookup caches are
4552 flushed. Lookups now release memory at the end of their result strings.
4553 This has involved some general refactoring of the lookup sources.
4555 7. Some code has been added to the store_xxx() functions to reduce the amount
4556 of flapping under certain conditions.
4558 8. log_incoming_interface used to affect only the <= reception log lines. Now
4559 it causes the local interface and port to be added to several more SMTP log
4560 lines, for example "SMTP connection from", and rejection lines.
4562 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4564 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4566 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4567 could be overwritten at the end of the current message (or the start of a
4568 new message if it was set in a HELO ACL). The value is now preserved for
4569 the duration of the SMTP connection.
4571 12. If a transport had a headers_rewrite setting, and a matching header line
4572 contained an unqualified address, that address was qualified, even if it
4573 did not match any rewriting rules. The underlying bug was that the values
4574 of the flags that permit the existence of unqualified sender and recipient
4575 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4576 non-local messages, and by -bnq for local messages) were not being
4577 preserved with the message after it was received.
4579 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4580 "next input=" as part of the text comprising the host identity instead of
4581 the correct text. The code was using the same buffer for two different
4582 strings. However, depending on which order the printing function evaluated
4583 its arguments, the bug did not always show up. Under Linux, for example, my
4584 test suite worked just fine.
4586 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4587 doesn't work with some older versions of Perl. It has been changed to "my",
4588 which in any case is probably the better facility to use.
4590 15. A really picky compiler found some instances of statements for creating
4591 error messages that either had too many or two few arguments for the format
4594 16. The size of the buffer for calls to the DNS resolver has been increased
4595 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4596 for addresses that have a lot of PTR records. This alleviates a problem; it
4597 does not fully solve it.
4599 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4600 buffer now truncates the list and logs the incident, which is the same
4601 action as happens when Exim is looking up a host name and its aliases.
4602 Previously in this situation something unpredictable would happen;
4603 sometimes it was "internal error: store_reset failed".
4605 18. If a server dropped the connection unexpectedly when an Exim client was
4606 using GnuTLS and trying to read a response, the client delivery process
4607 crashed while trying to generate an error log message.
4609 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4610 string, for example:
4612 warn message = H1: something\nH2: something
4614 the text was added as a single header line from Exim's point of view
4615 though it ended up OK in the delivered message. However, searching for the
4616 second and subsequent header lines using $h_h2: did not work. This has been
4617 fixed. Similarly, if a system filter added multiple headers in this way,
4618 the routers could not see them.
4620 20. Expanded the error message when iplsearch is called with an invalid key to
4621 suggest using net-iplsearch in a host list.
4623 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4624 ACLs are no longer actually imposed (and a message to that effect is
4627 22. If a "gecos" field in a passwd entry contained escaped characters, in
4628 particular, if it contained a \" sequence, Exim got it wrong when building
4629 a From: or a Sender: header from that name. A second bug also caused
4630 incorrect handling when an unquoted " was present following a character
4631 that needed quoting.
4633 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4634 was not being matched caselessly.
4636 24. Arranged for all hyphens in the exim.8 source to be escaped with
4639 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4640 properly. Recipient callout cache records were still being keyed to include
4641 the sender, even when use_sender was set false. This led to far more
4642 callouts that were necessary. The sender is no longer included in the key
4643 when use_sender is false.
4645 26. Added "control = submission" modifier to ACLs.
4647 27. Added the ${base62d: operator to decode base 62 numbers.
4649 28. dnsdb lookups can now access SRV records.
4651 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4652 the configuration file.
4654 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4655 (-v) mode. This makes the output for a verbose queue run more intelligible.
4657 31. Added a use_postmaster feature to recipient callouts.
4659 32. Added the $body_zerocount variable, containing the number of binary zero
4660 bytes in the message body.
4662 33. The time of last modification of the "new" subdirectory is now used as the
4663 "mailbox time last read" when there is a quota error for a maildir
4666 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4668 35. Added +ignore_unknown as a special item in host lists.
4670 36. Code for decoding IPv6 addresses in host lists is now included, even if
4671 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4672 address was recognized as an IP address, but was then not correctly decoded
4673 into binary, causing unexpected and incorrect effects when compared with
4680 1. Very minor rewording of debugging text in manualroute to say "list of
4681 hosts" instead of "hostlist".
4683 2. If verify=header_syntax was set, and a header line with an unqualified
4684 address (no domain) and a large number of spaces between the end of the
4685 name and the colon was received, the reception process suffered a buffer
4686 overflow, and (when I tested it) crashed. This was caused by some obsolete
4687 code that should have been removed. The fix is to remove it!
4689 3. When running in the test harness, delay a bit after writing a bounce
4690 message to get a bit more predictability in the log output.
4692 4. Added a call to search_tidyup() just before forking a reception process. In
4693 theory, someone could use a lookup in the expansion of smtp_accept_max_
4694 per_host which, without the tidyup, could leave open a database connection.
4696 5. Added the variables $recipient_data and $sender_data which get set from a
4697 lookup success in an ACL "recipients" or "senders" condition, or a router
4698 "senders" option, similar to $domain_data and $local_part_data.
4700 6. Moved the writing of debug_print from before to after the "senders" test
4703 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4704 problems for message scanning, either using a data ACL, or using
4705 local_scan() because the Received: header was not generated till after they
4706 were called (in order to set the time as the time of reception completion).
4707 I have revised the way this works. The header is now generated after the
4708 body is received, but before the ACL or local_scan() are called. After they
4709 are run, the timestamp in the header is updated.
4715 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4716 before starting a queue runner without re-exec. This happened only when
4717 deliver_drop_privilege was set or when the Exim user was set to root. The
4718 effect of the bug was that timeouts during subsequent deliveries caused
4719 crashes instead of being properly handled. The handler is now left at its
4720 default (and expected) setting.
4722 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4723 message, again when deliver_drop_privilege is set or Exim is run as root.
4724 The bug described in (1) was not present in this case, but the tidying up
4725 of the other signals was missing. I have made the two cases consistent.
4727 3. The ignore_target_hosts setting on a manualroute router was being ignored
4728 for hosts that were looked up using the /MX notation.
4730 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4733 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4734 operated on the sender address. After changing the $sender_address to <>
4735 for the sender address verify, Exim was re-instated it as the original
4736 (before rewriting) address, but remembering that it had rewritten it, so it
4737 wasn't rewriting it again. This bug also had the effect of breaking the
4738 sender address verification caching when the sender address was rewritten.
4740 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4741 This has been changed so that if the ip literal address matches
4742 ignore_target_hosts, the router declines.
4744 7. Added expansion conditions match_domain, match_address, and match_local_
4745 part (NOT match_host).
4747 8. The placeholder for the Received: header didn't have a length field set.
4749 9. Added code to Exim itself and to exim_lock to test for a specific race
4750 condition that could lead to file corruption when using MBX delivery. The
4751 issue is with the lockfile that is created in /tmp. If this file is removed
4752 after a process has opened it but before that process has acquired a lock,
4753 there is the potential for a second process to recreate the file and also
4754 acquire a lock. This could lead to two Exim processes writing to the file
4755 at the same time. The added code performs the same test as UW imapd; it
4756 checks after acquiring the lock that its file descriptor still refers to
4757 the same named file.
4759 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4760 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4763 11. Added the smtp_active_hostname option. If used, this will typically be made
4764 to depend on the incoming interface address. Because $interface_address is
4765 not set up until the daemon has forked a reception process, error responses
4766 that can happen earlier (such as "too many connections") no longer contain
4769 12. If an expansion in a condition on a "warn" statement fails because a lookup
4770 defers, the "warn" statement is abandoned, and the next ACL statement is
4771 processed. Previously this caused the whole ACL to be aborted.
4773 13. Added the iplsearch lookup type.
4775 14. Added ident_timeout as a log selector.
4777 15. Added tls_certificate_verified as a log selector.
4779 16. Added a global option tls_require_ciphers (compare the smtp transport
4780 option of the same name). This controls incoming TLS connections.
4782 17. I finally figured out how to make tls_require_ciphers do a similar thing
4783 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4784 before starting the TLS session.
4786 18. Tabs are now shown as \t in -bP output.
4788 19. If the log selector return_path_on_delivery was set, Exim crashed when
4789 bouncing a message because it had too many Received: header lines.
4791 20. If two routers both had headers_remove settings, and the first one included
4792 a superfluous trailing colon, the final name in the first list and the
4793 first name in the second list were incorrectly joined into one item (with a
4794 colon in the middle).
4800 1. Added -C and -D options to the exinext utility, mainly to make it easier
4801 to include in the automated testing, but these could be helpful when
4802 multiple configurations are in use.
4804 2. The exinext utility was not formatting the output nicely when there was
4805 an alternate port involved in the retry record key, nor when there was a
4806 message id as well (for retries that were specific to a specific message
4807 and a specific host). It was also confused by IPv6 addresses, because of
4808 the additional colons they contain. I have fixed the IPv4 problem, and
4809 patched it up to do a reasonable job for IPv6.
4811 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4812 delivery, the log line now contains "pipelined" if PIPELINING was used.
4814 4. An SMTP transport process used to panic and die if the bind() call to set
4815 an explicit outgoing interface failed. This has been changed; it is now
4816 treated in the same way as a connect() failure.
4818 5. A reference to $sender_host_name in the part of a conditional expansion
4819 that was being skipped was still causing a DNS lookup. This no longer
4822 6. The def: expansion condition was not recognizing references to header lines
4823 that used bh_ and bheader_.
4825 7. Added the _cache feature to named lists.
4827 8. The code for checking quota_filecount in the appendfile transport was
4828 allowing one more file than it should have been.
4830 9. For compatibility with Sendmail, the command line option
4838 and sets the incoming protocol and host name (for trusted callers). The
4839 host name and its colon can be omitted when only the protocol is to be set.
4840 Note the Exim already has two private options, -pd and -ps, that refer to
4841 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4842 "s", but I don't think that's a major issue.
4844 10. A number of refactoring changes to the code, none of which should affect
4847 (a) The number of logging options was getting close to filling up the
4848 32-bit word that was used as a bit map. I have split them into two classes:
4849 those that are passed in the argument to log_write(), and those that are
4850 only ever tested independently outside of that function. These are now in
4851 separate 32-bit words, so there is plenty of room for expansion again.
4852 There is no change in the user interface or the logging behaviour.
4854 (b) When building, for example, log lines, the code previously used a
4855 macro that called string_cat() twice, in order to add two strings. This is
4856 not really sufficiently general. Furthermore, there was one instance where
4857 it was actually wrong because one of the argument was used twice, and in
4858 one call a function was used. (As it happened, calling the function twice
4859 did not affect the overall behaviour.) The macro has been replaced by a
4860 function that can join an arbitrary number of extra strings onto a growing
4863 (c) The code for expansion conditions now uses a table and a binary chop
4864 instead of a serial search (which was left over from when there were very
4865 few conditions). Also, it now recognizes conditions like "pam" even when
4866 the relevant support is not compiled in: a suitably worded error message is
4867 given if an attempt is made to use such a condition.
4869 11. Added ${time_interval:xxxxx}.
4871 12. A bug was causing one of the ddress fields not to be passed back correctly
4872 from remote delivery subprocesses. The field in question was not being
4873 subsequently used, so this caused to problems in practice.
4875 13. Added new log selectors queue_time and deliver_time.
4877 14. Might have fixed a bug in maildirsizefile handling that threw up
4878 "unexpected character" debug warnings, and recalculated the data
4879 unnecessarily. In any case, I expanded the warning message to give more
4882 15. Added the message "Restricted characters in address" to the statements in
4883 the default ACL that block characters like @ and % in local parts.
4885 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4886 Three changes have been made:
4888 (a) There was a serious bug; a negative response to MAIL caused the whole
4889 recipient domain to be cached as invalid, thereby blocking all messages
4890 to all local parts at the same domain, from all senders. This bug has
4891 been fixed. The domain is no longer cached after a negative response to
4892 MAIL if the sender used is not empty.
4894 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4897 (c) A new callout option, "use_sender" has been added for people who want
4898 the modified behaviour.
4904 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4907 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4908 indeed breaks things for older releases.
4910 3. Added additional logging to the case where there is a problem reading data
4911 from a filter that is running in a subprocess using a pipe, in order to
4912 try to track down a specific problem.
4914 4. Testing facility fudge: when running in the test harness and attempting
4915 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4916 getting "No route to host". Convert this to a timeout.
4918 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4921 6. Some OS don't have socklen_t but use size_t instead. This affects the
4922 fifth argument of getsockopt() amongst other things. This is now
4923 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4924 can be set for individual OS. I have set it for SunOS5, OSF1, and
4925 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4926 some earlier ones do not.
4928 7. Change 4.30/15 was not doing the test caselessly.
4930 8. The standard form for an IPv6 address literal was being rejected by address
4931 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4932 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4933 this, as well as the form without the "IPv6" on the front (but only when
4934 address literals are enabled, of course).
4936 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4938 10. Exim crashed if a message with an empty sender address specified by -f
4939 encountered a router with an errors_to setting. This could be provoked only
4940 by a command such as
4944 where an empty string was supplied; "<>" did not hit this bug.
4946 11. Installed PCRE release 4.5.
4948 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4949 remained set. It is now erased.
4951 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4952 times from message ids (which are base 36 rather than the normal 62).
4954 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4955 were being counted as actual protocol errors, and logged if the log
4956 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4957 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4958 and DATA following a set of rejected RCPTs do not count as protocol errors.
4959 In other words, Exim assumes they were pipelined, though this may not
4960 actually be the case. Of course, in all cases the client gets an
4961 appropriate error code.
4963 15. If a lookup fails in an ACL condition, a message about the failure may
4964 be available; it is used if testing the ACL cannot continue, because most
4965 such messages specify what the cause of the deferral is. However, some
4966 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4967 that caused an old message to be retained and used if a later statement
4968 caused a defer, replacing the real cause of the deferral.
4970 16. If an IP address had so many PTR records that the DNS lookup buffer
4971 was not large enough to hold them, Exim could crash while trying to process
4972 the truncated data. It now detects and logs this case.
4974 17. Further to 4.21/58, another change has been made: if (and only if) the
4975 first line of a message (the first header line) ends with CRLF, a bare LF
4976 in a subsequent header line has a space inserted after it, so as not to
4977 terminate the header.
4979 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4980 unnecessarily, used atoi() instead of strtol(), and didn't check the
4981 termination when getting file sizes from file names by regex.
4983 19. Completely re-implemented the support for maildirsize files, in the light
4984 of a number of problems with the previous contributed implementation
4985 (4.30/29). In particular:
4987 . If the quota is zero, the maildirsize file is maintained, but no quota is
4990 . If the maildir directory does not exist, it is created before any attempt
4991 to write a maildirsize file.
4993 . The quota value in the file is just a cache; if the quota is changed in
4994 the transport, the new value overrides.
4996 . A regular expression is available for excluding directories from the
4999 20. The autoreply transport checks the characters in options that define the
5000 message's headers; it allows continued headers, but it was checking with
5001 isspace() after an embedded newline instead of explicitly looking for a
5004 21. If all the "regular" hosts to which an address was routed had passed their
5005 expiry times, and had not reached their retry times, the address was
5006 bounced, even if fallback hosts were defined. Now Exim should go on to try
5009 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5010 equivalent code in the SMTP transport. Some hosts send humungous responses
5011 to HELO/EHLO, more than 1024 it seems.
5013 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5014 gives compiler warnings in some environments. I've now done it "properly",
5017 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5018 (because of problems with the built-in one) was declared to return uschar *
5019 instead of char *, causing compiler failure.
5021 25. Fixed a file descriptor leak when processing alias/forward files.
5023 26. Fixed a minor format string issue in dbfn.c.
5025 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5027 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5028 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5031 29. When a sender address is verified, it is cached, to save repeating the test
5032 when there is more than one recipient in a message. However, when the
5033 verification involves a callout, it is possible for different callout
5034 options to be set for different recipients. It is too complicated to keep
5035 track of this in the cache, so now Exim always runs a verification when a
5036 callout is required, relying on the callout cache for the optimization.
5037 The overhead is duplication of the address routing, but this should not be
5040 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5041 to be verified with callout=postmaster, and the main callout worked but the
5042 postmaster check failed, the verification correctly failed. However, if a
5043 subsequent RCPT command asked for sender verification *without* the
5044 postmaster check, incorrect caching caused this verification also to fail,
5047 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5048 it was not caching the DNS options (qualify_single, search_parents) that
5049 were used when the lookup failed. A subsequent lookup with different
5050 options therefore always gave the same answer, though there were cases
5051 where it should not have. (Example: a "domains = !$mx_any" option on a
5052 dnslookup router: the "domains" option is always processed without any
5053 widening, but the router might have qualify_single set.) Now Exim uses the
5054 cached value only when the same options are set.
5056 32. Added John Jetmore's "exipick" utility to the distribution.
5058 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5059 than a timeout (e.g. a certificate is required, and is not provided), an
5060 Exim server now closes the connection immediately. Previously it waited for
5061 the client to close - but if the client is SSL, it seems that they each
5062 wait for each other, leading to a delay before one of them times out.
5064 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5065 maintained 0.8.x compatibility because I don't think many are using it, and
5066 it is clearly obsolete.
5068 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5071 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5072 client certificate was expired. A simple patch fixes this, though I don't
5073 understand the full logic of why the verify callback is called multiple
5076 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5077 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5078 which causes problems with some clients (such as the Certicom SSL Plus
5079 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5080 disables the coutermeasure allowing Eudora to connect."
5082 38. Exim was not checking that a write() to a log file succeeded. This could
5083 lead to Bad Things if a log got too big, in particular if it hit a file
5084 size limit. Exim now panics and dies if it cannot write to a log file, just
5085 as it does if it cannot open a log file.
5087 39. Modified OS/Makefile-Linux so that it now contains
5089 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5091 The two -D definitions ensure that Exim is compiled with large file
5092 support, which makes it possible to handle log files that are bigger than
5095 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5096 instance) a domain was checked against a named list that involved a lookup,
5097 causing $domain_data to be set, then another domain was checked against the
5098 same list, then the first domain was re-checked, the value of $domain_data
5099 after the final check could be wrong. In particular, if the second check
5100 failed, it could be set empty. This bug probably also applied to
5103 41. The strip_trailing_dot option was not being applied to the address given
5104 with the -f command-line option.
5106 42. The code for reading a message's header from the spool was incrementing
5107 $received_count, but never initializing it. This meant that the value was
5108 incorrect (doubled) while delivering a message in the same process in which
5109 it was received. In the most common configuration of Exim, this never
5110 happens - a fresh exec is done - but it can happen when
5111 deliver_drop_privilege is set.
5113 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5114 it now includes up to 150 characters of the unexpected data in the log
5117 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5118 and building data strings. The size of both of these buffers was 10 000
5119 bytes - far larger than anybody would *ever* want, thought I. Needless to
5120 say, somebody hit the limit. I have increased the maximum line length to
5121 20 000 and the maximum data length of concatenated lines to 100 000. I have
5122 also fixed two bugs, because there was no checking on these buffers. Tsk,
5123 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5124 buffer is too small.
5126 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5127 lsearch lookups. Now it does.
5129 46. When parsing a route_list item in a manualroute router, a fixed-length
5130 buffer was used for the list of hosts. I made this 1024 bytes long,
5131 thinking that nobody would ever have a list of hosts that long. Wrong.
5132 Somebody had a whole pile of complicated expansion conditions, and the
5133 string was silently truncated, leading to an expansion error. It turns out
5134 that it is easier to change to an unlimited length (owing to other changes
5135 that have happened since this code was originally written) than to build
5136 structure for giving a limitation error. The length of the item that
5137 expands into the list of hosts is now unlimited.
5139 47. The lsearch lookup could not handle data where the length of text line was
5140 more than 4095 characters. Such lines were truncated, leading to shortened
5141 data being returned. It should now handle lines of any length.
5143 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5144 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5147 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5148 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5149 no longer happens in comment lines. A list of replacements is now placed
5150 at the head of all of the source files, except those whose only change is
5151 to replace PERL_COMMAND in the very first #! line.
5153 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5154 messages on the queue, with a bottom-up merge sort, using code contributed
5155 by Michael Haardt. This should make operations like -bp somewhat faster on
5156 large queues. It won't affect queue runners, except when queue_run_in_order
5159 51. Installed eximstats 1.31 in the distribution.
5161 52. Added support for SRV lookups to the dnslookup router.
5163 53. If an ACL referred to $message_body or $message_body_end, the value was not
5164 reset for any messages that followed in the same SMTP session.
5166 54. The store-handling optimization for building very long strings was not
5167 differentiating between the different store pools. I don't think this
5168 actually made any difference in practice, but I've tidied it.
5170 55. While running the routers to verify a sender address, $sender_address
5171 was still set to the sender address. This is wrong, because when routing to
5172 send a bounce to the sender, it would be empty. Therefore, I have changed
5173 it so that, while verifying a sender address, $sender_address is set to <>.
5174 (There is no change to what happens when verifying a recipient address.)
5176 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5177 target A or AAAA records (if not already returned) without resetting the
5178 qualify_single or search_parents options of the DNS resolver. These are
5179 inappropriate in this case because the targets of MX and SRV records must
5180 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5181 target that, when qualified, matched something in the local domain. These
5182 two options are now turned off when doing these lookups.
5184 57. It seems that at least some releases of Reiserfs (which does not have the
5185 concept of a fixed number of inodes) returns zero and not -1 for the
5186 number of available inodes. This interacted badly with check_spool_inodes,
5187 which assumed that -1 was the "no such thing" setting. What I have done is
5188 to check that the total number of inodes is greater than zero before doing
5189 the test of how many are available.
5191 58. When a "warn" ACL statement has a log_message modifier, the message is
5192 remembered, and not repeated. This is to avoid a lot of repetition when a
5193 message has many recipients that cause the same warning to be written.
5194 Howewer, Exim was preserving the list of already written lines for an
5195 entire SMTP session, which doesn't seem right. The memory is now reset if a
5196 new message is started.
5198 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5199 debugging output unless log_rewrite was also set.
5201 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5202 of (int)(handle) when we know that handle contains (void *)(-1).
5204 61. The Exim daemon panic-logs an error return when it closes the incoming
5205 connection. However "connection reset by peer" seems to be common, and
5206 isn't really an error worthy of noting specially, so that particular error
5209 62. When Exim is trying to find all the local interfaces, it used to panic and
5210 die if the ioctl to get the interface flags failed. However, it seems that
5211 on at least one OS (Solaris 9) it is possible to have an interface that is
5212 included in the list of interfaces, but for which you get a failure error
5213 for this call. This happens when the interface is not "plumbed" into a
5214 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5215 failure of the "get flags" call assumes that the interface is down.
5217 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5218 makes life easier for people who are doing arithmetic on fields extracted
5219 from dates, where you often get leading zeros that should not be
5220 interpreted as octal.
5222 64. Added qualify_domain to the redirect router, to override the global
5225 65. If a pathologically long header line contained very many addresses (the
5226 report of this problem mentioned 10 000) and each of them was rewritten,
5227 Exim could use up a very large amount of memory. (It kept on making new
5228 copies of the header line as it rewrote, and never released the old ones.)
5229 At the expense of a bit more processing, the header rewriting function has
5230 been changed so that it no longer eats memory in this way.
5232 66. The generation of the Received: header has been moved from the time that a
5233 message starts to be received, to the time that it finishes. The timestamp
5234 in the Received: header should now be very close to that of the <= log
5235 line. There are two side-effects of this change:
5237 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5238 logged header lines no longer include the local Received: line, because
5239 it has not yet been created. The same applies to a copy of the message
5240 that is returned to a non-SMTP sender when a message is rejected.
5242 (b) When a filter file is tested using -bf, no additional Received: header
5243 is added to the test message. After some thought, I decided that this
5246 This change does not affect the value of $received_for. It is still set
5247 after address rewriting, but before local_scan() is called.
5249 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5251 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5252 gave an unhelpful panic error message, and a defer error. I have managed to
5253 change this behaviour so that it now rejects any supplied certificate,
5254 which seems right, as the list of acceptable certificates is empty.
5256 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5257 gave an unhelpful defer error. I have not managed to make this reject any
5258 supplied certificates, but the error message it gives is "no certificate
5259 supplied", which is not helpful.
5261 70. exigrep's output now also includes lines that are not associated with any
5262 message, but which match the given pattern. Implemented by a patch from
5263 Martin Sluka, which also tidied up the Perl a bit.
5265 71. Recipient callout verification, like sender verification, was using <> in
5266 the MAIL FROM command. This isn't really the right thing, since the actual
5267 sender may affect whether the remote host accepts the recipient or not. I
5268 have changed it to use the actual sender in the callout; this means that
5269 the cache record is now keyed on a recipient/sender pair, not just the
5270 recipient address. There doesn't seem to be a real danger of callout loops,
5271 since a callout by the remote host to check the sender would use <>.
5272 [SEE ABOVE: changed after hitting problems.]
5274 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5275 temporary errors. However, in the case of such a code being given after
5276 the end of a data transmission (i.e. after ".") Exim was failing to write
5277 a retry record for the message. (Yes, there was some broken host that was
5278 actually sending 8xx at this point.)
5280 73. An unknown lookup type in a host list could cause Exim to panic-die when
5281 the list was checked. (An example that provoked this was putting <; in the
5282 middle of a list instead of at the start.) If this happened during a DATA
5283 ACL check, a -D file could be left lying around. This kind of configuration
5284 error no longer causes Exim to die; instead it causes a defer errror. The
5285 incident is still logged to the main and panic logs.
5287 74. Buglet left over from Exim 3 conversion. The message "too many messages
5288 in one connection" was written to the rejectlog but not the mainlog, except
5289 when address rewriting (yes!) was being logged.
5291 75. Added write_rejectlog option.
5293 76. When a system filter was run not as root (that is, when system_filter_user
5294 was set), the values of the $n variables were not being returned to the
5295 main process; thus, they were not subsequently available in the $sn
5298 77. Added +return_path_on_delivery log selector.
5300 78. A connection timeout was being treated differently from recipients deferred
5301 when testing hosts_max_try with a message that was older than the host's
5302 retry timeout. (The host should not be counted, thus allowing all hosts to
5303 be tried at least once before bouncing.) This may have been the cause of an
5304 occasionally reported bug whereby a message would remain on the queue
5305 longer than the retry timeout, but would be bounced if a delivery was
5306 forced. I say "may" because I never totally pinned down the problem;
5307 setting up timeout/retry tests is difficult. See also the next item.
5309 79. The ultimate address timeout was not being applied to errors that involved
5310 a combination of host plus message (for example, a timeout on a MAIL
5311 command). When an address resolved to a number of possible hosts, and they
5312 were not all tried for each delivery (e.g. because of hosts_max_try), a
5313 message could remain on the queue longer than the retry timeout.
5315 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5318 81. Fixed an obscure SMTP outgoing bug which required at least the following
5319 conditions: (a) there was another message waiting for the same server;
5320 (b) the server returned 5xx to all RCPT commands in the first message so
5321 that the message was not completed; (c) the server dropped the connection
5322 or gave a negative response to the RSET that Exim sends to abort the
5323 transaction. The observed case was a dropped connection after DATA that had
5324 been sent in pipelining mode. That is, the server had advertised PIPELINING
5325 but was not implementing it correctly. The effect of the bug was incorrect
5326 behaviour, such as trying another host, and this could lead to a crash.
5332 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5333 and daemon.c were passed as pointers to ints; they should have been
5334 pointers to socklen_t variables (which are typically unsigned ints).
5336 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5339 3. Fixed a really odd bug that affected only the testing scheme; patching a
5340 certain fixed string in the binary changed the value of another string that
5341 happened to be identical to the end of the original first string.
5343 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5344 name", it returns that address as the IP address. On some operating
5345 systems (e.g. Solaris), it also passes back the IP address string as the
5346 "host name". However, on others (e.g. Linux), it passes back an empty
5347 string. Exim wasn't checking for this, and was changing the host name to an
5348 empty string, assuming it had been canonicized.
5350 5. Although rare, it is permitted to have more than one PTR record for a given
5351 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5352 all the names associated with an address, because they do in Solaris.
5353 However, it seems that they do not in Linux for data that comes from the
5354 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5355 I found this out when I moved to a new Linux workstation and tried to run
5356 the Exim test suite.
5358 To get round this problem I have changed the code so that it now does its
5359 own call to the DNS to look up PTR records when searching for a host name.
5360 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5361 addresses that are only in /etc/hosts are still found.
5363 This behaviour is, however, controlled by an option called host_lookup_
5364 order, which defaults to "bydns:byaddr". If people want to use the other
5365 order, or indeed, just use one or the other means of lookup, they can
5366 specify it in this variable.
5368 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5369 some operating systems, this comes back from gethostbyaddr() as an empty
5370 string, and this is what Exim used to test for. However, it seems that in
5371 other systems, "." is yielded. Exim now tests for this case too.
5373 7. The values of check_spool_space and check_log_space are now held internally
5374 as a number of kilobytes instead of an absolute number of bytes. If a
5375 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5376 kilobyte. This means that much larger values can be stored.
5378 8. Exim monitor: an attempt to get the action menu when not actually pointing
5379 at a message produces an empty menu entitled "No message selected". This
5380 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5381 no entries in it ("Shell widget menu has zero width and/or height"). So I
5382 have added a single, blank menu entry in this case.
5384 9. Added ${quote_local_part.
5386 10. MIME decoding is now applied to the contents of Subject: header lines when
5389 11. Now that a reference to $sender_host_address automatically causes a reverse
5390 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5391 host lookup before query-style lookups in lists that might use this
5392 variable. This has therefore been abolished, and the "net-" prefix is no
5393 longer necessary for query-style lookups.
5395 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5396 have been a typo for LFLAGS, so it has been changed.
5398 13. The install script calls Exim with "-C /dev/null" in order to find the
5399 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5400 to be output. Howeve, since Exim outputs its version number before the
5401 error, it didn't break the script. It just looked ugly. I fixed this by
5402 always allowing "-C /dev/null" if the caller is root.
5404 14. Ignore overlarge ACL variable number when reading spool file - insurance
5405 against a later release with more variables having written the file.
5407 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5408 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5409 this, as well as the form without the "IPv6" on the front.
5411 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5412 OS/Makefile-Darwin file.
5414 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5415 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5417 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5418 variable. However, some OS use ints for inodes. I've added cast to long int
5419 to get rid of the compiler warning.
5421 19. I had forgotten to lock out "/../" in configuration file names when
5422 ALT_CONFIG_PREFIX was set.
5424 20. Routers used for verification do not need to specify transports. However,
5425 if such a router generated a host list, and callout was configured, Exim
5426 crashed, because it could not find a port number from the (non-existent)
5427 transport. It now assumes port 25 in this circumstance.
5429 21. Added the -t option to exigrep.
5431 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5432 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5433 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5434 them is set, LOOKUP_LSEARCH is forced.
5436 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5438 24. Added sender and host information to the "rejected by local_scan()" log
5439 line; previously there was no indication of these.
5441 25. Added .include_if_exists.
5443 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5444 Linux. It turns out that not all file systems support this. Apparently some
5445 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5446 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5447 not supported on the file descriptor, is now ignored when Exim is trying to
5448 sync a directory. This applies only to Linux.
5450 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5452 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5453 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5456 29. Added support for maildirsize files from supplied patch (modified a bit).
5458 30. The use of :fail: followed by an empty string could lead Exim to respond to
5459 sender verification failures with (e.g.):
5461 550 Verification failed for <xxx>
5462 550 Sender verify failed
5464 where the first response line was missing the '-' that indicates it is not
5465 the final line of the response.
5467 31. The loop for finding the name of the user that called Exim had a hardwired
5468 limit of 10; it now uses the value of finduser_retries, which is used for
5469 all other user lookups.
5471 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5474 33. Exim was neglecting to zero errno before one call of strtol() when
5475 expanding a string and expecting an integer value. On some systems this
5476 resulted in spurious "integer overflow" errors. Also, it was casting the
5477 result into an int without checking.
5479 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5480 did not work. The code looks as if it has *never* worked, though it appears
5481 to have been documented since at least releast 1.62. I have made it work.
5483 35. The "timeout_DNS" error in retry rules, also documented since at least
5484 1.62, also never worked. As it isn't clear exactly what this means, and
5485 clearly it isn't a major issue, I have abolished the feature by treating it
5486 as "timeout", and writing a warning to the main and panic logs.
5488 36. The display of retry rules for -brt wasn't always showing the error code
5491 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5492 timeout_connect_A, timeout_connect_MX.
5494 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5495 to the empty sender.
5497 39. The daemon was not analysing the content of -oX till after it had closed
5498 stderr and disconnected from the controlling terminal. This meant that any
5499 syntax errors were only noted on the panic log, and the return code from
5500 the command was 0. By re-arranging the code a little, I've made the
5501 decoding happen first, so such errors now appear on stderr, and the return
5502 code is 1. However, the actual setting up of the sockets still happens in
5503 the disconnected process, so errors there are still only recorded on the
5506 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5507 connections (as happens on some IP stacks) was logged at start up time as
5508 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5509 it from "IPv6 and IPv4", which means that two separate sockets are being
5512 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5513 says whether AF_INET or AF_INET6 was passed as an argument.
5515 42. Exiwhat output was messed up when time zones were included in log
5518 43. Exiwhat now gives more information about the daemon's listening ports,
5519 and whether -tls-on-connect was used.
5521 44. The "port" option of the smtp transport is now expanded.
5523 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5524 silently ignored. Now an error message is written to the main and panic
5527 46. There's a new ACL modifier called "logwrite" which writes to a log file
5528 as soon as it is encountered.
5530 47. Added $local_user_uid and $local_user_gid at routing time.
5532 48. Exim crashed when trying to verify a sender address that was being
5535 49. Exim was recognizing only a space character after ".include". It now also
5536 recognizes a tab character.
5538 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5539 extracting the relevant information from the specification. The man page no
5540 longer contains scrambled data for the -d option, and I've added a section
5541 at the front about calling Exim under different names.
5543 51. Added "extra_headers" argument to the "mail" command in filter files.
5545 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5548 53. Installed eximstats 1.29.
5550 54. Added transport_filter_timeout as a generic transport option.
5552 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5553 Cc: header lines. This was required by RFC 822, but it not required by RFC
5556 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5557 incoming messages that did not have them. Now it does so only if the
5558 message originates locally, that is, if there is no associated remote host
5559 address. When Resent- header lines are present, this applies to the Resent-
5560 lines rather than the non-Resent- lines.
5562 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5563 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5565 58. Messages for configuration errors now include the name of the main
5566 configuration file - useful now that there may be more than one file in a
5567 list (.included file names were always shown).
5569 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5570 for those rare installations that do not start the daemon as root or run it
5571 setuid root. I've cut out the call to initgroups() if the daemon is not
5574 60. The Exim user and group can now be bound into the binary as text strings
5575 that are looked up at the start of Exim's processing.
5577 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5579 62. Added $mailstore_basename variable.
5581 63. Installed patch to sieve.c from Michael Haardt.
5583 64. When Exim failed to open the panic log after failing to open the main log,
5584 the original message it was trying to log was written to stderr and debug
5585 output, but if they were not available (the usual case in production), it
5586 was lost. Now it is written to syslog before the two lines that record the
5587 failures to open the logs.
5589 65. Users' Exim filters run in subprocesses under the user's uid. It is
5590 possible for a "deliver" command or an alias in a "personal" command to
5591 provoke an address rewrite. If logging of address rewriting is configured,
5592 this fails because the process is not running as root or exim. There may be
5593 a better way of dealing with this, but for the moment (because 4.30 needs
5594 to be released), I have disabled address rewrite logging when running a
5595 filter in a non-root, non-exim process.
5601 1. The buildconfig auxiliary program wasn't quoting the value set for
5602 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5603 not defined. This bug was masked in 4.22 by the effect that was fixed in
5606 2. Some messages that were rejected after a message id was allocated were
5607 shown as "incomplete" by exigrep. It no longer does this for messages that
5608 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5610 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5611 have allow_domain_literals set, the ID did not get logged in the <= line.
5612 Domain literals are now always recognized in Message-ID: header lines.
5614 4. The first argument for a ${extract expansion item is the key name or field
5615 number. Leading and trailing spaces in this item were not being ignored,
5616 causing some misleading effects.
5618 5. When deliver_drop_privilege was set, single queue runner processes started
5619 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5620 same command in the process it spins off) were not dropping privilege.
5622 6. When the daemon running as "exim" started a queue runner, it always
5623 re-executed Exim in the spun-off process. This is a waste of effort when
5624 deliver_drop_privilege is set. The new process now just calls the
5625 queue-runner function directly.
5631 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5634 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5635 ignored. Though the use of -oP was forcing the writing of a pid file, it
5636 was always written to the default place.
5638 3. If the message "no IP address found for host xxxx" is generated during
5639 incoming verification, it is now followed by identification of the incoming
5640 connection (so you can more easily find what provoked it).
5642 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5644 5. Added some features to "harden" Exim a bit more against certain attacks:
5646 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5647 be put in Local/Makefile. This is like the never_users runtime option,
5648 but it cannot be overridden. The default setting is "root".
5650 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5651 prefix string with which any file named in a -C command line option
5654 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5655 is retained for -C and -D only if the caller of Exim is root. Without
5656 it, the exim user may also use -C and -D and retain privilege.
5658 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5659 command line option is disabled.
5661 6. Macro names set by the -D option must start with an upper case letter, just
5662 like macro names defined in the configuration file.
5664 7. Added "dereference=" facility to LDAP.
5666 8. Two instances of the typo "uknown" in the source files are fixed.
5668 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5669 the Configure-Makefile script screwed up while processing it.
5671 10. Incorporated PCRE 4.4.
5673 11. The SMTP synchronization check was not operating right at the start of an
5674 SMTP session. For example, it could not catch a HELO sent before the client
5675 waited for the greeting. There is now a check for outstanding input at the
5676 point when the greeting is written. Because of the duplex, asynchronous
5677 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5678 way, but not yet received, when the check is performed.
5680 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5681 on TCP/IP sockets, because this apparently causes some broken clients to
5684 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5685 unchanged) from the Cygwin maintainer.
5687 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5688 was supported instead of testing for mbx. Effectively a typo.
5690 15. The spa authenticator server code was not checking that the input it
5691 received was valid base64.
5693 16. The debug output line for the "set" modifier in ACLs was not showing the
5694 name of the variable that was being set.
5696 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5698 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5699 lookup on its own. Something else was needed to trigger the lookup. For
5700 example, a match in host_lookup or the need for a host name in a host list.
5701 Now, if $sender_host_name is referenced and the host name has not yet been
5702 looked up, a lookup is performed. If the lookup fails, the variable remains
5703 empty, and $host_lookup_failed is set to "1".
5705 19. Added "eqi" as a case-independent comparison operator.
5707 20. The saslauthd authentication condition could segfault if neither service
5708 nor realm was specified.
5710 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5711 error message that was logged was misleading, and incoming SMTP
5712 connections were dropped. The message is now more accurate, and temporary
5713 errors are given to SMTP connections.
5715 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5716 (except RSET) with a 421 error, until QUIT is received. However, it was
5717 failing to send a response to QUIT.
5719 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5720 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5721 ACL is now run after the helo_try_verify_hosts code.
5723 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5724 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5725 case-independent, so other case variants are also recognized). Apparently
5726 some systems use these upper case variants.
5728 25. If more than two messages were waiting for the same host, and a transport
5729 filter was specified for the transport, Exim sent two messages over the
5730 same TCP/IP connection, and then failed with "socket operation on non-
5731 socket" when it tried to send the third.
5733 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5735 27. The extern definition of crypt16() in expand.c was not being excluded when
5736 the OS had its own crypt16() function.
5738 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5739 as a preferred synonym for return_size_limit, both as an option and as an
5742 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5744 30. Changed the default configuration ACL to relax the local part checking rule
5745 for addresses that are not in any local domains. For these addresses,
5746 slashes and pipe symbols are allowed within local parts, but the sequence
5747 /../ is explicitly forbidden.
5749 31. SPA server authentication was not clearing the challenge buffer before
5752 32. log_message in a "warn" ACL statement was writing to the reject log as
5753 well as to the main log, which contradicts the documentation and doesn't
5754 seem right (because no rejection is happening). So I have stopped it.
5756 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5757 However, I am unable to do any testing of this.
5759 34. Fixed an infelicity in the appendfile transport. When checking directories
5760 for a mailbox, to see if any needed to be created, it was accidentally
5761 using path names with one or more superfluous leading slashes; tracing
5762 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5764 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5765 contents are added to the log line that is written for every discarded
5766 recipient. (Previously a log_message setting was ignored.)
5768 36. The ${quote: operator now quotes the string if it is empty.
5770 37. The install script runs exim in order to find its version number. If for
5771 some reason other than non-existence or emptiness, which it checks, it
5772 could not run './exim', it was installing it with an empty version number,
5773 i.e. as "exim-". This error state is now caught, and the installation is
5776 38. An argument was missing from the function that creates an error message
5777 when Exim fails to connect to the socket for saslauthd authentication.
5778 This could cause Exim to crash, or give a corrupted message.
5780 39. Added isip, isip4, and isip6 to ${if conditions.
5782 40. The ACL variables $acl_xx are now saved with the message, and can be
5783 accessed later in routers, transports, and filters.
5785 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5786 strings in the file are not string-expanded.
5788 42. If a MAIL command specified a SIZE value that was too large to fit into an
5789 int variable, the check against message_size_limit failed. Such values are
5790 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5791 day this will have to be increased, but I don't think I want to be around
5792 when emails are that large.
5799 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5800 iconv() is not standard in FreeBSD.
5802 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5803 IPv6 enabled. The observed symptom was a segmentation fault on return from
5804 the function os_common_find_running_interfaces() in src/os.c.
5806 3. In the check_special_case() function in daemon.c I had used "errno" as an
5807 argument name, which causes warnings on some systems. This was basically a
5808 typo, since it was named "eno" in the comments!
5810 4. The code that waits for the clock to tick (at a resolution of some fraction
5811 of a second) so as to ensure message-id uniqueness was always waiting for
5812 at least one whole tick, when it could have waited for less. [This is
5813 almost certainly not relevant at current processor speeds, where it is
5814 unlikely to ever wait at all. But we try to future-proof.]
5816 5. The function that sleeps for a time interval that includes fractions of a
5817 second contained a race. It did not block SIGALRM between setting the
5818 timer, and suspending (a couple of lines later). If the interval was short
5819 and the sigsuspend() was delayed until after it had expired, the suspension
5820 never ended. On busy systems this could lead to processes getting stuck for
5823 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5824 process, before it forks any delivery processes. The open lookup caching
5825 mechanism meant that the open file or database connection was passed into
5826 the delivery process. The problem was that delivery processes always tidy
5827 up cached lookup data. This could cause a problem for the next delivery
5828 process started by the queue runner, because the external queue runner
5829 process does not know about the closure. So the next delivery process
5830 still has data in the lookup cache. In the case of a file lookup, there was
5831 no problem because closing a file descriptor in a subprocess doesn't affect
5832 the parent. However, if the lookup was caching a connection to a database,
5833 the connection was closed, and the second delivery process was likely to
5834 see errors such as "PGSQL: query failed: server closed the connection
5835 unexpectedly". The problem has been fixed by closing all cached lookups
5836 in a queue runner before running a delivery process.
5838 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5839 seem to have the "const" qualifier which it has on other OS. I've
5842 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5843 *of the same type* (client or server) with the same public name that an
5844 error should be diagnosed.
5846 9. When Exim looked up a host name for an IP address, but failed to find the
5847 original IP address when looking up the host name (a safety check), it
5848 output the message "<ip address> does not match any IP for NULL", which was
5849 confusing, to say the least. The bug was that the host name should have
5850 appeared instead of "NULL".
5852 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5853 user that is built into the binary, and the -C or -D options is used, root
5854 privilege is dropped before the configuration file is read. In addition,
5855 logging is switched to stderr instead of the normal log files. If the
5856 configuration then re-defines the Exim user, the unprivileged environment
5857 is probably not what is expected, so Exim logs a panic warning message (but
5860 However, if deliver_drop_privilege is set, the unprivileged state may well
5861 be exactly what is intended, so the warning has been cut out in that case,
5862 and Exim is allowed to try to write to its normal log files.
5868 1. smtp_return_error_details was not giving details for temporary sender
5869 or receiver verification errors.
5871 2. Diagnose a configuration error if two authenticators have the same public
5874 3. Exim used not to create the message log file for a message until the first
5875 delivery attempt. This could be confusing when incoming messages were held
5876 for policy or load reasons. The message log file is now created at the time
5877 the message is received, and an initial "Received" line is written to it.
5879 4. The automatically generated man page for command line options had a minor
5880 bug that caused no ill effects; however, a more serious problem was that
5881 the procedure for building the man page automatically didn't always
5882 operate. Consequently, release 4.20 contains an out-of-date version. This
5883 shouldn't happen again.
5885 5. When building Exim with embedded Perl support, the script that builds the
5886 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5887 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5889 6. The freeze_tell option was not being used for messages that were frozen on
5890 arrival, either by an ACL or by local_scan().
5892 7. Added the smtp_incomplete_transaction log selector.
5894 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5895 was accepting AUTH without a new EHLO.
5897 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5898 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5901 10. Logging of TCP/IP connections (when configured) now happens in the main
5902 daemon process instead of the child process, so that the TCP/IP connection
5903 count is more accurate (but it can never be perfect).
5905 11. The use of "drop" in a nested ACL was not being handled correctly in the
5906 outer ACL. Now, if condition failure induced by the nested "drop" causes
5907 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5908 or "require"), the connection is dropped.
5910 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5911 that yield "discard" can now be used with an "accept" or a "discard" verb,
5912 but an error is generated for any others (because I can't see a useful way
5913 to define what should happen).
5915 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5916 are now processed in the same way as lines in the Exim configuration file.
5917 In particular, continuation lines are supported.
5919 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5921 15. Added -ti meaning -t -i.
5923 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5924 domains, and warn by logging if others are found.
5926 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5927 returned from GIFCONF when Exim is trying to find the list of interfaces on
5928 a host. The code in os.c has been modified to copy each ifreq to an aligned
5929 structure in all cases.
5931 Also, in some cases, the returned ifreq's were being copied to a 'struct
5932 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5933 means the last couple of bytes of an IPv6 address could be chopped if the
5934 ifreq contained only a normal sockaddr (14 bytes storage).
5936 18. Named domain lists were not supported in the hosts_treat_as_local option.
5937 An entry such as +xxxx was not recognized, and was treated as a literal
5940 19. Ensure that header lines added by a DATA ACL are included in the reject log
5941 if the ACL subsequently rejects the message.
5943 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5944 MD5 (which is deprecated).
5946 21. When testing a filter file using -bf, Exim was writing a message when it
5947 took the sender from a "From " line in the message, but it was not doing so
5948 when it took $return_path from a Return-Path: header line. It now does.
5950 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5951 with a valid header line field name (a series of printing characters
5952 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5954 23. Changed "disc" in the source to "disk" to conform to the documentation and
5955 the book and for uniformity.
5957 24. Ignore Sendmail's -Ooption=value command line item.
5959 25. When execve() failed while trying to run a command in a pipe transport,
5960 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5961 could be confused with a return value of 69 from the command itself. This
5962 has been changed to 127, the value the shell returns if it is asked to run
5963 a non-existent command. The wording for the related log line suggests a
5964 non-existent command as the problem.
5966 26. If received_header_text expands to an empty string, do not add a Received:
5967 header line to the message. (Well, it adds a token one on the spool, but
5968 marks it "old" so that it doesn't get used or transmitted.)
5970 27. Installed eximstats 1.28 (addition of -nt option).
5972 28. There was no check for failure on the call to getsockname() in the daemon
5973 code. This can fail if there is a shortage of resources on the system, with
5974 ENOMEM, for example. A temporary error is now given on failure.
5976 29. Contrary to the C standard, it seems that in some environments, the
5977 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5978 program. Exim now does this explicitly; it affects the formatting of
5979 timestamps using strftime().
5981 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5982 complaints. I've now initialized all the variables, to avoid this.
5984 32. Header lines added by a system filter were not being "seen" during
5985 transport-time rewrites.
5987 33. The info_callback() function passed to OpenSSL is set up with type void
5988 (*)(SSL *, int, int), as described somewhere. However, when calling the
5989 function (actually a macro) that sets it up, the type void(*)() is
5990 expected. I've put in a cast to prevent warnings from picky compilers.
5992 34. If a DNS black list lookup found a CNAME record, but there were no A
5993 records associated with the domain it pointed at, Exim crashed.
5995 35. If a DNS black list lookup returned more than one A record, Exim ignored
5996 all but the first. It now scans all returned addresses if a particular IP
5997 value is being sought. In this situation, the contents of the
5998 $dnslist_value variable are a list of all the addresses, separated by a
6001 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6002 to accept a host name and all its aliases if the forward lookup for any of
6003 them yielded the IP address of the incoming connection. Now it accepts only
6004 those names whose forward lookup yields the correct IP address. Any other
6005 names are discarded. This closes a loophole whereby a rogue DNS
6006 administrator could create reverse DNS records to break through a
6007 wildcarded host restriction in an ACL.
6009 37. If a user filter or a system filter that ran in a subprocess used any of
6010 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6011 the wrong values were passed to the pipe command ($thisaddress had the
6012 value of $0, $0 had the value of $1, etc). This bug was introduced by
6013 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6015 38. Improved the line breaking for long SMTP error messages from ACLs.
6016 Previously, if there was no break point between 40 and 75 characters, Exim
6017 left the rest of the message alone. Two changes have been made: (a) I've
6018 reduced the minimum length to 35 characters; (b) if it can't find a break
6019 point between 35 and 75 characters, it looks ahead and uses the first one
6020 that it finds. This may give the occasional overlong line, but at least the
6021 remaining text gets split now.
6023 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6024 file descriptors might be low, and that setting 1000 would always raise it.
6025 It turns out that in some environments, the limit is already over 1000 and
6026 that lowering it causes trouble. So now Exim takes care not to decrease it.
6028 40. When delivering a message, the value of $return_path is set to $sender_
6029 address at the start of routing (routers may change the value). By an
6030 oversight, this default was not being set up when an address was tested by
6031 -bt or -bv, which affected the outcome if any router or filter referred to
6034 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6035 log only when "message" or "log_message" are set. However, if one of the
6036 conditions was an address verification, or a call to a nested ACL, the
6037 messages generated by the underlying test were being passed through. This
6038 no longer happens. The underlying message is available in $acl_verify_
6039 message for both "message" and "log_message" expansions, so it can be
6040 passed through if needed.
6042 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6043 new expansion $bh_ to give the encoded byte string without charset
6044 translation. Translation happens only if iconv() is available; HAVE_ICONV
6045 indicates this at build time. HEADERS_CHARSET gives the charset to
6046 translate to; headers_charset can change it in the configuration, and
6047 "headers charset" can change it in an individual filter file.
6049 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6050 that creates RFC 2047 encoded "words" labels them as that charset instead
6051 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6052 expansion operator; (ii) when Exim creates a From: line for a local
6053 message; (iii) when a header line is rewritten to include a "phrase" part.
6055 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6056 buggy, causing it to skip the first lines of messages whose message ID
6057 ended in 'D'. This would not have bitten before Exim release 4.14, because
6058 message IDs were unlikely to end in 'D' before then. The effect was to have
6059 incorrect size information for certain domains.
6061 45. #include "config.h" was missing at the start of the crypt16.c module. This
6062 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6065 46. If there was a timeout during a "random" callout check, Exim treated it as
6066 a failure of the random address, and carried on sending RSET and the real
6067 address. If the delay was just some slowness somewhere, the response to the
6068 original RCPT would be taken as a response to RSET and so on, causing
6069 mayhem of various kinds.
6071 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6072 when I implemented it. It didn't allow for the fact that some option values
6073 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6074 the right test for positive values.
6076 48. Domain names in DNS records are case-independent. Exim always looks them up
6077 in lower case. Some resolvers return domain names in exactly the case they
6078 appear in the zone file, that is, they may contain uppercase letters. Not
6079 all resolvers do this - some return always lower case. Exim was treating a
6080 change of case by a resolver as a change of domain, similar to a widening
6081 of a domain abbreviation. This triggered its re-routing code and so it was
6082 trying to route what was effectively the same domain again. This normally
6083 caused routing to fail (because the router wouldn't handle the domain
6084 twice). Now Exim checks for this case specially, and just changes the
6085 casing of the domain that it ultimately uses when it transmits the message
6088 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6091 50. If a filter generated a file delivery with a non-absolute name (possible if
6092 no home directory exists for the router), the forbid_file option was not
6095 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6096 the existing equality matching.
6098 52. Exim was using ints instead of ino_t variables in some places where it was
6099 dealing with inode numbers.
6101 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6102 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6103 called TMPDIR, and if it finds it is different, it changes its value.
6105 54. The smtp_printf() function is now made available to local_scan() so
6106 additional output lines can be written before returning. There is also an
6107 smtp_fflush() function to enable the detection of a dropped connection.
6108 The variables smtp_input and smtp_batched_input are exported to
6111 55. Changed the default runtime configuration: the message "Unknown user"
6112 has been removed from the ACL, and instead placed on the localuser router,
6113 using the cannot_route_message feature. This means that any verification
6114 failures that generate their own messages won't get overridden. Similarly,
6115 the "Unrouteable address" message that was in the ACL for unverifiable
6116 relay addresses has also been removed.
6118 56. Added hosts_avoid_esmtp to the smtp transport.
6120 57. The exicyclog script was not checking for the esoteric option
6121 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6122 will work only if exicyclog is run under the appropriate euid.
6124 58. Following a discussion on the list, the rules by which Exim recognises line
6125 endings on incoming messages have been changed. The -dropcr and drop_cr
6126 options are now no-ops, retained only for backwards compatibility. The
6127 following line terminators are recognized: LF CRLF CR. However, special
6128 processing applies to CR:
6130 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6131 nor a local message in the state where . is a terminator.
6133 (ii) If a bare CR is encountered in a header line, an extra space is added
6134 after the line terminator so as not to end the header. The reasoning
6135 behind this is that bare CRs in header lines are most likely either
6136 to be mistakes, or people trying to play silly games.
6138 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6139 was being incorrectly given as 18 bytes larger than it should have been.
6140 This is a VOB (very old bug).
6142 60. This may never have affected anything current, but just in case it has:
6143 When the local host is found other than at the start of a list of hosts,
6144 the local host, those with the same MX, and any that follow, are discarded.
6145 When the list in question was part of a longer list of hosts, the following
6146 hosts (not currently being processed) were also being discarded. This no
6147 longer happens. I'm not sure if this situation could ever has previously
6150 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6153 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6154 header. This is something that is recommended in a new Internet Draft, and
6155 is something that is documented as being done by Sendmail. There are two
6156 possible values. For messages generated by the autoreply transport, Exim
6159 Auto-Submitted: auto-replied
6161 whereas for all other generated messages (e.g. bounces) it adds
6163 Auto-Submitted: auto-generated
6165 63. The "personal" condition in filters now includes a test for the
6166 Auto-Submitted: header. If it contains the string "auto-" the message it
6167 not considered personal.
6169 64. Added rcpt_include_affixes as a generic transport option.
6171 65. Added queue_only_override (default true).
6173 66. Added the syslog_duplication option.
6175 67. If what should have been the first header line of a message consisted of
6176 a space followed by a colon, Exim was mis-interpreting it as a header line.
6177 It isn't of course - it is syntactically invalid and should therefore be
6178 treated as the start of the message body. The misbehaviour could have
6179 caused a number of strange effects, including loss of data in subsequent
6180 header lines, and spool format errors.
6182 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6183 client host had authenticated. This control can now be exercised by an ACL
6184 for more flexibility.
6186 69. By default, callouts do not happen when testing with -bh. There is now a
6187 variant, -bhc, which does actually run the callout code, including
6188 consulting and updating the callout cache.
6190 70. Added support for saslauthd authentication, courtesy of Alexander
6193 71. If statvfs() failed on the spool or log directories while checking their
6194 size for availability, Exim confusingly gave the error "space shortage".
6195 Furthermore, in debugging mode it crashed with a floating point exception.
6196 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6197 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6198 serious problem, Exim now writes to the main and panic logs when this
6199 happens, with details of the failure. It then refuses to accept the
6200 incoming message, giving the message "spool directory problem" or "log
6201 directory problem" with a 421 code for SMTP messages.
6203 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6204 0, 1, and 2 exist, because some OS complain for execs without them (see
6205 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6206 descriptors. However, the code omitted to check that the open succeeded,
6207 causing mysterious errors if for some reason the permissions on /dev/null
6208 got screwed. Now Exim writes a message to the main and panic logs, and
6209 bombs out if it can't open /dev/null.
6211 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6212 interact so that it is all more flexible. It is supposed to remain
6213 backwards compatible. Also added extra_local_interfaces.
6215 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6216 to bomb out with an assertion failure - to the client this appears as a
6217 connection drop. This problem occurs in the part of the code that was taken
6218 from the Samba project. Fortunately, the assertion is in a very simple
6219 function, so I have fixed this by reproducing the function inline in the
6220 one place where it is called, and arranging for authentication to fail
6221 instead of killing the process with assert().
6223 75. The SPA client code was not working when the server requested OEM rather
6224 than Unicode encoding.
6226 76. Added code to make require_files with a specific uid setting more usable in
6227 the case where statting the file as root fails - usually a non-root-mounted
6228 NFS file system. When this happens and the failure is EACCES, Exim now
6229 forks a subprocess and does the per-uid checking as the relevant uid.
6231 77. Added process_log_path.
6233 78. If log_file_path was not explicitly set, a setting of check_log_space or
6234 check_log_inodes was ignored.
6236 79. If a space check for the spool or log partitions fails, the incident is now
6237 logged. Of course, in the latter case the data may get lost...
6239 80. Added the %p formatting code to string_format() so that it can be used to
6240 print addresses in debug_print(). Adjusted all the address printing in the
6241 debugging in store.c to use %p rather than %d.
6243 81. There was a concern that a line of code in smtp_in.c could overflow a
6244 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6245 initially expressed, the concern was not well-founded, because trailing
6246 spaces are removed early. However, if the trailing spaces were followed by
6247 a NULL, they did not get removed, so the overflow was possible. Two fixes
6250 (a) I re-wrote the offending code in a cleaner fashion.
6251 (b) If an incoming SMTP command contains a NULL character, it is rejected
6254 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6255 runs initgroups(), so that if the Exim user is in any additional groups,
6256 they will be used during message reception.
6262 The change log for 4.20 and earlier releases has been archived.