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.
15 TL/01 When running the test suite, the README says that variables such as
16 no_msglog_check are global and can be placed anywhere in a specific
17 test's script, however it was observed that placement needed to be near
18 the beginning for it to behave that way. Changed the runtest perl
19 script to read through the entire script once to detect and set these
20 variables, reset to the beginning of the script, and then run through
21 the script parsing/test process like normal.
23 TL/02 The BSD's have an arc4random API. One of the functions to induce
24 adding randomness was arc4random_stir(), but it has been removed in
25 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
26 function when detected.
28 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
29 cause callback expansion.
31 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
32 syntax errors in an expansion can be treated as a string instead of
33 logging or causing an error, due to the internal use of bool_lax
34 instead of bool when processing it.
36 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
37 server certificates when making smtp deliveries.
39 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
41 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
43 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
45 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
46 Merged patch from Sebastian Wiedenroth.
48 JH/05 Fix results-pipe from transport process. Several recipients, combined
49 with certificate use, exposed issues where response data items split
50 over buffer boundaries were not parsed properly. This eventually
51 resulted in duplicates being sent. This issue only became common enough
52 to notice due to the introduction of conection certificate information,
53 the item size being so much larger. Found and fixed by Wolfgang Breyha.
55 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
56 size buffer was used, resulting in syntax errors when an expansion
59 JH/07 Add support for directories of certificates when compiled with a GnuTLS
60 version 3.3.6 or later.
62 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
63 is EXPERIMENTAL_EVENT, the main-configuration and transport options
64 both become "event_action", the variables become $event_name, $event_data
65 and $event_defer_errno. There is a new variable $verify_mode, usable in
66 routers, transports and related events. The tls:cert event is now also
67 raised for inbound connections, if the main configuration event_action
70 TL/06 In test suite, disable OCSP for old versions of openssl which contained
71 early OCSP support, but no stapling (appears to be less than 1.0.0).
73 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
74 server certificate names available under the smtp transport option
75 "tls_verify_cert_hostname" now do not permit multi-component wildcard
78 JH/10 Time-related extraction expansions from certificates now use the main
79 option "timezone" setting for output formatting, and are consistent
80 between OpenSSL and GnuTLS compilations. Bug 1541.
82 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
83 encoded parameter in the incoming message. Bug 1558.
85 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
86 include certificate info, eximon was claiming there were spoolfile
89 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
91 JH/14 Log delivery-related information more consistently, using the sequence
92 "H=<name> [<ip>]" wherever possible.
94 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
95 are problematic for Debian distribution, omit them from the release
98 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
100 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
103 JH/17 Fix a null-indirection in certextract expansions when a nondefault
104 output list separator was used.
109 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
110 checkers that were complaining about end of non-void function with no
113 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
114 This was a regression intruduced in 4.83 by another bugfix.
116 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
118 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
119 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
125 TF/01 Correctly close the server side of TLS when forking for delivery.
127 When a message was received over SMTP with TLS, Exim failed to clear up
128 the incoming connection properly after forking off the child process to
129 deliver the message. In some situations the subsequent outgoing
130 delivery connection happened to have the same fd number as the incoming
131 connection previously had. Exim would try to use TLS and fail, logging
132 a "Bad file descriptor" error.
134 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
135 utilities have not been installed.
137 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
138 temporary space as the ACL may create new global variables.
140 TL/01 LDAP support uses per connection or global context settings, depending
141 upon the detected version of the libraries at build time.
143 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
144 to extract and use the src ip:port in logging and expansions as if it
145 were a direct connection from the outside internet. PPv2 support was
146 updated based on HAProxy spec change in May 2014.
148 JH/02 Add ${listextract {number}{list}{success}{fail}}.
150 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
151 Properly escape header and check for NULL return.
153 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
156 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
158 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
159 characters in header names, implemented as a verify condition.
160 Contributed by Michael Fischer v. Mollard.
162 TL/05 Rename SPF condition results err_perm and err_temp to standardized
163 results permerror and temperror. Previous values are deprecated but
164 still accepted. In a future release, err_perm and err_temp will be
165 completely removed, which will be a backward incompatibility if the
166 ACL tests for either of these two old results. Patch contributed by
167 user bes-internal on the mailing list.
169 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
171 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
172 selectors, in both main and reject logs.
174 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
177 JH/07 Add malware type "sock" for talking to simple daemon.
179 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
181 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
182 routers/transports under cutthrough routing.
184 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
185 numbers. Touch up "bool" conditional to keep the same definition.
187 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
189 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
191 JH/12 Expand items in router/transport headers_add or headers_remove lists
192 individually rather than the list as a whole. Bug 1452.
194 Required for reasonable handling of multiple headers_ options when
195 they may be empty; requires that headers_remove items with embedded
196 colons must have them doubled (or the list-separator changed).
198 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
199 view the policy declared in the DMARC record. Currently, $dmarc_status
200 is a combined value of both the record presence and the result of the
203 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
205 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
206 dnslookup router and the smtp transport (applying to the forward
209 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
210 of ldap servers used for a specific lookup. Patch provided by Heiko
213 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
214 New variable $lookup_dnssec_authenticated for observability.
216 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
217 Patch submitted by Lars Timman.
219 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
221 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
222 Requires trusted mode and valid format message id, aborts otherwise.
223 Patch contributed by Heiko Schlichting.
225 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
226 certextract with support for various fields. Bug 1358.
228 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
229 is requested by default, modifiable by smtp transport option
232 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
233 operate on certificate variables to give certificate fingerprints
234 Also new ${sha256:cert_variable}.
236 JH/23 The PRDR feature is moved from being Experimental into the mainline.
238 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
239 Christian Aistleitner.
241 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
243 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
244 file. Patch from Wolfgang Breyha.
246 JH/25 Expand the coverage of the delivery $host and $host_address to
247 client authenticators run in verify callout. Bug 1476.
249 JH/26 Port service names are now accepted for tls_on_connect_ports, to
250 align with daemon_smtp_ports. Bug 72.
252 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
253 support and error reporting did not work properly.
255 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
256 and is readable. Patch from Andrew Colin Kissa.
258 TL/14 Enhance documentation of ${run expansion and how it parses the
259 commandline after expansion, particularly in the case when an
260 unquoted variable expansion results in an empty value.
262 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
264 PP/02 Fix internal collision of T_APL on systems which support RFC3123
265 by renaming away from it. Addresses GH issue 15, reported by
268 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
270 TL/15 SECURITY: prevent double expansion in math comparison functions
271 (can expand unsanitized data). Not remotely exploitable.
278 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
280 PP/02 Make -n do something, by making it not do something.
281 When combined with -bP, the name of an option is not output.
283 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
286 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
287 $sender_host_name and config options to manage this, and basic check
290 PP/05 DSCP support for outbound connections and control modifier for inbound.
292 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
293 (Only plugin which currently uses this is kerberos4, which nobody should
294 be using, but we should make it available and other future plugins might
295 conceivably use it, even though it would break NAT; stuff *should* be
296 using channel bindings instead).
298 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
299 name; added for Sendmail compatibility; requires admin caller.
300 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
301 just ignore it); requires trusted caller.
302 Also parse but ignore: -Ac -Am -X<logfile>
305 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
307 TL/02 Add +smtp_confirmation as a default logging option.
309 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
310 Patch by Magnus Holmgren from 2007-02-20.
312 TL/04 Bugzilla 1281 - Spec typo.
313 Bugzilla 1283 - Spec typo.
314 Bugzilla 1290 - Spec grammar fixes.
316 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
318 TL/06 Add Experimental DMARC support using libopendmarc libraries.
320 TL/07 Fix an out of order global option causing a segfault. Reported to dev
321 mailing list by by Dmitry Isaikin.
323 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
325 JH/02 Support "G" suffix to numbers in ${if comparisons.
327 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
329 NM/01 Bugzilla 1197 - Spec typo
330 Bugzilla 1196 - Spec examples corrections
332 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
334 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
335 gnutls_enable_pkcs11, but renamed to more accurately indicate its
338 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
339 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
341 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
342 "acl {{name}{arg}...}", and optional args on acl condition
345 JH/05 Permit multiple router/transport headers_add/remove lines.
347 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
349 JH/07 Avoid using a waiting database for a single-message-only transport.
350 Performance patch from Paul Fisher. Bugzilla 1262.
352 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
355 JH/09 Add $headers_added variable, with content from use of ACL modifier
356 add_header (but not yet added to the message). Bugzilla 199.
358 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
359 Pulled from Bugzilla 817 by Wolfgang Breyha.
361 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
363 (nb: this is the same fix as in Exim 4.80.1)
365 JH/11 Add A= logging on delivery lines, and a client_set_id option on
368 JH/12 Add optional authenticated_sender logging to A= and a log_selector
371 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
373 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
374 advertise SMTP AUTH mechanism to us, instead of a generic
375 protocol violation error. Also, make Exim more robust to bad
376 data from the Dovecot auth socket.
378 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
380 When a queue runner is handling a message, Exim first routes the
381 recipient addresses, during which it prunes them based on the retry
382 hints database. After that it attempts to deliver the message to
383 any remaining recipients. It then updates the hints database using
386 So if a recipient address works intermittently, it can get repeatedly
387 deferred at routing time. The retry hints record remains fresh so the
388 address never reaches the final cutoff time.
390 This is a fairly common occurrence when a user is bumping up against
391 their storage quota. Exim had some logic in its local delivery code
392 to deal with this. However it did not apply to per-recipient defers
393 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
395 This change adds a proper retry rule check during routing so that the
396 final cutoff time is checked against the message's age. We only do
397 this check if there is an address retry record and there is not a
398 domain retry record; this implies that previous attempts to handle
399 the address had the retry_use_local_parts option turned on. We use
400 this as an approximation for the destination being like a local
401 delivery, as in LMTP.
403 I suspect this new check makes the old local delivery cutoff check
404 redundant, but I have not verified this so I left the code in place.
406 TF/02 Correct gecos expansion when From: is a prefix of the username.
408 Test 0254 submits a message to Exim with the header
412 When I ran the test suite under the user fanf2, Exim expanded
413 the header to contain my full name, whereas it should have added
414 a Resent-Sender: header. It erroneously treats any prefix of the
415 username as equal to the username.
417 This change corrects that bug.
419 GF/01 DCC debug and logging tidyup
420 Error conditions log to paniclog rather than rejectlog.
421 Debug lines prefixed by "DCC: " to remove any ambiguity.
423 TF/03 Avoid unnecessary rebuilds of lookup-related code.
425 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
426 Bug spotted by Jeremy Harris; was flawed since initial commit.
427 Would have resulted in OCSP responses post-SNI triggering an Exim
428 NULL dereference and crash.
430 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
432 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
433 Bug detection, analysis and fix by Samuel Thibault.
434 Bugzilla 1331, Debian bug #698092.
436 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
438 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
439 Server implementation by Todd Lyons, client by JH.
440 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
441 config variable "prdr_enable" controls whether the server
442 advertises the facility. If the client requests PRDR a new
443 acl_data_smtp_prdr ACL is called once for each recipient, after
444 the body content is received and before the acl_smtp_data ACL.
445 The client is controlled by bolth of: a hosts_try_prdr option
446 on the smtp transport, and the server advertisement.
447 Default client logging of deliveries and rejections involving
448 PRDR are flagged with the string "PRDR".
450 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
451 fclose(). Diagnosis by Todd Lyons.
453 PP/17 Update configure.default to handle IPv6 localhost better.
454 Patch by Alain Williams (plus minor tweaks).
457 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
458 This is now consistent with GnuTLS, and is now documented: the
459 previous undocumented portable approach to treating the option as
460 unset was to force an expansion failure. That still works, and
461 an empty string is now equivalent.
463 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
464 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
465 not performing validation itself.
467 PP/20 Added force_command boolean option to pipe transport.
468 Patch from Nick Koston, of cPanel Inc.
470 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
473 TF/04 Added udpsend ACL modifer and hexquote expansion operator
475 PP/21 Fix eximon continuous updating with timestamped log-files.
476 Broken in a format-string cleanup in 4.80, missed when I repaired the
477 other false fix of the same issue.
478 Report and fix from Heiko Schlichting.
481 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
482 Report from Prashanth Katuri.
484 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
485 It's SecureTransport, so affects any MacOS clients which use the
486 system-integrated TLS libraries, including email clients.
488 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
489 using a MIME ACL for non-SMTP local injection.
490 Report and assistance in diagnosis by Warren Baker.
492 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
494 JH/16 Fix comparisons for 64b. Bugzilla 1385.
496 TL/09 Add expansion variable $authenticated_fail_id to keep track of
497 last id that failed so it may be referenced in subsequent ACL's.
499 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
502 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
503 ldap library initialization, allowing self-signed CA's to be
504 used. Also properly sets require_cert option later in code by
505 using NULL (global ldap config) instead of ldap handle (per
506 session). Bug diagnosis and testing by alxgomz.
508 TL/12 Enhanced documentation in the ratelimit.pl script provided in
509 the src/util/ subdirectory.
511 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
512 renamed to Transport Post Delivery Action by Jeremy Harris, as
515 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
516 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
517 redis_servers = needs to be configured which will be used by the redis
518 lookup. Patch from Warren Baker, of The Packet Hub.
520 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
522 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
523 hostname or reverse DNS when processing a host list. Used suggestions
524 from multiple comments on this bug.
526 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
528 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
529 Missed a few lines, added it to make the runtest require no keyboard
532 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
533 contains upper case chars. Make router use caseful_local_part.
535 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
536 support when GnuTLS has been built with p11-kit.
542 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
544 This, or similar/improved, will also be change PP/11 of 4.82.
550 PP/01 Handle short writes when writing local log-files.
551 In practice, only affects FreeBSD (8 onwards).
552 Bugzilla 1053, with thanks to Dmitry Isaikin.
554 NM/01 Bugzilla 949 - Documentation tweak
556 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
559 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
561 PP/02 Implemented gsasl authenticator.
563 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
565 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
566 `pkg-config foo` for cflags/libs.
568 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
569 with rest of GSASL and with heimdal_gssapi.
571 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
572 `pkg-config foo` for cflags/libs for the TLS implementation.
574 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
575 properties get this fed in as external SSF. A number of robustness
576 and debugging improvements to the cyrus_sasl authenticator.
578 PP/08 cyrus_sasl server now expands the server_realm option.
580 PP/09 Bugzilla 1214 - Log authentication information in reject log.
581 Patch by Jeremy Harris.
583 PP/10 Added dbmjz lookup type.
585 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
587 PP/12 MAIL args handles TAB as well as SP, for better interop with
588 non-compliant senders.
589 Analysis and variant patch by Todd Lyons.
591 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
592 Bug report from Lars Müller <lars@samba.org> (via SUSE),
593 Patch from Dirk Mueller <dmueller@suse.com>
595 PP/13 tls_peerdn now print-escaped for spool files.
596 Observed some $tls_peerdn in wild which contained \n, which resulted
597 in spool file corruption.
599 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
600 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
601 or write after TLS renegotiation, which otherwise led to messages
604 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
605 as a tracking header (ie: a signed header comes before the signature).
606 Patch from Wolfgang Breyha.
608 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
609 comma-sep list; embedded commas doubled.
611 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
613 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
615 Report and patch from Dmitry Banschikov.
617 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
618 Removed SSL_clear() after SSL_new() which led to protocol negotiation
619 failures. We appear to now support TLS1.1+ with Exim.
621 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
622 lets Exim select keys and certificates based upon TLS SNI from client.
623 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
624 before an outbound SMTP session. New log_selector, +tls_sni.
626 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
627 NULL dereference. Report and patch from Alun Jones.
629 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
630 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
631 Not seeing resolver debug output on NetBSD, but suspect this is a
632 resolver implementation change.
634 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
635 Left warnings. Added "eximon gdb" invocation mode.
637 PP/21 Defaulting "accept_8bitmime" to true, not false.
639 PP/22 Added -bw for inetd wait mode support.
641 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
642 locate the relevant includes and libraries. Made this the default.
644 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
645 Bugzilla 1246, report and most of solution from Tomasz Kusy.
647 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
648 This may cause build issues on older platforms.
650 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
651 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
652 gnutls_require_mac & gnutls_require_protocols (no longer supported).
653 Added SNI support via GnuTLS too.
654 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
656 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
658 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
659 Applied second patch from Janne, implementing suggestion to default
660 multiple-strings-in-record handling to match SPF spec.
662 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
664 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
665 read-only, out of scope).
666 Patch from Wolfgang Breyha, report from Stuart Northfield.
668 PP/29 Fix three issues highlighted by clang analyser static analysis.
669 Only crash-plausible issue would require the Cambridge-specific
670 iplookup router and a misconfiguration.
671 Report from Marcin Mirosław.
673 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
675 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
676 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
677 As part of this, removing so much warning spew let me fix some minor
678 real issues in debug logging.
680 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
681 assignment on my part. Fixed.
683 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
684 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
685 Janne Snabb (who went above and beyond: thank you).
687 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
688 string otherwise requires a connection and a bunch more work and it's
689 relatively easy to get wrong. Should also expose TLS library linkage
692 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
693 64-bit ${eval} (JH/03).
695 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
696 GNU libc to support some of the 64-bit stuff, should not lead to
697 conflicts. Defined before os.h is pulled in, so if a given platform
698 needs to override this, it can.
700 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
701 protection layer was required, which is not implemented.
702 Bugzilla 1254, patch from Wolfgang Breyha.
704 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
705 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
706 tls_dhparam take prime identifiers. Also unbreak combination of
707 OpenSSL+DH_params+TLSSNI.
709 PP/39 Disable SSLv2 by default in OpenSSL support.
715 PP/01 Solaris build fix for Oracle's LDAP libraries.
716 Bugzilla 1109, patch from Stephen Usher.
718 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
720 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
723 TF/02 Fix a couple more cases where we did not log the error message
724 when unlink() failed. See also change 4.74-TF/03.
726 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
727 lock up or crash if it happened to be inside a call to libc when it
728 got a SIGUSR1 from exiwhat.
730 The SIGUSR1 handler appends the current process status to the process
731 log which is later printed by exiwhat. It used to use the general
732 purpose logging code to do this, but several functions it calls are
733 not safe for signals.
735 The new output code in the SIGUSR1 handler is specific to the process
736 log, and simple enough that it's easy to inspect for signal safety.
737 Removing some special cases also simplifies the general logging code.
738 Removing the spurious timestamps from the process log simplifies
741 TF/04 Improved ratelimit ACL condition.
743 The /noupdate option has been deprecated in favour of /readonly which
744 has clearer semantics. The /leaky, /strict, and /readonly update modes
745 are mutually exclusive. The update mode is no longer included in the
746 database key; it just determines when the database is updated. (This
747 means that when you upgrde Exim will forget old rate measurements.)
749 Exim now checks that the per_* options are used with an update mode that
750 makes sense for the current ACL. For example, when Exim is processing a
751 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
752 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
753 must specify per_mail/readonly. If you omit the update mode it defaults to
754 /leaky where that makes sense (as before) or /readonly where required.
756 The /noupdate option is now undocumented but still supported for
757 backwards compatibility. It is equivalent to /readonly except that in
758 ACLs where /readonly is required you may specify /leaky/noupdate or
759 /strict/noupdate which are treated the same as /readonly.
761 A useful new feature is the /count= option. This is a generalization
762 of the per_byte option, so that you can measure the throughput of other
763 aggregate values. For example, the per_byte option is now equivalent
764 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
766 The per_rcpt option has been generalized using the /count= mechanism
767 (though it's more complicated than the per_byte equivalence). When it is
768 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
769 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
770 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
771 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
772 Note that using per_rcpt with a non-readonly update mode in more than
773 one ACL will cause the recipients to be double-counted. (The per_mail
774 and per_byte options don't have this problem.)
776 The handling of very low rates has changed slightly. If the computed rate
777 is less than the event's count (usually one) then this event is the first
778 after a long gap. In this case the rate is set to the same as this event's
779 count, so that the first message of a spam run is counted properly.
781 The major new feature is a mechanism for counting the rate of unique
782 events. The new per_addr option counts the number of different
783 recipients that someone has sent messages to in the last time period. It
784 behaves like per_rcpt if all the recipient addresses are different, but
785 duplicate recipient addresses do not increase the measured rate. Like
786 the /count= option this is a general mechanism, so the per_addr option
787 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
788 example, measure the rate that a client uses different sender addresses
789 with the options per_mail/unique=$sender_address. There are further
790 details in the main documentation.
792 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
794 TF/06 Removed a few PCRE remnants.
796 TF/07 Automatically extract Exim's version number from tags in the git
797 repository when doing development or release builds.
799 PP/02 Raise smtp_cmd_buffer_size to 16kB.
800 Bugzilla 879. Patch from Paul Fisher.
802 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
803 Heavily based on revision 40f9a89a from Simon Arlott's tree.
806 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
808 PP/05 Variable $av_failed, true if the AV scanner deferred.
809 Bugzilla 1078. Patch from John Horne.
811 PP/06 Stop make process more reliably on build failure.
812 Bugzilla 1087. Patch from Heiko Schlittermann.
814 PP/07 Make maildir_use_size_file an _expandable_ boolean.
815 Bugzilla 1089. Patch from Heiko Schlittermann.
817 PP/08 Handle ${run} returning more data than OS pipe buffer size.
818 Bugzilla 1131. Patch from Holger Weiß.
820 PP/09 Handle IPv6 addresses with SPF.
821 Bugzilla 860. Patch from Wolfgang Breyha.
823 PP/10 GnuTLS: support TLS 1.2 & 1.1.
825 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
828 PP/11 match_* no longer expand right-hand-side by default.
829 New compile-time build option, EXPAND_LISTMATCH_RHS.
830 New expansion conditions, "inlist", "inlisti".
832 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
834 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
836 PP/14 fix log_write() format string regression from TF/03.
837 Bugzilla 1152. Patch from Dmitry Isaikin.
843 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
845 PP/02 Harmonised TLS library version reporting; only show if debugging.
846 Layout now matches that introduced for other libraries in 4.74 PP/03.
848 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
850 PP/04 New "dns_use_edns0" global option.
852 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
855 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
856 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
858 TK/01 Updated PolarSSL code to 0.14.2.
859 Bugzilla 1097. Patch from Andreas Metzler.
861 PP/07 Catch divide-by-zero in ${eval:...}.
864 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
867 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
868 format-string attack -- SECURITY: remote arbitrary code execution.
870 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
871 time unintentionally subject to list matching rules, letting the header
872 cause arbitrary Exim lookups (of items which can occur in lists, *not*
873 arbitrary string expansion). This allowed for information disclosure.
875 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
876 INT_MIN/-1 -- value coerced to INT_MAX.
882 NM/01 Workround for PCRE version dependency in version reporting
885 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
886 This fixes portability to compilers other than gcc, notably
887 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
889 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
890 makefiles for portability to HP-UX and POSIX correctness.
892 PP/01 Permit LOOKUP_foo enabling on the make command-line.
893 Also via indented variable definition in the Makefile.
894 (Debugging by Oliver Heesakkers).
896 PP/02 Restore caching of spamd results with expanded spamd_address.
897 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
899 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
900 Improves build reliability. Fix from: Frank Elsner
902 NM/02 Fix wide character breakage in the rfc2047 coding
903 Fixes bug 1064. Patch from Andrey N. Oktyabrski
905 NM/03 Allow underscore in dnslist lookups
906 Fixes bug 1026. Patch from Graeme Fowler
908 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
909 Code patches from Adam Ciarcinski of NetBSD.
911 NM/04 Fixed exiqgrep to cope with mailq missing size issue
914 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
915 is logged, to avoid truncation. Patch from John Horne.
917 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
918 Patch from Jakob Hirsch.
920 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
921 SQL string expansion failure details.
922 Patch from Andrey Oktyabrski.
924 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
925 Patch from Simon Arlott.
927 PP/09 New lookups functionality failed to compile on old gcc which rejects
928 extern declarations in function scope.
929 Patch from Oliver Fleischmann
931 PP/10 Use sig_atomic_t for flags set from signal handlers.
932 Check getgroups() return and improve debugging.
933 Fixed developed for diagnosis in bug 927 (which turned out to be
936 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
937 Patch from Mark Zealey.
939 PP/12 Bugzilla 1056: Improved spamd server selection.
940 Patch from Mark Zealey.
942 PP/13 Bugzilla 1086: Deal with maildir quota file races.
943 Based on patch from Heiko Schlittermann.
945 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
946 Patch from Uwe Doering, sign-off by Michael Haardt.
948 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
949 variable declaration deep within a block. Bug and patch from
952 PP/15 lookups-Makefile IRIX compatibilty coercion.
954 PP/16 Make DISABLE_DKIM build knob functional.
956 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
957 Patch by Simon Arlott
959 TF/03 Fix valgrind.h portability to C89 compilers that do not support
960 variable argument macros. Our copy now differs from upstream.
966 TF/01 Failure to get a lock on a hints database can have serious
967 consequences so log it to the panic log.
969 TF/02 Log LMTP confirmation messages in the same way as SMTP,
970 controlled using the smtp_confirmation log selector.
972 TF/03 Include the error message when we fail to unlink a spool file.
974 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
975 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
976 for maintaining out-of-tree patches for some time.
978 PP/01 Bugzilla 139: Documentation and portability issues.
979 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
980 Handle per-OS dynamic-module compilation flags.
982 PP/02 Let /dev/null have normal permissions.
983 The 4.73 fixes were a little too stringent and complained about the
984 permissions on /dev/null. Exempt it from some checks.
985 Reported by Andreas M. Kirchwitz.
987 PP/03 Report version information for many libraries, including
988 Exim version information for dynamically loaded libraries. Created
989 version.h, now support a version extension string for distributors
990 who patch heavily. Dynamic module ABI change.
992 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
993 privilege escalation vulnerability whereby the Exim run-time user
994 can cause root to append content of the attacker's choosing to
997 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1000 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1001 If dropping privileges for untrusted macros, we disabled normal logging
1002 on the basis that it would fail; for the Exim run-time user, this is not
1003 the case, and it resulted in successful deliveries going unlogged.
1004 Fixed. Reported by Andreas Metzler.
1010 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1011 only prepend for the Resent-* case. Fixes regression introduced in
1012 Exim 4.70 by NM/22 for Bugzilla 607.
1014 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1015 increasing numbers of administrators be bitten by this.
1017 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1019 PP/03 Bugzilla 994: added openssl_options main configuration option.
1021 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1023 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1025 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1027 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1028 without a peer certificate, leading to a segfault because of an
1029 assumption that peers always have certificates. Be a little more
1030 paranoid. Problem reported by Martin Tscholak.
1032 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1033 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1034 NB: ClamAV planning to remove STREAM in "middle of 2010".
1035 CL also introduces -bmalware, various -d+acl logging additions and
1036 more caution in buffer sizes.
1038 PP/09 Implemented reverse_ip expansion operator.
1040 PP/10 Bugzilla 937: provide a "debug" ACL control.
1042 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1044 PP/12 Bugzilla 973: Implement --version.
1046 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1048 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1050 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1052 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1053 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1054 ignore trailing whitespace.
1056 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1058 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1061 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1062 Notification from Dr Andrew Aitchison.
1064 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1065 ExtendedDetectionInfo response format.
1066 Notification from John Horne.
1068 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1071 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1072 XSL and documented dependency on system catalogs, with examples of how
1075 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1078 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1079 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1080 configuration file which is writeable by the Exim user or group.
1082 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1083 of configuration files to cover files specified with the -C option if
1084 they are going to be used with root privileges, not just the default
1087 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1088 option (effectively making it always true).
1090 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1091 files to be used while preserving root privileges.
1093 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1094 that rogue child processes cannot use them.
1096 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1097 run-time user, instead of root.
1099 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1100 Exim run-time user without dropping privileges.
1102 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1103 result string, instead of calling string_vformat() twice with the same
1106 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1107 for other users. Others should always drop root privileges if they use
1108 -C on the command line, even for a whitelisted configure file.
1110 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1112 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1118 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1119 $data_path, and $header_path variables; fixed documentation bugs and
1122 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1123 exipick to access non-standard spools, including the "frozen" queue
1126 NM/01 Bugzilla 965: Support mysql stored procedures.
1127 Patch from Alain Williams
1129 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1131 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1132 Patch from Andreas Metzler
1134 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1135 Patch from Kirill Miazine
1137 NM/05 Bugzilla 671: Added umask to procmail example.
1139 JJ/03 installed exipick 20100323.0, fixing doc bug
1141 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1142 directory. Notification and patch from Dan Rosenberg.
1144 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1146 TK/02 Improve log output when DKIM signing operation fails.
1148 MH/01 Treat the transport option dkim_domain as a colon separated
1149 list, not as a single string, and sign the message with each element,
1150 omitting multiple occurences of the same signer.
1152 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1153 Bugzilla 985, 986. Patch by Simon Arlott
1155 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1156 Patch by Simon Arlott
1158 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1159 MBX locking. Notification from Dan Rosenberg.
1165 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1167 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1169 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1171 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1173 NM/04 Bugzilla 915: Flag broken perl installation during build.
1179 TK/01 Added patch by Johannes Berg that expands the main option
1180 "spamd_address" if it starts with a dollar sign.
1182 TK/02 Write list of recipients to X-Envelope-Sender header when building
1183 the mbox-format spool file for content scanning (suggested by Jakob
1186 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1187 (http://www.dcc-servers.net/) support via dccifd. Activated by
1188 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1190 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1191 by Mark Daniel Reidel <mr@df.eu>.
1193 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1194 When building exim an external PCRE library is now needed -
1195 PCRE is a system library on the majority of modern systems.
1196 See entry on PCRE_LIBS in EDITME file.
1198 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1199 conversation. Added nologin parameter to request.
1200 Patch contributed by Kirill Miazine.
1202 TF/01 Do not log submission mode rewrites if they do not change the address.
1204 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1206 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1207 log files in place. Contributed by Roberto Lima.
1209 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1211 TF/03 Bugzilla 615: When checking the local_parts router precondition
1212 after a local_part_suffix or local_part_prefix option, Exim now
1213 does not use the address's named list lookup cache, since this
1214 contains cached lookups for the whole local part.
1216 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1217 Robert Millan. Documentation is in experimental-spec.txt.
1219 TF/04 Bugzilla 668: Fix parallel build (make -j).
1221 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1223 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1224 Patch provided by Jan Srzednicki.
1226 TF/05 Leading white space used to be stripped from $spam_report which
1227 wrecked the formatting. Now it is preserved.
1229 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1230 that they are available at delivery time.
1232 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1234 TF/08 TLS error reporting now respects the incoming_interface and
1235 incoming_port log selectors.
1237 TF/09 Produce a more useful error message if an SMTP transport's hosts
1238 setting expands to an empty string.
1240 NM/06 Bugzilla 744: EXPN did not work under TLS.
1241 Patch provided by Phil Pennock.
1243 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1244 Patch provided by Richard Godbee.
1246 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1247 acl_smtp_notquit, added index entry.
1249 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1250 Patch provided by Eugene Bujak.
1252 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1253 accept(). Patch provided by Maxim Dounin.
1255 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1256 Patch provided by Phil Pennock.
1258 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1260 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1261 Patch provided by Brad "anomie" Jorsch.
1263 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1264 Patch provided by Dean Brooks.
1266 TK/05 Add native DKIM support (does not depend on external libraries).
1268 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1269 Patch provided by Graeme Fowler.
1271 NM/16 Bugzilla 851: Documentation example syntax fix.
1273 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1275 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1278 NM/19 Bugzilla 745: TLS version reporting.
1279 Patch provided by Phil Pennock.
1281 NM/20 Bugzilla 167: bool: condition support.
1282 Patch provided by Phil Pennock.
1284 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1285 clients. Patch provided by Phil Pennock.
1287 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1288 Patch provided by Brad "anomie" Jorsch.
1290 NM/23 Bugzilla 687: Fix misparses in eximstats.
1291 Patch provided by Heiko Schlittermann.
1293 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1294 Patch provided by Heiko Schlittermann.
1296 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1297 plus update to original patch.
1299 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1301 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1302 Patch provided by David Brownlee.
1304 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1306 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1308 NM/30 Bugzilla 888: TLS documentation bugfixes.
1310 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1312 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1313 Unlike the original bugzilla I have changed all shell scripts in src tree.
1315 NM/33 Bugzilla 898: Transport filter timeout fix.
1316 Patch by Todd Rinaldo.
1318 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1319 Patch by Serge Demonchaux.
1321 NM/35 Bugzilla 39: Base64 decode bug fixes.
1322 Patch by Jakob Hirsch.
1324 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1326 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1328 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1330 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1336 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1337 ALT-N's libdkim that I have put here:
1338 http://duncanthrax.net/exim-experimental/
1340 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1341 were called 'true' and it seems that C99 defines that as a reserved
1342 keyword to be used with 'bool' variable types. That means you could
1343 not include C99-style headers which use bools without triggering
1344 build errors in sieve.c.
1346 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1347 as mailq or other aliases. Changed the --help handling significantly
1348 to do whats expected. exim_usage() emits usage/help information.
1350 SC/01 Added the -bylocaldomain option to eximstats.
1352 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1354 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1356 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1362 PH/01 Another patch from the Sieve maintainer.
1364 PH/02 When an IPv6 address is converted to a string for single-key lookup
1365 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1366 dots are used instead of colons so that keys in lsearch files need not
1367 contain colons. This was done some time before quoting was made available
1368 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1369 (notated using the quote facility) so as to distinguish them from IPv4
1370 keys. This meant that lookups for IP addresses in host lists did not work
1371 for iplsearch lookups.
1373 This has been fixed by arranging for IPv6 addresses to be expressed with
1374 colons if the lookup type is iplsearch. This is not incompatible, because
1375 previously such lookups could never work.
1377 The situation is now rather anomolous, since one *can* have colons in
1378 ordinary lsearch keys. However, making the change in all cases is
1379 incompatible and would probably break a number of configurations.
1381 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1384 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1385 conversion specification without a maximum field width, thereby enabling
1386 a rogue spamd server to cause a buffer overflow. While nobody in their
1387 right mind would setup Exim to query an untrusted spamd server, an
1388 attacker that gains access to a server running spamd could potentially
1389 exploit this vulnerability to run arbitrary code as the Exim user.
1391 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1392 $primary_hostname instead of what libspf2 thinks the hosts name is.
1394 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1395 a directory entry by the name of the lookup key. Previously, if a
1396 symlink pointed to a non-existing file or a file in a directory that
1397 Exim lacked permissions to read, a lookup for a key matching that
1398 symlink would fail. Now it is enough that a matching directory entry
1399 exists, symlink or not. (Bugzilla 503.)
1401 PH/03 The body_linecount and body_zerocount variables are now exported in the
1404 PH/04 Added the $dnslist_matched variable.
1406 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1407 This means they are set thereafter only if the connection becomes
1410 PH/06 Added the client_condition to authenticators so that some can be skipped
1411 by clients under certain conditions.
1413 PH/07 The error message for a badly-placed control=no_multiline_responses left
1414 "_responses" off the end of the name.
1416 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1418 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1419 (without spaces) instead of just copying the configuration text.
1421 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1423 PH/11 Added $max_received_linelength.
1425 PH/12 Added +ignore_defer and +include_defer to host lists.
1427 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1428 way in which PCRE > 7.0 is built.
1430 PH/14 Implemented queue_only_load_latch.
1432 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1433 MAIL command. The effect was to mangle the value on 64-bit systems.
1435 PH/16 Another patch from the Sieve maintainer.
1437 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1439 PH/18 If a system quota error occurred while trying to create the file for
1440 a maildir delivery, the message "Mailbox is full" was not appended to the
1441 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1442 only to a quota excession during the actual writing of the file.
1444 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1445 characters?) which causes problems in log lines. The DN values are now
1446 passed through string_printing() before being added to log lines.
1448 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1449 and InterBase are left for another time.)
1451 PH/21 Added message_body_newlines option.
1453 PH/22 Guard against possible overflow in moan_check_errorcopy().
1455 PH/23 POSIX allows open() to be a macro; guard against that.
1457 PH/24 If the recipient of an error message contained an @ in the local part
1458 (suitably quoted, of course), incorrect values were put in $domain and
1459 $local_part during the evaluation of errors_copy.
1465 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1466 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1469 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1470 issue a MAIL command.
1472 PH/02 In an ACL statement such as
1474 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1476 if a client was not listed at all, or was listed with a value other than
1477 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1478 the condition was not true (as it should be), so access was not denied.
1479 The bug was that the ! inversion was incorrectly passed on to the second
1480 item. This has been fixed.
1482 PH/03 Added additional dnslists conditions == and =& which are different from
1483 = and & when the dns lookup returns more than one IP address.
1485 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1486 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1488 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1489 FSYNC, which compiles an option called disable_fsync that allows for
1490 bypassing fsync(). The documentation is heavily laced with warnings.
1492 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1494 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1495 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1496 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1497 including adding "make clean"; (3) Added -fPIC when compiling the test
1498 dynamically loaded module, to get rid of a warning.
1500 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1501 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1502 = 0s. The bug is otherwise harmless.
1504 PH/07 There was a bug in the dovecot authenticator such that the value of
1505 $auth1 could be overwritten, and so not correctly preserved, after a
1506 successful authentication. This usually meant that the value preserved by
1507 the server_setid option was incorrect.
1509 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1511 PH/09 Installed PCRE release 7.0.
1513 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1514 run for batched SMTP input. It is now run at the start of every message
1515 in the batch. While fixing this I discovered that the process information
1516 (output by running exiwhat) was not always getting set for -bs and -bS
1517 input. This is fixed, and it now also says "batched" for BSMTP.
1519 PH/11 Added control=no_pipelining.
1521 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1522 patch, slightly modified), and move the expansion of helo_data till after
1523 the connection is made in the smtp transport (so it can use these
1526 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1528 PH/14 Added log_selector = +pid.
1530 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1532 PH/16 Add ${if forany and ${if forall.
1534 PH/17 Added dsn_from option to vary the From: line in DSNs.
1536 PH/18 Flush SMTP output before performing a callout, unless control =
1537 no_callout_flush is set.
1539 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1540 was true (the default) a successful delivery failed to delete the retry
1541 item, thus causing premature timeout of the address. The bug is now
1544 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1546 PH/21 Long custom messages for fakedefer and fakereject are now split up
1547 into multiline reponses in the same way that messages for "deny" and
1548 other ACL rejections are.
1550 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1551 with slight modification.
1553 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1554 draft, changing the syntax and factoring some duplicate code".
1556 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1557 for deliveries of the second and subsequent messages over the same SMTP
1560 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1561 ${reduce, with only minor "tidies".
1563 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1565 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1566 expansion side effects.
1568 PH/27 When a message times out after an over-quota error from an Exim-imposed
1569 quota, the bounce message says "mailbox is full". This message was not
1570 being given when it was a system quota that was exceeded. It now should
1573 MH/03 Made $recipients available in local_scan(). local_scan() already has
1574 better access to the recipient list through recipients_list[], but
1575 $recipients can be useful in postmaster-provided expansion strings.
1577 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1578 in the case of a MAIL command with additional options following the
1579 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1580 were accidentally chopped off.
1582 PH/29 SMTP synchronization checks are implemented when a command is read -
1583 there is a check that no more input is waiting when there shouldn't be
1584 any. However, for some commands, a delay in an ACL can mean that it is
1585 some time before the response is written. In this time, more input might
1586 arrive, invalidly. So now there are extra checks after an ACL has run for
1587 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1588 pipelining has not been advertised.
1590 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1592 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1593 returned in the SMTP response when smtp_return_error_details was false.
1594 This has been fixed.
1596 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1597 instead of the C I/O that was originally supplied, because problems were
1598 reported on Solaris.
1600 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1601 Exim which did not show up earlier: it was assuming that a call to
1602 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1603 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1604 was a macro that became an assignment, so it seemed to work. This has
1605 changed to a proper function call with a void return, hence the compile
1606 error. Exim's code has been fixed.
1608 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1611 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1613 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1615 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1616 because it turns out to be too big or there is a timeout) there is no
1617 "Completed" line in the log. When some messages of this type were
1618 selected by exigrep, they were listed as "not completed". Others were
1619 picked up by some special patterns. I have improved the selection
1620 criteria to be more general.
1622 PH/38 The host_find_failed option in the manualroute router can now be set
1623 to "ignore", to completely ignore a host whose IP address cannot be
1624 found. If all hosts are ignored, the behaviour is controlled by the new
1625 host_all_ignored option.
1627 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1628 homing or because of multiple MX records with /mx) generated more than
1629 one IP address, and the following item turned out to be the local host,
1630 all the secondary addresses of the first item were incorrectly removed
1631 from the list, along with the local host and any following hosts (which
1632 is what is supposed to happen).
1634 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1635 whoever called Exim into the -H file. In the case of the daemon it was
1636 behaving confusingly. When first started, it used values for whoever
1637 started the daemon, but after a SIGHUP it used the Exim user (because it
1638 calls itself on a restart). I have changed the code so that it now always
1641 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1642 message are rejected with the same error (e.g. no authentication or bad
1643 sender address), and a DATA command is nevertheless sent (as can happen
1644 with PIPELINING or a stupid MUA), the error message that was given to the
1645 RCPT commands is included in the rejection of the DATA command. This is
1646 intended to be helpful for MUAs that show only the final error to their
1649 PH/42 Another patch from the Sieve maintainer.
1651 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1652 Eximstats - Fixed some broken HTML links and added missing column headers
1654 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1655 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1657 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1659 PH/43 Yet another patch from the Sieve maintainer.
1661 PH/44 I found a way to check for a TCP/IP connection going away before sending
1662 the response to the final '.' that terminates a message, but only in the
1663 case where the client has not sent further data following the '.'
1664 (unfortunately, this is allowed). However, in many cases there won't be
1665 any further data because there won't be any more messages to send. A call
1666 to select() can be used: if it shows that the input is "ready", there is
1667 either input waiting, or the socket has been closed. An attempt to read
1668 the next input character can distinguish the two cases. Previously, Exim
1669 would have sent an OK response which the client would never have see.
1670 This could lead to message repetition. This fix should cure that, at
1671 least in a lot of common cases.
1673 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1674 advertised in response to EHLO.
1680 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1681 fixed by 4.65/MH/01 (is this a record?) are fixed:
1683 (i) An empty string was always treated as zero by the numeric comparison
1684 operators. This behaviour has been restored.
1686 (ii) It is documented that the numeric comparison operators always treat
1687 their arguments as decimal numbers. This was broken in that numbers
1688 starting with 0 were being interpreted as octal.
1690 While fixing these problems I realized that there was another issue that
1691 hadn't been noticed. Values of message_size_limit (both the global option
1692 and the transport option) were treated as octal if they started with 0.
1693 The documentation was vague. These values are now always treated as
1694 decimal, and I will make that clear in the documentation.
1700 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1701 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1704 MH/01 Don't check that the operands of numeric comparison operators are
1705 integers when their expansion is in "skipping" mode (fixes bug
1706 introduced by 4.64-PH/07).
1708 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1709 child addresses, Exim now panics and dies. Previously, because the count
1710 is held in a short int, deliveries were likely to be lost. As such a
1711 large number of recipients for a single message is ridiculous
1712 (performance will be very, very poor), I have chosen to impose a limit
1713 rather than extend the field.
1719 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1720 leftover -K file (the existence of which was triggered by #402).
1721 While we were at it, introduced process PID as part of the -K
1722 filename. This should rule out race conditions when creating
1725 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1726 processing considerably. Previous code took too long for large mails,
1727 triggering a timeout which in turn triggers #401.
1729 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1730 in the DK code in transports.c. sendfile() is not really portable,
1731 hence the _LINUX specificness.
1733 TF/01 In the add_headers option to the mail command in an Exim filter,
1734 there was a bug that Exim would claim a syntax error in any
1735 header after the first one which had an odd number of characters
1738 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1739 callout verification, Exim cached a "reject" for the entire domain. This
1740 is correct for most verifications, but it is not correct for a recipient
1741 verification with use_sender or use_postmaster set, because in that case
1742 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1743 case of MAIL FROM:<> rejection from other early rejections (e.g.
1744 rejection of HELO). When verifying a recipient using a non-null MAIL
1745 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1746 Whatever the result of the callout, the value of the domain cache is
1747 left unchanged (for any other kind of callout, getting as far as trying
1748 RCPT means that the domain itself is ok).
1750 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1753 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1754 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1755 session, the daemon ignores ECONNECT errors and logs others; it now
1756 ignores EPIPE as well.
1758 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1759 (quoted-printable decoding).
1761 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1762 later the small subsequent patch to fix an introduced bug.
1764 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1766 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1768 PH/08 An error is now given if message_size_limit is specified negative.
1770 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1771 to be given (somewhat) arbitrary names.
1773 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1776 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1777 miscellaneous code fixes
1779 PH/10 Added the log_reject_target ACL modifier to specify where to log
1782 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1783 hostname. This is wrong, because it relates to the incoming message (and
1784 probably the interface on which it is arriving) and not to the outgoing
1785 callout (which could be using a different interface). This has been
1786 changed to use the value of the helo_data option from the smtp transport
1787 instead - this is what is used when a message is actually being sent. If
1788 there is no remote transport (possible with a router that sets up host
1789 addresses), $smtp_active_hostname is used.
1791 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1792 tweaks were necessary in order to get it to work (see also 21 below):
1793 (a) The code assumed that strncpy() returns a negative number on buffer
1794 overflow, which isn't the case. Replaced with Exim's string_format()
1796 (b) There were several signed/unsigned issues. I just did the minimum
1797 hacking in of casts. There is scope for a larger refactoring.
1798 (c) The code used strcasecmp() which is not a standard C function.
1799 Replaced with Exim's strcmpic() function.
1800 (d) The code set only $1; it now sets $auth1 as well.
1801 (e) A simple test gave the error "authentication client didn't specify
1802 service in request". It would seem that Dovecot has changed its
1803 interface. Fortunately there's a specification; I followed it and
1804 changed what the client sends and it appears to be working now.
1806 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1809 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1810 address A is aliased to B and C, where B exists and C does not. Without
1811 -v the output is "A verified" because verification stops after a
1812 successful redirection if more than one address is generated. However,
1813 with -v the child addresses are also verified. Exim was outputting "A
1814 failed to verify" and then showing the successful verification for C,
1815 with its parentage. It now outputs "B failed to verify", showing B's
1816 parentage before showing the successful verification of C.
1818 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1819 look up a TXT record in a specific list after matching in a combined
1822 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1823 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1824 they consult the DNS. I had assumed they would set it the way they
1825 wanted; and indeed my experiments on Linux seem to show that in some
1826 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1827 To be on the safe side, however, I have now made the interface to
1828 host_find_byname() similar to host_find_bydns(), with an argument
1829 containing the DNS resolver options. The host_find_byname() function now
1830 sets these options at its start, just as host_find_bydns() does. The smtp
1831 transport options dns_qualify_single and dns_search_parents are passed to
1832 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1833 of host_find_byname() use the default settings of RES_DEFNAMES
1834 (qualify_single) but not RES_DNSRCH (search_parents).
1836 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1837 spool_read_header() do less string testing, by means of a preliminary
1838 switch on the second character of optional "-foo" lines. (This is
1839 overdue, caused by the large number of possibilities that now exist.
1840 Originally there were few.) While I was there, I also converted the
1841 str(n)cmp tests so they don't re-test the leading "-" and the first
1842 character, in the hope this might squeeze out yet more improvement.
1844 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1845 flag allowing group syntax was set by the header_syntax check but not
1846 turned off, possible causing trouble later; (2) The flag was not being
1847 set at all for the header_verify test, causing "group"-style headers to
1848 be rejected. I have now set it in this case, and also caused header_
1849 verify to ignore an empty address taken from a group. While doing this, I
1850 came across some other cases where the code for allowing group syntax
1851 while scanning a header line wasn't quite right (mostly, not resetting
1852 the flag correctly in the right place). These bugs could have caused
1853 trouble for malformed header lines. I hope it is now all correct.
1855 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1856 with the "reply" argument non-NULL. The code, however (which originally
1857 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1858 but it didn't always do it. This confused somebody who was copying the
1859 code for some other use. I have removed all the tests.
1861 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1862 feature that was used to support insecure browsers during the U.S. crypto
1863 embargo. It requires special client support, and Exim is probably the
1864 only MTA that supported it -- and would never use it because real RSA is
1865 always available. This code has been removed, because it had the bad
1866 effect of slowing Exim down by computing (never used) parameters for the
1867 RSA_EXPORT functionality.
1869 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1870 authenticator to fail if there's a tab character in the incoming data
1871 (there should never be unless someone is messing about, as it's supposed
1872 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1873 if the connection is using TLS or if the remote IP is the same as the
1874 local IP, and the "valid-client-cert option" if a client certificate has
1877 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1878 authenticators. This can be used for authorization after authentication
1879 succeeds. (In the case of plaintext, it servers for both authentication
1882 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1883 if any retry times were supplied.
1885 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1886 connection, where there is no client IP address to check. In this
1887 situation, the verify now always succeeds.
1889 PH/25 Applied John Jetmore's -Mset patch.
1891 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1893 PH/27 In a string expansion for a processed (not raw) header when multiple
1894 headers of the same name were present, leading whitespace was being
1895 removed from all of them, but trailing whitespace was being removed only
1896 from the last one. Now trailing whitespace is removed from each header
1897 before concatenation. Completely empty headers in a concatenation (as
1898 before) are ignored.
1900 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1901 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1903 PH/29 [Removed. This was a change that I later backed out, and forgot to
1904 correct the ChangeLog entry (that I had efficiently created) before
1905 committing the later change.]
1907 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1908 address errors (4xx response to RCPT) over the same connection as other
1909 messages routed to the same hosts. Such deliveries are always "forced",
1910 so retry times are not inspected. This resulted in far too many retries
1911 for the affected addresses. The effect occurred only when there were more
1912 hosts than the hosts_max_try setting in the smtp transport when it had
1913 the 4xx errors. Those hosts that it had tried were not added to the list
1914 of hosts for which the message was waiting, so if all were tried, there
1915 was no problem. Two fixes have been applied:
1917 (i) If there are any address or message errors in an SMTP delivery, none
1918 of the hosts (tried or untried) are now added to the list of hosts
1919 for which the message is waiting, so the message should not be a
1920 candidate for sending over the same connection that was used for a
1921 successful delivery of some other message. This seems entirely
1922 reasonable: after all the message is NOT "waiting for some host".
1923 This is so "obvious" that I'm not sure why it wasn't done
1924 previously. Hope I haven't missed anything, but it can't do any
1925 harm, as the worst effect is to miss an optimization.
1927 (ii) If, despite (i), such a delivery is accidentally attempted, the
1928 routing retry time is respected, so at least it doesn't keep
1929 hammering the server.
1931 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1932 in ${readsocket because some servers need this prod.
1934 PH/32 Added some extra debug output when updating a wait-xxx database.
1936 PH/33 The hint "could be header name not terminated by colon", which has been
1937 given for certain expansion errors for a long time, was not being given
1938 for the ${if def:h_colon_omitted{... case.
1940 PH/34 The spec says: "With one important exception, whenever a domain list is
1941 being scanned, $domain contains the subject domain." There was at least
1942 one case where this was not true.
1944 PH/35 The error "getsockname() failed: connection reset by peer" was being
1945 written to the panic log as well as the main log, but it isn't really
1946 panic-worthy as it just means the connection died rather early on. I have
1947 removed the panic log writing for the ECONNRESET error when getsockname()
1950 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1951 runs only) independently of the message's sender address. This meant
1952 that, if the 4xx error was in fact related to the sender, a different
1953 message to the same recipient with a different sender could confuse
1954 things. In particualar, this can happen when sending to a greylisting
1955 server, but other circumstances could also provoke similar problems.
1956 I have changed the default so that the retry time for these errors is now
1957 based a combination of the sender and recipient addresses. This change
1958 can be overridden by setting address_retry_include_sender=false in the
1961 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1962 remote server are returned as part of bounce messages. This was not
1963 happening for LMTP over a pipe (the lmtp transport), but now it is the
1964 same for both kinds of LMTP.
1966 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1967 in header lines that were in fact CNAMEs. This is no longer the case.
1969 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1970 and queue runs started by the daemon processed all messages. This has
1971 been fixed so that -R and -S can now usefully be given with -q<time>.
1973 PH/40 Import PCRE release 6.7 (fixes some bugs).
1975 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1977 PH/42 Give an error if -q is specified more than once.
1979 PH/43 Renamed the variables $interface_address and $interface_port as
1980 $received_ip_address and $received_port, to make it clear that these
1981 values apply to message reception, and not to the outgoing interface when
1982 a message is delivered. (The old names remain recognized, of course.)
1984 PH/44 There was no timeout on the connect() call when using a Unix domain
1985 socket in the ${readsocket expansion. There now is.
1987 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1988 be meaningful with "accept".
1990 SC/01 Eximstats V1.43
1991 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1993 SC/02 Eximstats V1.44
1994 Use a glob alias rather than an array ref in the generated
1995 parser. This improves both readability and performance.
1997 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1998 Collect SpamAssassin and rejection statistics.
1999 Don't display local sender or destination tables unless
2000 there is data to show.
2001 Added average volumes into the top table text output.
2003 SC/04 Eximstats V1.46
2004 Collect data on the number of addresses (recipients)
2005 as well as the number of messages.
2007 SC/05 Eximstats V1.47
2008 Added 'Message too big' to the list of mail rejection
2009 reasons (thanks to Marco Gaiarin).
2011 SC/06 Eximstats V1.48
2012 Mainlog lines which have GMT offsets and are too short to
2013 have a flag are now skipped.
2015 SC/07 Eximstats V1.49 (Alain Williams)
2016 Added the -emptyok flag.
2018 SC/08 Eximstats V1.50
2019 Fixes for obtaining the IP address from reject messages.
2021 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2022 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2023 whitesspace changes from 4.64-PH/27
2025 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2028 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2029 are found, allow negative numbers in numeric criteria)
2031 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2033 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2034 to match changes made in 4.64-PH/43
2036 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2038 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2039 "domain" is actually an IP address, and force a failure. This locks out
2040 those revolvers/nameservers that support "A-for-A" lookups, in
2041 contravention of the specifications.
2043 PH/48 When a host name was looked up from an IP address, and the subsequent
2044 forward lookup of the name timed out, the host name was left in
2045 $sender_host_name, contrary to the specification.
2047 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2048 restricted to single-key lookups, Exim was not diagnosing an error if
2049 * or *@ was used with a query-style lookup.
2051 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2053 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2054 long ago, but noone interested enough thought of it. Let's just say that
2055 the "1.1" means that there are some new functions that weren't there at
2056 some point in the past.
2058 PH/51 Error processing for expansion failure of helo_data from an smtp
2059 transport during callout processing was broken.
2061 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2062 tested/used via the -bh/-bhc/-bs options.
2064 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2065 bug, fixed in subsequent PCRE releases).
2067 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2068 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2070 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2076 SC/01 Use a glob alias rather than an array ref in eximstats generated
2077 parser. This improves both readability and performance.
2079 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2080 Don't display local sender or destination tables in eximstats unless
2081 there is data to show.
2082 Added average volumes into the eximstats top table text output.
2084 SC/03 Collect data on the number of addresses (recipients) as well
2085 as the number of messages in eximstats.
2087 TF/01 Correct an error in the documentation for the redirect router. Exim
2088 does (usually) call initgroups() when daemonizing.
2090 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2091 with consistent privilege compared to when running as a daemon.
2093 TF/03 Note in the spec that $authenticated_id is not set for local
2094 submissions from trusted users.
2096 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2097 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2099 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2100 by adding some example configuration directives to the default
2101 configuration file. A little bit of work is required to uncomment the
2102 directives and define how usernames and passwords are checked, but
2103 there is now a framework to start from.
2105 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2106 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2107 without this. I don't know how relevant this is to other LDAP libraries.
2109 PH/02 Add the verb name to the "unknown ACL verb" error.
2111 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2113 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2115 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2116 directory not expanded when it should be if an expanded home directory
2117 was set for the address (which is overridden by the transport).
2119 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2122 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2123 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2124 because it is too late at that time, and has no effect.
2126 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2127 security issue with \' (bugzilla #107). I could not use the
2128 PQescapeStringConn() function, because it needs a PGconn value as one of
2131 PH/08 When testing addresses using -bt, indicate those final addresses that
2132 are duplicates that would not cause an additional delivery. At least one
2133 person was confused, thinking that -bt output corresponded to deliveries.
2134 (Suppressing duplicates isn't a good idea as you lose the information
2135 about possibly different redirections that led to the duplicates.)
2137 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2138 systems where poll() doesn't work, in particular OS X.
2140 PH/10 Added more information to debugging output for retry time not reached.
2142 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2143 operations in malware.c.
2145 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2148 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2149 syslog_duplication set false, log lines that would normally be written
2150 both the the main log and to the reject log were not written to syslog at
2153 PH/14 In the default configuration, change the use of "message" in ACL warn
2154 statements to "add_header".
2156 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2157 not followed by a command (e.g. "seen endif").
2159 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2160 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2163 PH/17 Added extra conditions to the default value of delay_warning_condition
2167 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2168 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2169 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2172 The Auto-Submitted: and various List- headers are standardised, whereas I
2173 don't think Precedence: ever was.
2175 PH/18 Refactored debugging code in route_finduser() to show more information,
2176 in particular, the error code if getpwnam() issues one.
2178 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2179 This is apparently needed in addition to the PH/07 change above to avoid
2180 any possible encoding problems.
2182 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2183 but not after initializing Perl.
2185 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2186 output them only if debugging. By default they are written stderr,
2187 apparently, which is not desirable.
2189 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2192 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2195 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2197 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2198 authenticated or an ident call has been made. Suppress the default
2199 values for $authenticated_id and $authenticated_sender (but permit -oMai
2200 and -oMas) when testing with -bh.
2202 PH/24 Re-jigged the order of the tests in the default configuration so that the
2203 tests for valid domains and recipients precede the DNS black list and CSA
2204 tests, on the grounds that those ones are more expensive.
2206 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2207 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2208 This bug exists in every version of Exim that I still have, right back to
2211 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2212 However, an attempt to turn on case-sensitivity in a regex key by
2213 including (?-i) didn't work because the subject string was already
2214 lowercased, and the effects were non-intuitive. It turns out that a
2215 one-line patch can be used to allow (?-i) to work as expected.
2221 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2222 other effects) broke the use of negated acl sub-conditions.
2224 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2227 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2228 "Deny" causes Exim to reject the incoming connection with a 554 error.
2229 Unfortunately, if there is a major crisis, such as a disk failure,
2230 tcp-wrappers gives "deny", whereas what one would like would be some
2231 kind of temporary error. A kludge has been added to help with this.
2232 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2233 554 error is used if errno is still zero or contains ENOENT (which occurs
2234 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2237 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2239 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2240 errors. Otherwise a message that provokes a temporary error (when other
2241 messages do not) can cause a whole host to time out.
2243 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2244 addresses were routed directly to files or pipes from a redirect router.
2245 File deliveries just didn't batch; pipe deliveries might have suffered
2248 PH/06 A failure to get a lock for a hints database would erroneously always say
2249 "Failed to get write lock", even when it was really a read lock.
2251 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2252 of 0600. This has been changed to use the value of the lockfile_mode
2253 option (which defaults to 0600).
2255 PH/08 Applied small patch from the Sieve maintainer.
2257 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2258 folder from quota calculations, a direct delivery into this folder messed
2259 up the contents of the maildirsize file. This was because the regex was
2260 used only to exclude .Trash (or whatever) when the size of the mailbox
2261 was calculated. There was no check that a delivery was happening into an
2262 excluded directory. This bug has been fixed by ignoring all quota
2263 processing for deliveries into excluded directories.
2265 PH/10 Added the maildirfolder_create_regex option to appendfile.
2271 PH/01 The code for finding all the local interface addresses on a FreeBSD
2272 system running IPv6 was broken. This may well have applied to all BSD
2273 systems, as well as to others that have similar system calls. The broken
2274 code found IPv4 interfaces correctly, but gave incorrect values for the
2275 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2276 that it would not match correctly against @[] and not recognize the IPv6
2279 PH/02 The ipliteral router was not recognizing addresses of the form user@
2280 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2282 PH/03 Added disable_ipv6.
2284 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2285 decoded form, because it is most often used to construct To: headers
2286 lines in autoreplies, and the decoded form may well be syntactically
2287 invalid. However, $reply_address has leading white space removed, and all
2288 newlines turned into spaces so that the autoreply transport does not
2291 PH/05 If group was specified without a user on a router, and no group or user
2292 was specified on a transport, the group from the router was ignored.
2294 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2295 for visible compile-time settings that can be used to change these
2296 numbers, for those that want even more. Backwards compatibility with old
2297 spool files has been maintained. However, going back to a previous Exim
2298 release will lost any variables that are in spool files.
2300 PH/07 Two small changes when running in the test harness: increase delay when
2301 passing a TCP/IP connection to a new process, in case the original
2302 process has to generate a bounce, and remove special handling of
2303 127.0.0.2 (sic), which is no longer necessary.
2305 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2306 be the same on different OS.
2308 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2311 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2312 whether --show-vars was specified or not
2314 JJ/02 exipick: Added support for new ACL variable spool format introduced
2317 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2318 syntactically invalid From: or Reply-to: line, and a filter used this to
2319 generate an autoreply, and therefore failed to obtain an address for the
2320 autoreply, Exim could try to deliver to a non-existent relative file
2321 name, causing unrelated and misleading errors. What now happens is that
2322 it logs this as a hard delivery error, but does not attempt to create a
2325 PH/11 The exinext utility has a -C option for testing purposes, but although
2326 the given file was scanned by exinext itself; it wasn't being passed on
2327 when Exim was called.
2329 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2330 an end-of-file indication when reading a command response.
2332 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2333 compiled. In many other places in Exim, IPv6 addresses are always
2334 recognized, so I have changed this. It also means that IPv4 domain
2335 literals of the form [IPV4:n.n.n.n] are now always recognized.
2337 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2338 used if the router is not running as root, for example, when verifying at
2339 ACL time, or when using -bh. The debugging output from this situation was
2340 non-existent - all you got was a failure to exec. I have made two
2343 (a) Failures to set uid/gid, the current directory, or a process leader
2344 in a subprocess such as that created by queryprogram now generate
2345 suitable debugging ouput when -d is set.
2347 (b) The queryprogram router detects when it is not running as root,
2348 outputs suitable debugging information if -d is set, and then runs
2349 the subprocess without attempting to change uid/gid.
2351 PH/15 Minor change to Makefile for building test_host (undocumented testing
2354 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2355 additional section of a DNS packet that returns MX or SRV records.
2356 Instead, it always explicitly searches for A/AAAA records. This avoids
2357 major problems that occur when a DNS server includes only records of one
2358 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2359 fixed another bug: if SRV records were looked up and the corresponding
2360 address records were *not* found in the additional section, the port
2361 values from the SRV records were lost.
2363 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2364 using the correct key (the original address) when searching the retry
2365 rules in order to find which one to use for generating the retry hint.
2367 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2368 adding the default one. Similarly, if it contains a Reply-To: header, the
2369 errors_reply_to option, if set, is not used.
2371 PH/19 When calculating a retry time, Exim used to measure the "time since
2372 failure" by looking at the "first failed" field in the retry record. Now
2373 it does not use this if it is later than than the arrival time of the
2374 message. Instead it uses the arrival time. This makes for better
2375 behaviour in cases where some deliveries succeed, thus re-setting the
2376 "first failed" field. An example is a quota failure for a huge message
2377 when small messages continue to be delivered. Without this change, the
2378 "time since failure" will always be short, possible causing more frequent
2379 delivery attempts for the huge message than are intended.
2380 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2382 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2383 $1, $2, $3) because the numerical variables can be reset during some
2384 expansion items (e.g. "match"), thereby losing the authentication data.
2386 PH/21 Make -bV show the size of off_t variables so that the test suite can
2387 decide whether to run tests for quotas > 2G.
2389 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2390 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2391 filecount value is greater than 2G or if a quota value is greater than 2G
2392 on a system where the size of off_t is not greater than 4, a panic error
2395 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2396 never match. The debug and -bh output now contains an explicit error
2397 message indicating a malformed IPv4 address or mask.
2399 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2400 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2401 PH/23 above applies.
2403 PH/25 Do not write to syslog when running in the test harness. The only
2404 occasion when this arises is a failure to open the main or panic logs
2405 (for which there is an explicit test).
2407 PH/26 Added the /no_tell option to "control=freeze".
2409 PH/27 If a host name lookup failed very early in a connection, for example, if
2410 the IP address matched host_lookup and the reverse lookup yielded a name
2411 that did not have a forward lookup, an error message of the form "no IP
2412 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2413 could be logged. Now it outputs the IP address instead of "NULL".
2415 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2416 allows the sender and the authenticated sender to be set when
2417 submitting a message from within Exim. Since child_open_exim() is
2418 documented for local_scan(), the new function should be too.
2420 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2421 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2422 results in an empty string is now treated as unset.
2424 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2426 PH/31 Added sender_verify_fail logging option.
2428 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2429 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2430 tidied the source and removed it altogether.
2432 PH/33 When a queue run was abandoned because the load average was too high, a
2433 log line was always written; now it is written only if the queue_run log
2434 selector is set. In addition, the log line for abandonment now contains
2435 information about the queue run such as the pid. This is always present
2436 in "start" and "stop" lines but was omitted from the "abandon" line.
2438 PH/34 Omit spaces between a header name and the colon in the error message that
2439 is given when verify = headers_syntax fails (if there are lots of them,
2440 the message gets confusing).
2442 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2443 names, as there are now some PTR records that contain slashes. This check
2444 is only to protect against broken name servers that fall over on strange
2445 characters, so the fact that it applies to all lookups doesn't matter.
2447 PH/36 Now that the new test suite is complete, we can remove some of the
2448 special code in Exim that was needed for the old test suite. For example,
2449 sorting DNS records because real resolvers return them in an arbitrary
2450 order. The new test suite's fake resolver always returns records in the
2453 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2454 bounces) except when queue_only is set, to avoid logging races between
2455 the different processes.
2457 PH/38 Panic-die if .include specifies a non-absolute path.
2459 PH/39 A tweak to the "H" retry rule from its user.
2461 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2462 a label. They prevented compilation on older perls.
2464 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2465 a warning to be raised on newish perls.
2467 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2468 on queue. Changes to match documented behaviour of showing count of
2469 messages matching specified criteria.
2471 PH/40 Changed the default ident timeout from 30s to 5s.
2473 PH/41 Added support for the use of login_cap features, on those BSD systems
2474 that have them, for controlling the resources used by pipe deliveries.
2476 PH/42 The content-scanning code uses fopen() to create files in which to put
2477 message data. Previously it was not paying any attention to the mode of
2478 the files. Exim runs with umask(0) because the rest of the code creates
2479 files with open(), and sets the required mode explicitly. Thus, these
2480 files were ending up world-writeable. This was not a big issue, because,
2481 being within the spool directory, they were not world-accessible. I have
2482 created a function called modefopen, which takes an additional mode
2483 argument. It sets umask(777), creates the file, chmods it to the required
2484 mode, then resets the umask. All the relevant calls to fopen() in the
2485 content scanning code have been changed to use this function.
2487 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2488 to 24 hours. This avoids potential overflow problems when processing G
2489 and H retry rules. I suspect nobody ever tinkers with this value.
2491 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2493 PH/45 When the plaintext authenticator is running as a client, the server's
2494 challenges are checked to ensure they are valid base64 strings. By
2495 default, the authentication attempt is cancelled if an invalid string is
2496 received. Setting client_ignore_invalid_base64 true ignores these errors.
2497 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2498 they are received. Thus, the responses can be made to depend on the
2499 challenges. If an invalid string is ignored, an empty string is placed in
2502 PH/46 Messages that are created by the autoreply transport now contains a
2503 References: header, in accordance with RFCs 2822 and 3834.
2505 PH/47 Added authenticated_sender_force to the smtp transport.
2507 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2509 PH/49 Installed latest patch from the Sieve maintainer.
2511 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2512 was also set, the appendfile transport was unnecessarily scanning a
2513 directory of message files (e.g. for maildir delivery) to find the count
2514 of files (along with the size), even though it did not need this
2515 information. It now does the scan only if it needs to find either the
2516 size of the count of files.
2518 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2520 PH/52 Two bugs concerned with error handling when the smtp transport is
2523 (i) Exim was not creating retry information for temporary errors given
2524 for individual recipients after the DATA command when the smtp transport
2525 was used in LMTP mode. This meant that they could be retried too
2526 frequently, and not timed out correctly.
2528 (ii) Exim was setting the flag that allows error details to be returned
2529 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2530 recipients that were returned after the DATA command.
2532 PH/53 This is related to PH/52, but is more general: for any failing address,
2533 when detailed error information was permitted to be returned to the
2534 sender, but the error was temporary, then after the final timeout, only
2535 "retry timeout exceeded" was returned. Now it returns the full error as
2536 well as "retry timeout exceeded".
2538 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2539 do this, and (what is worse) MTAs that accept it.
2541 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2542 will now be deprecated.
2544 PH/56 New os.c-cygwin from the Cygwin maintainer.
2546 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2547 formats (previously only available in exim formats via -bpr, -bpru,
2548 and -bpra. Now also available in native and exiqgrep formats)
2550 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2551 with very large, slow to parse queues
2553 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2555 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2557 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2558 responses to authentication challenges, though it was showing the
2559 challenges; (ii) I've removed the CR characters from the debug output for
2562 PH/58 Allow for the insertion of a newline as well as a space when a string
2563 is turned into more than one encoded-word during RFC 2047 encoding. The
2564 Sieve code now uses this.
2566 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2567 data_4xx, lost_connection, tls_required.
2569 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2570 message was being sent as an SMTP response.
2572 PH/61 Add -l and -k options to exicyclog.
2574 PH/62 When verifying, if an address was redirected to one new address, so that
2575 verification continued, and the new address failed or deferred after
2576 having set something in $address_data, the value of $address_data was not
2577 passed back to the ACL. This was different to the case when no
2578 redirection occurred. The value is now passed back in both cases.
2580 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2581 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2582 use login_cap.h, so on its own it isn't the distinguishing feature. The
2583 new name refers directly to the setclassresources() function.
2585 PH/65 Added configuration files for NetBSD3.
2587 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2589 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2590 is preferred over IPv4.
2592 PH/68 The bounce_return_message and bounce_return_body options were not being
2593 honoured for bounces generated during the reception of non-SMTP messages.
2594 In particular, this applied to messages rejected by the ACL. This bug has
2595 been fixed. However, if bounce_return_message is true and bounce_return_
2596 body is false, the headers that are returned for a non-SMTP message
2597 include only those that have been read before the error was detected.
2598 (In the case of an ACL rejection, they have all been read.)
2600 PH/69 The HTML version of the specification is now built in a directory called
2601 spec_html instead of spec.html, because the latter looks like a path with
2602 a MIME-type, and this confuses some software.
2604 PH/70 Catch two compiler warnings in sieve.c.
2606 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2607 function verify_get_ident() calls ip_connect() to connect a socket, but
2608 if the "connect()" function timed out, ip_connect() used to close the
2609 socket. However, verify_get_ident() also closes the socket later, and in
2610 between Exim writes to the log, which may get opened at this point. When
2611 the socket was closed in ip_connect(), the log could get the same file
2612 descriptor number as the socket. This naturally causes chaos. The fix is
2613 not to close the socket in ip_connect(); the socket should be closed by
2614 the function that creates it. There was only one place in the code where
2615 this was missing, in the iplookup router, which I don't think anybody now
2616 uses, but I've fixed it anyway.
2618 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2619 well as to direct DNS lookups. Otherwise the handling of names in host
2620 lists is inconsistent and therefore confusing.
2626 PH/01 Two changes to the default runtime configuration:
2628 (1) Move the checks for relay_from_hosts and authenticated clients from
2629 after to before the (commented out) DNS black list checks.
2631 (2) Add control=submission to the relay_from_hosts and authenticated
2632 clients checks, on the grounds that messages accepted by these
2633 statements are most likely to be submissions.
2635 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2637 (1) Generate an error if the third argument for the ${prvs expansion is
2640 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2643 (3) Reset the variables that are obtained from the first argument of
2644 ${prvscheck and used in the second argument before leaving the code,
2645 because their memory is reclaimed, so using them afterwards may do
2648 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2649 one (it's much easier than Tom thought :-).
2651 (5) Because of (4), we can now allow for the use of $prvscheck_result
2652 inside the third argument.
2654 PH/03 For some reason, the default setting of PATH when running a command from
2655 a pipe transport was just "/usr/bin". I have changed it to
2658 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2659 anything to be listed in the output from -bV.
2661 PH/05 When a filter generated an autoreply, the entire To: header line was
2662 quoted in the delivery log line, like this:
2664 => >A.N.Other <ano@some.domain> <original@ddress> ...
2666 This has been changed so that it extracts the operative address. There
2667 may be more than one such address. If so, they are comma-separated, like
2670 => >ano@some.domain,ona@other.domain <original@ddress> ...
2672 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2673 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2674 not being looked up in the rDNS to get a host name, Exim was showing the
2675 IP address twice in Received: lines, even though the IP addresses were
2676 identical. For example:
2678 Received: from [1.2.3.4] (helo=[1.2.3.4])
2680 However, if the real host name was known, it was omitting the HELO data
2681 if it matched the actual IP address. This has been tidied up so that it
2682 doesn't show the same IP address twice.
2684 PH/07 When both +timestamp and +memory debugging was on, the value given by
2685 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2686 called by the debug printing, thereby overwriting the timestamp buffer.
2687 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2689 PH/08 When the original message was included in an autoreply transport, it
2690 always said "this is a copy of the message, including all the headers",
2691 even if body_only or headers_only was set. It now gives an appropriate
2694 PH/09 Applied a patch from the Sieve maintainer which:
2696 o fixes some comments
2697 o adds the (disabled) notify extension core
2698 o adds some debug output for the result of if/elsif tests
2699 o points to the current vacation draft in the documentation
2700 and documents the missing references header update
2704 o fixes a bug in processing the envelope test (when testing
2705 multiple envelope elements, the last element determinted the
2708 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2709 Electronic Mail") by including:
2711 Auto-submitted: auto-generated
2713 in the messages that it generates (bounce messages and others, such as
2714 warnings). In the case of bounce messages for non-SMTP mesages, there was
2715 also a typo: it was using "Auto_submitted" (underscore instead of
2716 hyphen). Since every message generated by Exim is necessarily in response
2717 to another message, thes have all been changed to:
2719 Auto-Submitted: auto-replied
2721 in accordance with these statements in the RFC:
2723 The auto-replied keyword:
2725 - SHOULD be used on messages sent in direct response to another
2726 message by an automatic process,
2728 - MUST NOT be used on manually-generated messages,
2730 - MAY be used on Delivery Status Notifications (DSNs) and Message
2731 Disposition Notifications (MDNs),
2733 - MUST NOT be used on messages generated by automatic or periodic
2734 processes, except for messages which are automatic responses to
2737 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2738 to the default Received: header definition.
2740 PH/12 Added log selector acl_warn_skipped (default on).
2742 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2743 variables because (a) they are in the wrong storage pool and (b) even if
2744 they were copied, it wouldn't work properly because of the caching.
2746 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2747 checking when decoding. Apparently there are clients that generate
2748 overlong encoded strings. Why am I not surprised?
2750 PH/15 If the first argument of "${if match_address" was not empty, but did not
2751 contain an "@" character, Exim crashed. Now it writes a panic log message
2752 and treats the condition as false.
2754 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2756 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2757 extension "envelope-auth". The code is finished and in agreement with
2758 other implementations, but there is no documentation so far and in fact,
2759 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2760 not changing the active code.
2762 Print executed "if" and "elsif" statements when debugging is used. This
2763 helps a great deal to understand what a filter does.
2765 Document more things not specified clearly in RFC3028. I had all this
2766 sorted out, when out of a sudden new issues came to my mind. Oops."
2768 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2771 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2772 canonical form (as documented). However, after a host name lookup from
2773 the IP address, check_host() was doing a simple string comparison with
2774 addresses acquired from the DNS when checking that the found name did
2775 have the original IP as one of its addresses. Since any found IPv6
2776 addresses are likely to be in abbreviated form, the comparison could
2777 fail. Luckily, there already exists a function for doing the comparison
2778 by converting both addresses to binary, so now that is used instead of
2779 the text comparison.
2781 PH/20 There was another similar case to PH/19, when a complete host name was
2782 given in a host list; looking up its IP address could give an abbreviated
2783 form, whereas the current host's name might or might not be abbreviated.
2784 The same fix has been applied.
2790 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2791 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2794 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2795 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2797 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2799 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2800 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2801 Also, for "Auto-Submitted", treat anything other than "no" as
2802 non-personal, in accordance with RFC 3834. (Previously it treated
2803 anything starting "auto-" as non-personal.)
2805 TF/01 The control=submission/name=... option had a problem with syntax
2806 errors if the name included a slash character. The /name= option
2807 now slurps the rest of the string, so it can include any characters
2808 but it must come last in the list of options (after /sender_retain
2811 PH/05 Some modifications to the interface to the fake nameserver for the new
2819 TK/01 Added the "success_on_redirect" address verification option. See
2820 NewStuff for rationale and an example.
2822 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2824 PH/02 Patch to exigrep to allow it to work on syslog lines.
2826 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2827 fread() to skip over the body file's header line, because in Cygwin the
2828 header line is locked and is inaccessible.
2830 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2831 co-exist for some time) to make it clear that it is the Exim ID that is
2832 referenced, not the Message-ID: header line.
2834 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2835 string_format() function, because snprintf() does not exist on all
2838 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2839 ${stat: expansion item.
2841 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2842 protocol synchronization error", to keep the pedants happy.
2844 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2845 well as for IRIX systems, when gcc is being used. See the host.c source
2848 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2850 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2853 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2854 transport and to the smtp transport in LMTP mode.
2856 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2858 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2859 run a filter in a subprocess. This could lead to confusion in subsequent
2860 lookups in the parent process. There should also be a search_tidyup() at
2861 the end of the subprocess.
2863 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2864 only if the host matched helo_try_verify_hosts, which caused the
2865 verification to occur when the EHLO/HELO command was issued. The ACL just
2866 tested the remembered result. Now, if a previous verification attempt has
2867 not happened, "verify = helo" does it there and then.
2869 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2871 TK/03 Fix log output including CR from clamd.
2873 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2874 exist provoked a memory error which could cause a segfault.
2876 PH/15 Installed PCRE 6.2
2878 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2880 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2881 of the problem. Specifically, suggested +O2 rather than +O1 for the
2884 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2886 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2887 record, the port was not correctly logged when the outgoing_port log
2888 selector was set (it logged the transort's default port).
2890 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2891 fallback_hosts, and "hosts" in the smtp transport.
2893 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2894 host errors such as "Connection refused".
2896 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2897 authentication with radiusclient 0.4.9:
2899 - Error returned from rc_read_config was caught wrongly
2900 - Username/password not passed on to radius server due to wrong length.
2902 The presumption is that some radiusclient API changes for 4.51/PH/17
2903 were not taken care of correctly. The code is still untested by me (my
2904 Linux distribution still has 0.3.2 of radiusclient), but it was
2905 contributed by a Radius user.
2907 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2908 expanding the "port" option of the smtp transport.
2910 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2911 while reading a MIME header. Thanks to Tom Hughes for a patch.
2913 PH/24 Include config.h inside local_scan.h so that configuration settings are
2916 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2917 that in an ACL for RCPT (for example), you can examine exactly what was
2920 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2921 commands, but it was not correctly comparing the address with the actual
2922 client host address. Thus, it would show the EHLO address in Received:
2923 header lines when this was not necessary.
2925 PH/27 Added the % operator to ${eval:}.
2927 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2928 it should be ignoring failures (because with -C, for example, it has lost
2929 privilege). It wasn't ignoring creation failures other than "already
2932 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2935 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2936 because an input line was too long, either on its own, or by virtue of
2937 too many continuations, the temporary file was not being removed, and the
2938 return code was incorrect.
2940 PH/31 Missing "BOOL" in function definition in filtertest.c.
2942 PH/32 Applied Sieve patches from the maintainer.
2944 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2946 PH/33 Added "verify = not_blind".
2948 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2949 Local/Makefile (with some defaults set). These are used in built scripts
2950 such as exicyclog, but they have never been used in the exim_install
2951 script (though there are many overriding facilities there). I have
2952 arranged that the exim_install script now takes note of these two
2955 PH/35 Installed configuration files for Dragonfly.
2957 PH/36 When a locally submitted message by a trusted user did not contain a
2958 From: header, and the sender address was obtained from -f or from an SMTP
2959 MAIL command, and the trusted user did not use -F to supply a sender
2960 name, $originator_name was incorrectly used when constructing a From:
2961 header. Furthermore, $originator_name was used for submission mode
2962 messages from external hosts without From: headers in a similar way,
2963 which is clearly wrong.
2965 PH/37 Added control=suppress_local_fixups.
2967 PH/38 When log_selector = +received_sender was set, and the addition of the
2968 sender made the log line's construction buffer exactly full, or one byte
2969 less than full, an overflow happened when the terminating "\n" was
2972 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2973 when the result of a list match is failure because a DNS lookup failed.
2975 PH/40 RM_COMMAND is now used in the building process.
2977 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2978 the "build-* directories that it finds.
2980 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2981 address in a domain literal was a prefix of an interface address.
2983 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2984 when verifying a sender address, unless rewrite_headers is false.
2986 PH/44 Wrote a long comment about why errors_to addresses are verified as
2987 recipients, not senders.
2989 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2990 the ratelimit ACL was added.
2992 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2994 PH/46 Added extra information about PostgreSQL errors to the error string.
2996 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2997 avoiding the need to install special zones in a real server. This is
2998 backwards compatible; if it can't find the fake resolver, it drops back.
2999 Thus, both old and new test suites can be run.
3001 TF/02 Added util/ratelimit.pl
3003 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3004 clock is set back in time.
3006 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3007 Candler <B.Candler@pobox.com>.
3009 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3010 OK for addresses that are the result of redirections.
3012 PH/48 A number of further additions for the benefit of the new test suite,
3013 including a fake gethostbyname() that interfaces to the fake DNS resolver
3016 TF/06 The fix for widen_domains has also been applied to qualify_single and
3017 search_parents which are the other dnslookup options that can cause
3020 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3023 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3025 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3026 service" if it encountered an MX record with an empty target hostname.
3027 The message is now "an MX or SRV record indicated no SMTP service".
3029 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3030 if the DNS of the sending site is misconfigured. This is quite a
3031 common situation. This change restores the behaviour of treating a
3032 helo verification defer as a failure.
3034 PH/51 If self=fail was set on a router, the bounce message did not include the
3035 actual error message.
3041 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3043 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3044 command itself ended in error, the underlying message about the transport
3045 filter timeout was being overwritten with the pipe command error. Now the
3046 underlying error message should be appended to the second error message.
3048 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3050 PH/02 Reduce the amount of output that "make" produces by default. Full output
3051 can still be requested.
3053 PH/03 The warning log line about a condition test deferring for a "warn" verb
3054 was being output only once per connection, rather than after each
3055 occurrence (because it was using the same function as for successful
3056 "warn" verbs). This seems wrong, so I have changed it.
3058 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3059 it should not have, which might have caused a crash in the right
3060 circumstances, but probably never did.
3062 PH/04 Installed a modified version of Tony Finch's patch to make submission
3063 mode fix the return path as well as the Sender: header line, and to
3064 add a /name= option so that you can make the user's friendly name appear
3067 TF/03 Added the control = fakedefer ACL modifier.
3069 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3070 Mark Lowes for thorough testing.
3072 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3074 TK/03 Merged latest SRS patch from Miles Wilton.
3076 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3077 with the definition in sysexits.h (which is #included earlier).
3078 Fortunately, Exim does not actually use EX_OK. The code used to try to
3079 preserve the sysexits.h value, by assumimg that macro definitions were
3080 scanned for macro replacements. I have been disabused of this notion,
3081 so now the code just undefines EX_OK before #including unistd.h.
3083 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3084 in the smtp transport. When a block could not be written in a single
3085 write() function, the timeout was being re-applied to each part-write.
3086 This seems wrong - if the receiver was accepting one byte at a time it
3087 would take for ever. The timeout is now adjusted when this happens. It
3088 doesn't have to be particularly precise.
3090 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3091 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3093 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3094 without a domain if the check to <postmaster@domain> fails.
3096 SC/01 Eximstats: added -xls and the ability to specify output files
3097 (patch written by Frank Heydlauf).
3099 SC/02 Eximstats: use FileHandles for outputing results.
3101 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3103 SC/04 Eximstats: fixed display of large numbers with -nvr option
3105 SC/05 Eximstats: fixed merging of reports with empty tables.
3107 SC/06 Eximstats: added the -include_original_destination flag
3109 SC/07 Eximstats: removed tabs and trailing whitespace.
3111 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3113 TK/06 MBOX spool code: Add real "From " MBOX separator line
3114 so the .eml file is really in mbox format (even though
3115 most programs do not really care). Patch from Alex Miller.
3117 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3118 The latter is generated from $received_to and is only set if the
3119 message has one envelope recipient. SA can use these headers,
3120 obviously out-of-the-box. Patch from Alex Miller.
3122 PH/08 The ${def test on a variable was returning false if the variable's
3123 value was "0", contrary to what the specification has always said!
3124 The result should be true unless the variable is empty.
3126 PH/09 The syntax error of a character other than { following "${if
3127 def:variable_name" (after optional whitespace) was not being diagnosed.
3128 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3129 accidental colon was present, for example, could give incorrect results.
3131 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3132 result is used (not including appendfile, where other changes are about
3135 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3136 This involved changing a lot of size variables from int to off_t. It
3137 should work with maildirs and everything.
3139 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3140 spamd dying while we are connected to it.
3142 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3145 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3146 function for BDB 4.3.
3148 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3150 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3151 the sources for PCRE have been re-arranged and more files are now
3154 PH/15 The code I had for printing potentially long long variables in PH/11
3155 above was not the best (it lost precision). The length of off_t variables
3156 is now inspected at build time, and an appropriate printing format (%ld
3157 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3158 to be "long long int" or "long int". This is needed for the internal
3159 formatting function string_vformat().
3161 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3162 the configuration file to be ":syslog", then the script "guesses" where
3163 the logs files are, rather than using the compiled in default. In our
3164 case the guess is not the same as the compiled default, so the script
3165 suddenly stopped working when I started to use syslog. The patch checks
3166 to see if log_file_path is "". If so, it attempts to read it from exim
3167 with no configuration file to get the compiled in version, before it
3168 falls back to the previous guessing code."
3170 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3171 implementing BATV in an Exim configuration. See NewStuff for the gory
3174 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3175 Makefile that are specific to HP-UX.
3177 PH/18 If the "use_postmaster" option was set for a recipient callout together
3178 with the "random" option, the postmaster address was used as the MAIL
3179 FROM address for the random test, but not for the subsequent recipient
3180 test. It is now used for both.
3182 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3183 patch removes a few documentation additions to RFC 3028, because the
3184 latest draft now contains them. It adds the new en;ascii-case comparator
3185 and a new error check for 8bit text in MIME parts. Comparator and
3186 require names are now matched exactly. I enabled the subaddress
3187 extension, but it is not well tested yet (read: it works for me)."
3189 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3190 rework some of the code of TK/09 above to avoid the hardwired use of
3191 "%lld" and "long long". Replaced the call to snprintf() with a call to
3194 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3195 records point to non-existent hosts", "retry timeout exceeded", and
3196 "retry time not reached for any host after a long failure period".
3198 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3199 experimental DomainKeys support:
3201 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3202 (2) On an error such as an illegally used "control", the wrong name for
3203 the control was given.
3205 These problems did NOT occur unless DomainKeys support was compiled.
3207 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3209 PH/24 Added ${if match_ip condition.
3211 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3212 databases so that it will be absolutely obvious if a crash occurs in the
3213 DB library. This is a regular occurrence (often caused by mis-matched
3216 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3217 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3218 picked up on a user's system that detects such things. There doesn't seem
3219 to be a gcc warning option for this - only an attribute that has to be
3220 put on the function's prototype. It seems that in Fedora Core 4 they have
3221 set this on a number of new functions. No doubt there will be more in due
3224 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3225 specify a transport. However, if an address that was verified by such a
3226 router was the subject of a callout, Exim crashed because it tried to
3227 read the rcpt_include_affixes from the non-existent transport. Now it
3228 just assumes that the setting of that option is false. This bug was
3229 introduced by 4.51/PH/31.
3231 PH/28 Changed -d+all to exclude +memory, because that information is very
3232 rarely of interest, but it makes the output a lot bigger. People tend to
3233 do -d+all out of habit.
3235 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3236 code in os-type was giving problems when libc.so lives in lib64, like on
3239 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3240 aren't the modern standard, and it seems that some systems' include files
3241 don't always have them. Exim was already checking for some of the newer
3242 ones like T_AAAA, and defining it itself. I've added checks for all the
3243 record types that Exim uses.
3245 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3246 not automatically generating a new one, as it is supposed to. This
3247 prevented TLS from working. If the file did exist, but contained invalid
3248 data, a new version was generated, as expected. It was only the case of a
3249 non-existent file that was broken.
3251 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3252 with a change in libdomainkeys > 0.64.
3254 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3255 from DNS. If the selector record carries the flag, it now has
3256 precedence over the domain-wide flag.
3258 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3260 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3261 the use of a port name that isn't defined in /etc/services) occurred, the
3262 message was deferred as in a normal delivery, and thus remained on the
3263 spool, instead of being failed because of the mua_wrapper setting. This
3264 is now fixed, and I tidied up some of the mua_wrapper messages at the
3267 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3268 the messages in a hash of arrays rather than using individual hashes.
3269 This is a bit cleaner and results in dramatic memory savings, albeit
3270 at a slight CPU cost.
3272 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3273 as requested by Marc Sherman.
3275 SC/10 Eximstats: added histograms for user specified patterns as requested
3278 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3280 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3281 fopen() in the content-scanning modules that did not already have it.
3287 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3288 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3290 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3292 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3294 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3295 to test Sieve filters that use "vacation".
3297 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3298 that changes the way the GnuTLS parameters are stored in the cache file.
3299 The new format can be generated externally. For backward compatibility,
3300 if the data in the cache doesn't make sense, Exim assumes it has read an
3301 old-format file, and it generates new data and writes a new file. This
3302 means that you can't go back to an older release without removing the
3305 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3306 work if there are any delivery delays because "one_time" forces the
3307 parent to be marked "delivered", so its unseen clone is never tried
3308 again. For this reason, Exim now forbids the simultaneous setting of
3311 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3312 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3313 entry if you want to know the details. The fix, however, neglected to
3314 consider the case when local delivery batching is involved. The test for
3315 "previously delivered" was not happening when checking to see if an
3316 address could be batched with a previous (undelivered) one; under
3317 certain circumstances this could lead to multiple deliveries to the same
3320 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3321 in its include files, and this causes problems building Exim.
3323 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3324 header_syntax) but Exim was just ignoring anything given after a slash.
3325 In particular, this caused confusion with an attempt to use "verify =
3326 reverse_host_lookup/defer_ok". An error is now given when options are
3327 supplied for verify items that do not have them. (Maybe reverse_host_
3328 lookup should have a defer_ok option, but that's a different point.)
3330 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3331 defined by RFC 821) to 2048, because there were problems with some AUTH
3332 commands, and RFC 1869 says the size should be increased for extended
3333 SMTP commands that take arguments.
3335 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3338 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3339 "unknown" error; now it says that the functionality isn't in the binary.
3341 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3342 an address' error message when a string expansion fails (syntax or
3343 whatever). Otherwise the password may appear in the log. Following change
3344 PH/42 below, there is no longer a chance of it appearing in a bounce
3347 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3349 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3350 of its addresses was ever tried. (Bugzilla bug #2.)
3352 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3353 the result incorrectly in the debug output. (It correctly added a newline
3354 to what was transported.)
3356 TF/01 Added $received_time.
3358 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3359 commented out examples of how to interface to a virus scanner and to
3360 SpamAssassin. Also added commented examples of av_scanner and
3361 spamd_address settings.
3363 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3364 and controls are allowed in which ACLs. There were a couple of minor
3365 errors. Some of the entries in the conditions table (which is a table of
3366 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3367 negation of where the condition IS allowed.
3369 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3371 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3372 header file does not have a version number, so I've had to invent a new
3373 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3374 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3375 radiusclient), but it was contributed by a Radius user.
3377 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3378 files or directories, for OpenSSL.
3380 PH/19 When an Exim process that is running as root has to create an Exim log
3381 file, it does so in a subprocess that runs as exim:exim so as to get the
3382 ownership right at creation (otherwise, other Exim processes might see
3383 the file with the wrong ownership). There was no test for failure of this
3384 fork() call, which would lead to the process getting stuck as it waited
3385 for a non-existent subprocess. Forks do occasionally fail when resources
3386 run out. I reviewed all the other calls to fork(); they all seem to check
3389 PH/20 When checking for unexpected SMTP input at connect time (before writing
3390 the banner), Exim was not dealing correctly with a non-positive return
3391 from the read() function. If the client had disconnected by this time,
3392 the result was a log entry for a synchronization error with an empty
3393 string after "input=" when read() returned zero. If read() returned -1
3394 (an event I could not check), uninitialized data bytes were printed.
3395 There were reports of junk text (parts of files, etc) appearing after
3398 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3400 PH/22 Added support for macro redefinition, and (re)definition in between
3401 driver and ACL definitions.
3403 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3404 forgetting to use the resulting value; it was using the unexpanded value.
3406 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3407 hadn't been configured. The fix is from Juergen Kreileder, who
3408 understands it better than I do:
3410 "Here's what I see happening with three configured cyrus_sasl
3411 authenticators configured (plain, login, cram-md5):
3413 On startup auth_cyrus_sasl_init() gets called for each of these.
3414 This means three calls to sasl_listmech() without a specified mech_list.
3415 => SASL tests which mechs of all available mechs actually work
3416 => three warnings about OTP not working
3417 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3419 With the patch, sasl_listmech() also gets called three times. But now
3420 SASL's mech_list option is set to the server_mech specified in the the
3421 authenticator. Or in other words, the answer from sasl_listmech()
3422 gets limited to just the mech you're testing for (which is different
3424 => the return list contains just 'plain' or 'login', 'cram-md5' or
3425 nothing depending on the value of ob->server_mech.
3427 I've just tested the patch: Authentication still works fine,
3428 unavailable mechs specified in the exim configuration are still
3429 caught, and the auth.log warnings about OTP are gone."
3431 PH/25 When debugging is enabled, the contents of the command line are added
3432 to the debugging output, even when log_selector=+arguments is not
3435 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3436 answer is "GNU", and only if the return is "GNU/something" is the answer
3439 PH/27 $acl_verify_message is now set immediately after the failure of a
3440 verification in an ACL, and so is available in subsequent modifiers. In
3441 particular, the message can be preserved by coding like this:
3443 warn !verify = sender
3444 set acl_m0 = $acl_verify_message
3446 Previously, $acl_verify_message was set only while expanding "message"
3447 and "log_message" when a very denied access.
3449 PH/28 Modified OS/os.c-Linux with
3451 -#ifndef OS_LOAD_AVERAGE
3452 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3454 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3455 nomenclature these days.)
3457 PH/29 Installed patch from the Sieve maintainer that adds the options
3458 sieve_useraddress and sieve_subaddress to the redirect router.
3460 PH/30 In these circumstances:
3461 . Two addresses routed to the same list of hosts;
3462 . First host does not offer TLS;
3463 . First host accepts first address;
3464 . First host gives temporary error to second address;
3465 . Second host offers TLS and a TLS session is established;
3466 . Second host accepts second address.
3467 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3468 and peerdn, if requested) that were in fact used only for the second
3471 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3472 attention to any local part prefix or suffix that was matched by the
3473 router that accepted the address. It now behaves in the same way as it
3474 does for delivery: the affixes are removed from the local part unless
3475 rcpt_include_affixes is set on the transport.
3477 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3478 timeout during the DATA phase of an incoming message.
3480 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3481 applied a patch sanctioned by the Sieve maintainer.
3483 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3484 the uid or gid is negative. A case of a negative gid caused this to be
3485 noticed. The fix allows for either to be negative.
3487 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3488 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3491 PH/36 The change PH/12 above was broken. Fixed it.
3493 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3494 the grounds that routing the same address twice would always produce the
3495 same answer. This might have been true once, but it is certainly no
3496 longer true now. Routing a child address may depend on the previous
3497 routing that produced that child. Some complicated redirection strategies
3498 went wrong when messages had multiple recipients, and made Exim's
3499 behaviour dependent on the order in which the addresses were given.
3501 I have moved the duplicate checking until after the routing is complete.
3502 Exim scans the addresses that are assigned to local and remote
3503 transports, and removes any duplicates. This means that more work will be
3504 done, as duplicates will always all be routed, but duplicates are
3505 presumably rare, so I don't expect this is of any significance.
3507 For deliveries to pipes, files, and autoreplies, the duplicate checking
3508 still happens during the routing process, since they are not going to be
3511 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3512 It corrects a timeout issue with spamd. This is Ian's comment: "The
3513 background is that sometimes spamd either never reads data from a
3514 connection it has accepted, or it never writes response data. The exiscan
3515 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3516 blindly assumes that writes won't block so it may never time out."
3518 PH/39 Allow G after quota size as well as K and M.
3520 PH/40 The value set for $authenticated_id in an authenticator may not contain
3521 binary zeroes or newlines because the value is written to log lines and
3522 to spool files. There was no check on this. Now the value is run through
3523 the string_printing() function so that such characters are converted to
3524 printable escape sequences.
3526 PH/41 $message_linecount is a new variable that contains the total number of
3527 lines in the message. Compare $body_linecount, which is the count for the
3530 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3531 bounce and delay warning messages, except in certain special cases, which
3534 (a) An SMTP error message from a remote host;
3535 (b) A message specified in a :fail: redirection;
3536 (c) A message specified in a "fail" command in a system filter;
3537 (d) A message specified in a FAIL return from the queryprogram router;
3538 (e) A message specified by the cannot_route_message router option.
3540 In these cases only, Exim does include the error details in bounce and
3541 warning messages. There are also a few cases where bland messages such
3542 as "unrouteable address" or "local delivery error" are given.
3544 PH/43 $value is now also set for the "else" part of a ${run expansion.
3546 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3547 being worked on, but at least Exim now implements the latest version to
3550 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3551 process to complete was treated as a delivery failure, a timeout while
3552 writing the message to the pipe was logged, but erroneously treated as a
3553 successful delivery. Such timeouts include transport filter timeouts. For
3554 consistency with the overall process timeout, these timeouts are now
3555 treated as errors, giving rise to delivery failures by default. However,
3556 there is now a new Boolean option for the pipe transport called
3557 timeout_defer, which, if set TRUE, converts the failures into defers for
3558 both kinds of timeout. A transport filter timeout is now identified in
3561 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3562 systems where "make" and "gmake" are different, calling "gmake" at top
3563 level broke things. I've arranged for the value of $(MAKE) to be passed
3564 from the Makefile to this script so that it can call the same version of
3568 A note about Exim versions 4.44 and 4.50
3569 ----------------------------------------
3571 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3572 changes of various kinds. As a consequence, a big documentation update was
3573 needed. This delayed the release for rather longer than seemed good, especially
3574 in the light of a couple of (minor) security issues. Therefore, the changes
3575 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3576 release. So 4.44 and 4.50 are in effect two different branches that both start
3579 I have left the 4.50 change log unchanged; it contains all the changes since
3580 4.43. The change log for 4.44 is below; many of its items are identical to
3581 those for 4.50. This seems to be the most sensible way to preserve the
3582 historical information.
3588 1. Minor wording change to the doc/README.SIEVE file.
3590 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3591 computation of the current number of files was incorrect.
3593 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3594 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3597 4. Give more explanation in the error message when the command for a transport
3598 filter fails to execute.
3600 5. There are several places where Exim runs a non-Exim command in a
3601 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3602 was being done only for the command run by the queryprogram router. It is
3603 now done for all such subprocesses. The other cases are: ${run, transport
3604 filters, and the commands run by the lmtp and pipe transports.
3606 6. Added CONFIGURE_GROUP build-time option.
3608 7. Some older OS have a limit of 256 on the maximum number of file
3609 descriptors. Exim was using setrlimit() to set 1000 as a large value
3610 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3611 systems. I've change it so that if it can't get 1000, it tries for 256.
3613 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3614 was an oversight, and furthermore, ever since the addition of extra
3615 controls (e.g. 4.43/32), the checks on when to allow different forms of
3616 "control" were broken. There should now be diagnostics for all cases when a
3617 control that does not make sense is encountered.
3619 9. Added the /retain_sender option to "control=submission".
3621 10. $recipients is now available in the predata ACL (oversight).
3623 11. Tidy the search cache before the fork to do a delivery from a message
3624 received from the command line. Otherwise the child will trigger a lookup
3625 failure and thereby defer the delivery if it tries to use (for example) a
3626 cached ldap connection that the parent has called unbind on.
3628 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3629 of $address_data from the recipient verification was clobbered by the
3630 sender verification.
3632 13. The value of address_data from a sender verification is now available in
3633 $sender_address_data in subsequent conditions in the ACL statement.
3635 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3637 15. Added a new option "connect=<time>" to callout options, to set a different
3640 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3641 was its contents. (It was OK if the option was not defined at all.)
3643 17. A "Completed" log line is now written for messages that are removed from
3644 the spool by the -Mrm option.
3646 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3647 information about exactly what failed.
3649 19. Added -dd to debug only the daemon process.
3651 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3652 handles timeouts, both on the server side and network timeouts. Renamed the
3653 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3655 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3656 It is now set to "smtps".
3658 22. $host_address is now set to the target address during the checking of
3659 ignore_target_hosts.
3661 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3662 being passed; this would have caused $sender_host_name to have been used if
3663 matching the list had actually called for a host name (not very likely,
3664 since this list is usually IP addresses). A host name is now passed as
3667 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3668 code that specifies a non-restarting handler (typically sigaction() in
3669 modern systems) in an attempt to fix a rare and obscure crash bug.
3671 25. Narrowed the window for a race in the daemon that could cause it to ignore
3672 SIGCHLD signals. This is not a major problem, because they are used only to
3673 wake it up if nothing else does.
3675 26. A malformed maildirsize file could cause Exim to calculate negative values
3676 for the mailbox size or file count. Odd effects could occur as a result.
3677 The maildirsize information is now recalculated if the size or filecount
3680 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3681 support for a long time. Removed HAVE_SYS_VFS_H.
3683 28. Installed the latest version of exipick from John Jetmore.
3685 29. In an address list, if the pattern was not a regular expression, an empty
3686 subject address (from a bounce message) matched only if the pattern was an
3687 empty string. Non-empty patterns were not even tested. This was the wrong
3688 because it is perfectly reasonable to use an empty address as part of a
3689 database query. An empty address is now tested by patterns that are
3690 lookups. However, all the other forms of pattern expect the subject to
3691 contain a local part and a domain, and therefore, for them, an empty
3692 address still always fails if the pattern is not itself empty.
3694 30. Exim went into a mad DNS loop when attempting to do a callout where the
3695 host was specified on an smtp transport, and looking it up yielded more
3696 than one IP address.
3698 31. Re-factored the code for checking spool and log partition space into a
3699 function that finds that data and another that does the check. The former
3700 is then used to implement four new variables: $spool_space, $log_space,
3701 $spool_inodes, and $log_inodes.
3703 32. The RFC2047 encoding function was originally intended for short strings
3704 such as real names; it was not keeping to the 75-character limit for
3705 encoded words that the RFC imposes. It now respects the limit, and
3706 generates multiple encoded words if necessary. To be on the safe side, I
3707 have increased the buffer size for the ${rfc2047: expansion operator from
3710 33. It is now permitted to omit both strings after an "if" condition; if the
3711 condition is true, the result is "true". As before, when the second string
3712 is omitted, a false condition yields an empty string. This makes it less
3713 cumbersome to write custom ACL and router conditions.
3715 34. Failure to deliver a bounce message always caused it to be frozen, even if
3716 there was an errors_to setting on the router. The errors_to setting is now
3719 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3720 canonical form (fully expanded) before being placed in
3721 $sender_host_address.
3723 36. The table in the code that translates DNS record types into text (T_A to
3724 "A" for instance) was missing entries for NS and CNAME. It is just possible
3725 that this could have caused confusion if both these types were looked up
3726 for the same domain, because the text type is used as part of Exim's
3727 per-process caching. But the chance of anyone hitting this buglet seems
3730 37. The dnsdb lookup has been extended in a number of ways.
3732 (1) There is a new type, "zns", which walks up the domain tree until it
3733 finds some nameserver records. It should be used with care.
3735 (2) There is a new type, "mxh", which is like "mx" except that it returns
3736 just the host names, not the priorities.
3738 (3) It is now possible to give a list of domains (or IP addresses) to be
3739 looked up. The behaviour when one of the lookups defers can be
3740 controlled by a keyword.
3742 (4) It is now possible to specify the separator character for use when
3743 multiple records are returned.
3745 38. The dnslists ACL condition has been extended: it is now possible to supply
3746 a list of IP addresses and/or domains to be looked up in a particular DNS
3749 39. Added log_selector=+queue_time_overall.
3751 40. When running the queue in the test harness, wait just a tad after forking a
3752 delivery process, to get repeatability of debugging output.
3754 41. Include certificate and key file names in error message when GnuTLS fails
3755 to set them up, because the GnuTLS error message doesn't include the name
3756 of the failing file when there is a problem reading it.
3758 42. Allow both -bf and -bF in the same test run.
3760 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3762 44. The "Exiscan patch" is now merged into the mainline Exim source.
3764 45. Sometimes the final signoff response after QUIT could fail to get
3765 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3766 < 0 before doing a fflush(). This bug looks as though it goes back to the
3767 introduction of TLS in release 3.20, but "sometimes" must have been rare
3768 because the tests only now provoked it.
3770 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3771 (this can affect the format of dates).
3773 47. exim_tidydb, when checking for the continued existence of a message for
3774 which it has found a message-specific retry record, was not finding
3775 messages that were in split spool directories. Consequently, it was
3776 deleting retry records that should have stayed in existence.
3778 48. Steve fixed some bugs in eximstats.
3780 49. The SPA authentication driver was not abandoning authentication and moving
3781 on to the next authenticator when an expansion was forced to fail,
3782 contradicting the general specification for all authenticators. Instead it
3783 was generating a temporary error. It now behaves as specified.
3785 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3786 (the order specifies the preference for clients). The order is now AES256,
3787 AES128, 3DES, ARCFOUR128.
3789 51. Small patch to Sieve code - explicitly set From: when generating an
3792 52. Exim crashed if a remote delivery caused a very long error message to be
3793 recorded - for instance if somebody sent an entire SpamAssassin report back
3794 as a large number of 550 error lines. This bug was coincidentally fixed by
3795 increasing the size of one of Exim's internal buffers (big_buffer) that
3796 happened as part of the Exiscan merge. However, to be on the safe side, I
3797 have made the code more robust (and fixed the comments that describe what
3800 53. Now that there can be additional text after "Completed" in log lines (if
3801 the queue_time_overall log selector is set), a one-byte patch to exigrep
3802 was needed to allow it to recognize "Completed" as not the last thing in
3805 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3806 patch that reportedly fixes this has been added. I am not expert enough to
3807 create a test for it. This is what the patch creator wrote:
3809 "I found a little strange behaviour of ldap code when working with
3810 Windows 2003 AD Domain, where users was placed in more than one
3811 Organization Units. When I tried to give exim partial DN, the exit code
3812 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3813 But simultaneously result of request was absolutely normal ldap result,
3814 so I produce this patch..."
3816 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3817 REFERENCE, so I have modified the code to exclude the patch when that macro
3820 55. Some experimental protocols are using DNS PTR records for new purposes. The
3821 keys for these records are domain names, not reversed IP addresses. The
3822 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3823 leaves it alone. Component reversal etc. now happens only for IP addresses.
3826 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3828 57. Double the size of the debug message buffer (to 2048) so that more of very
3829 long debug lines gets shown.
3831 58. The exicyclog utility now does better if the number of log files to keep
3832 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3834 59. Two changes related to the smtp_active_hostname option:
3836 (1) $smtp_active_hostname is now available as a variable.
3837 (2) The default for smtp_banner uses $smtp_active_hostname instead
3838 of $primary_hostname.
3840 60. The host_aton() function is supposed to be passed a string that is known
3841 to be a valid IP address. However, in the case of IPv6 addresses, it was
3842 not checking this. This is a hostage to fortune. Exim now panics and dies
3843 if the condition is not met. A case was found where this could be provoked
3844 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3845 components; fortuitously, this particular loophole had already been fixed
3846 by change 4.50/55 above.
3848 If there are any other similar loopholes, the new check in host_aton()
3849 itself should stop them being exploited. The report I received stated that
3850 data on the command line could provoke the exploit when Exim was running as
3851 exim, but did not say which command line option was involved. All I could
3852 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3853 running as the user.
3856 61. There was a buffer overflow vulnerability in the SPA authentication code
3857 (which came originally from the Samba project). I have added a test to the
3858 spa_base64_to_bits() function which I hope fixes it.
3861 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3862 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3864 63. The daemon start-up calls getloadavg() while still root for those OS that
3865 need the first call to be done as root, but it missed one case: when
3866 deliver_queue_load_max is set with deliver_drop_privilege. This is
3867 necessary for the benefit of the queue runner, because there is no re-exec
3868 when deliver_drop_privilege is set.
3870 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3871 This has been fixed.
3873 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3874 was in use, was not putting the data itself into the right store pool;
3875 consequently, it could be overwritten for a subsequent message in the same
3876 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3879 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3881 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3882 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3883 the function were treating the return as a boolean value, which happened to
3884 work because 0=false and not-0=true, but is not correct code.
3886 68. The host_aton() function was not handling scoped IPv6 addresses (those
3887 with, for example, "%eth0" on the end) correctly.
3889 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3890 negated items (that is, ~something) in unsigned ints. Some compilers
3891 apparently mutter when there is no cast.
3893 70. If an address verification called from an ACL failed, and did not produce a
3894 user-specific message (i.e. there was only a "system" message), nothing was
3895 put in $acl_verify_message. In this situation, it now puts the system
3898 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3899 session; change 4.31/43 added the unwanted input to the log line - except
3900 that it did not do this in the start of session case. It now does.
3902 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3903 This is wrong and can cause the other end to generate a synchronization
3904 error if it is another Exim or anything else that does the synchronization
3905 check. A QUIT command is no longer sent after a timeout.
3907 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3908 during host lookups.
3910 74. The defer_ok option of callout verification was not working if it was used
3911 when verifying addresses in header lines, that is, for this case:
3913 verify = header_sender/callout=defer_ok
3915 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3916 those file descriptors could be used for SMTP connections. If anything
3917 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3918 could be sent to the SMTP client, causing chaos. The daemon now opens
3919 stdin, stdout, and stderr to /dev/null when it puts itself into the
3922 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3923 log by default. The user can override this with suitable Perl magic.
3925 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3926 the log message when discard triggers, was not working for the DATA ACL or
3927 for the non-SMTP ACL.
3929 78. Error message wording change in sieve.c.
3931 79. If smtp_accept_max_per_host was set, the number of connections could be
3932 restricted to fewer than expected, because the daemon was trying to set up
3933 a new connection before checking whether the processes handling previous
3934 connections had finished. The check for completed processes is now done
3935 earlier. On busy systems, this bug wouldn't be noticed because something
3936 else would have woken the daemon, and it would have reaped the completed
3939 80. If a message was submitted locally by a user whose login name contained one
3940 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3941 It caused a spool format error. I have fixed the spool reading code. A
3942 related problem was that the "from" clause in the Received: line became
3943 illegal because of the space(s). It is now covered by ${quote_local_part.
3945 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3948 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3950 83. Patch from Sieve maintainer to fix unterminated string problem in
3951 "vacation" handling.
3953 84. Some minor changes to the Linux configuration files to help with other
3954 OS variants using glibc.
3956 85. One more patch for Sieve to update vacation handling to latest spec.
3959 ----------------------------------------------------
3960 See the note above about the 4.44 and 4.50 releases.
3961 ----------------------------------------------------
3967 1. Change 4.43/35 introduced a bug that caused file counts to be
3968 incorrectly computed when quota_filecount was set in an appendfile
3971 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3972 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3975 3. Give more explanation in the error message when the command for a transport
3976 filter fails to execute.
3978 4. There are several places where Exim runs a non-Exim command in a
3979 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3980 was being done only for the command run by the queryprogram router. It is
3981 now done for all such subprocesses. The other cases are: ${run, transport
3982 filters, and the commands run by the lmtp and pipe transports.
3984 5. Some older OS have a limit of 256 on the maximum number of file
3985 descriptors. Exim was using setrlimit() to set 1000 as a large value
3986 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3987 systems. I've change it so that if it can't get 1000, it tries for 256.
3989 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3990 was an oversight, and furthermore, ever since the addition of extra
3991 controls (e.g. 4.43/32), the checks on when to allow different forms of
3992 "control" were broken. There should now be diagnostics for all cases when a
3993 control that does not make sense is encountered.
3995 7. $recipients is now available in the predata ACL (oversight).
3997 8. Tidy the search cache before the fork to do a delivery from a message
3998 received from the command line. Otherwise the child will trigger a lookup
3999 failure and thereby defer the delivery if it tries to use (for example) a
4000 cached ldap connection that the parent has called unbind on.
4002 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4003 of $address_data from the recipient verification was clobbered by the
4004 sender verification.
4006 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4007 was its contents. (It was OK if the option was not defined at all.)
4009 11. A "Completed" log line is now written for messages that are removed from
4010 the spool by the -Mrm option.
4012 12. $host_address is now set to the target address during the checking of
4013 ignore_target_hosts.
4015 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4016 being passed; this would have caused $sender_host_name to have been used if
4017 matching the list had actually called for a host name (not very likely,
4018 since this list is usually IP addresses). A host name is now passed as
4021 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4022 code that specifies a non-restarting handler (typically sigaction() in
4023 modern systems) in an attempt to fix a rare and obscure crash bug.
4025 15. Narrowed the window for a race in the daemon that could cause it to ignore
4026 SIGCHLD signals. This is not a major problem, because they are used only to
4027 wake it up if nothing else does.
4029 16. A malformed maildirsize file could cause Exim to calculate negative values
4030 for the mailbox size or file count. Odd effects could occur as a result.
4031 The maildirsize information is now recalculated if the size or filecount
4034 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4035 support for a long time. Removed HAVE_SYS_VFS_H.
4037 18. Updated exipick to current release from John Jetmore.
4039 19. Allow an empty sender to be matched against a lookup in an address list.
4040 Previously the only cases considered were a regular expression, or an
4043 20. Exim went into a mad DNS lookup loop when doing a callout where the
4044 host was specified on the transport, if the DNS lookup yielded more than
4047 21. The RFC2047 encoding function was originally intended for short strings
4048 such as real names; it was not keeping to the 75-character limit for
4049 encoded words that the RFC imposes. It now respects the limit, and
4050 generates multiple encoded words if necessary. To be on the safe side, I
4051 have increased the buffer size for the ${rfc2047: expansion operator from
4054 22. Failure to deliver a bounce message always caused it to be frozen, even if
4055 there was an errors_to setting on the router. The errors_to setting is now
4058 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4059 canonical form (fully expanded) before being placed in
4060 $sender_host_address.
4062 24. Updated eximstats to version 1.33
4064 25. Include certificate and key file names in error message when GnuTLS fails
4065 to set them up, because the GnuTLS error message doesn't include the name
4066 of the failing file when there is a problem reading it.
4068 26. Expand error message when OpenSSL has problems setting up cert/key files.
4071 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4072 (this can affect the format of dates).
4074 28. exim_tidydb, when checking for the continued existence of a message for
4075 which it has found a message-specific retry record, was not finding
4076 messages that were in split spool directories. Consequently, it was
4077 deleting retry records that should have stayed in existence.
4079 29. eximstats updated to version 1.35
4080 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4081 1.35 - bugfix such that pie charts by volume are generated correctly
4083 30. The SPA authentication driver was not abandoning authentication and moving
4084 on to the next authenticator when an expansion was forced to fail,
4085 contradicting the general specification for all authenticators. Instead it
4086 was generating a temporary error. It now behaves as specified.
4088 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4089 (the order specifies the preference for clients). The order is now AES256,
4090 AES128, 3DES, ARCFOUR128.
4092 31. Small patch to Sieve code - explicitly set From: when generating an
4095 32. Exim crashed if a remote delivery caused a very long error message to be
4096 recorded - for instance if somebody sent an entire SpamAssassin report back
4097 as a large number of 550 error lines. This bug was coincidentally fixed by
4098 increasing the size of one of Exim's internal buffers (big_buffer) that
4099 happened as part of the Exiscan merge. However, to be on the safe side, I
4100 have made the code more robust (and fixed the comments that describe what
4103 33. Some experimental protocols are using DNS PTR records for new purposes. The
4104 keys for these records are domain names, not reversed IP addresses. The
4105 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4106 leaves it alone. Component reversal etc. now happens only for IP addresses.
4109 34. The host_aton() function is supposed to be passed a string that is known
4110 to be a valid IP address. However, in the case of IPv6 addresses, it was
4111 not checking this. This is a hostage to fortune. Exim now panics and dies
4112 if the condition is not met. A case was found where this could be provoked
4113 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4114 components; fortuitously, this particular loophole had already been fixed
4115 by change 4.50/55 or 4.44/33 above.
4117 If there are any other similar loopholes, the new check in host_aton()
4118 itself should stop them being exploited. The report I received stated that
4119 data on the command line could provoke the exploit when Exim was running as
4120 exim, but did not say which command line option was involved. All I could
4121 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4122 running as the user.
4125 35. There was a buffer overflow vulnerability in the SPA authentication code
4126 (which came originally from the Samba project). I have added a test to the
4127 spa_base64_to_bits() function which I hope fixes it.
4130 36. The daemon start-up calls getloadavg() while still root for those OS that
4131 need the first call to be done as root, but it missed one case: when
4132 deliver_queue_load_max is set with deliver_drop_privilege. This is
4133 necessary for the benefit of the queue runner, because there is no re-exec
4134 when deliver_drop_privilege is set.
4136 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4137 was in use, was not putting the data itself into the right store pool;
4138 consequently, it could be overwritten for a subsequent message in the same
4139 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4142 38. Sometimes the final signoff response after QUIT could fail to get
4143 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4144 < 0 before doing a fflush(). This bug looks as though it goes back to the
4145 introduction of TLS in release 3.20, but "sometimes" must have been rare
4146 because the tests only now provoked it.
4152 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4153 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4154 Now it returns an integer. A number of calls were still expecting a T/F
4155 return. Fortuitously, in all cases, the tests worked in OK situations,
4156 which is the norm. However, things would have gone wrong on any write
4157 failures on the smtp file descriptor. This function is used when sending
4158 messages over SMTP and also when doing verify callouts.
4160 2. When Exim is called to do synchronous delivery of a locally submitted
4161 message (the -odf or -odi options), it no longer closes stderr before doing
4164 3. Implemented the mua_wrapper option.
4166 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4168 5. Implemented the functions header_remove(), header_testname(),
4169 header_add_at_position(), and receive_remove_recipient(), and exported them
4172 6. If an ACL "warn" statement specified the addition of headers, Exim already
4173 inserted X-ACL-Warn: at the start if there was no header name. However, it
4174 was not making this test for the second and subsequent header lines if
4175 there were newlines in the string. This meant that an invalid header could
4176 be inserted if Exim was badly configured.
4178 7. Allow an ACL "warn" statement to add header lines at the start or after all
4179 the Received: headers, as well as at the end.
4181 8. Added the rcpt_4xx retry error code.
4183 9. Added postmaster_mailfrom=xxx to callout verification option.
4185 10. Added mailfrom=xxxx to the callout verification option, for verify=
4188 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4189 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4191 12. Inserted some casts to stop certain compilers warning when using pointer
4192 differences as field lengths or precisions in printf-type calls (mostly
4193 affecting debugging statements).
4195 13. Added optional readline() support for -be (dynamically loaded).
4197 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4198 same clock tick as a message's arrival, so that its received time was the
4199 same as the "first fail" time on the retry record, and that message
4200 remained on the queue past the ultimate address timeout, every queue runner
4201 would try a delivery (because it was past the ultimate address timeout) but
4202 after another failure, the ultimate address timeout, which should have then
4203 bounced the address, did not kick in. This was a "< instead of <=" error;
4204 in most cases the first failure would have been in the next clock tick
4205 after the received time, and all would be well.
4207 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4208 being recognized when the domain list was tested by the match_domain
4209 condition in an expansion string.
4211 16. Added the ${str2b64: operator.
4213 17. Exim was always calling setrlimit() to set a large limit for the number of
4214 processes, without checking whether the existing limit was already
4215 adequate. (It did check for the limit on file descriptors.) Furthermore,
4216 errors from getrlimit() and setrlimit() were being ignored. Now they are
4217 logged to the main and panic logs, but Exim does carry on, to try to do its
4218 job under whatever limits there are.
4220 18. Imported PCRE 5.0.
4222 19. Trivial typo in log message " temporarily refused connection" (the leading
4225 20. If the log selector return_path_on_delivery was set and an address was
4226 redirected to /dev/null, the delivery process crashed because it assumed
4227 that a return path would always be set for a "successful" delivery. In this
4228 case, the whole delivery is bypassed as an optimization, and therefore no
4231 21. Internal re-arrangement: the function for sending a challenge and reading
4232 a response while authentication was assuming a zero-terminated challenge
4233 string. It's now changed to take a pointer and a length, to allow for
4234 binary data in such strings.
4236 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4238 23. Exim was not respecting finduser_retries when seeking the login of the
4239 uid under which it was called; it was always trying 10 times. (The default
4240 setting of finduser_retries is zero.) Also, it was sleeping after the final
4241 failure, which is pointless.
4243 24. Implemented tls_on_connect_ports.
4245 25. Implemented acl_smtp_predata.
4247 26. If the domain in control=submission is set empty, Exim assumes that the
4248 authenticated id is a complete email address when it generates From: or
4249 Sender: header lines.
4251 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4252 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4253 chown and chgrp in /bin and hostname in /usr/bin.
4255 28. Exim was keeping the "process log" file open after each use, just as it
4256 does for the main log. This opens the possibility of it remaining open for
4257 long periods when the USR1 signal hits a daemon. Occasional processlog
4258 errors were reported, that could have been caused by this. Anyway, it seems
4259 much more sensible not to leave this file open at all, so that is what now
4262 29. The long-running daemon process does not normally write to the log once it
4263 has entered its main loop, and it closes the log before doing so. This is
4264 so that log files can straightforwardly be renamed and moved. However,
4265 there are a couple of unusual error situations where the daemon does write
4266 log entries, and I had neglected to close the log afterwards.
4268 30. The text of an SMTP error response that was received during a remote
4269 delivery was being truncated at 512 bytes. This is too short for some of
4270 the long messages that one sometimes sees. I've increased the limit to
4273 31. It is now possible to make retry rules that apply only when a message has a
4274 specific sender, in particular, an empty sender.
4276 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4277 it possible to be selective about when SMTP synchronization is enforced.
4279 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4281 32. Implemented hosts_connection_nolog.
4283 33. Added an ACL for QUIT.
4285 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4288 35. Added mailbox_size and mailbox_filecount to appendfile.
4290 36. Added control = no_multiline_responses to ACLs.
4292 37. There was a bug in the logic of the code that waits for the clock to tick
4293 in the case where the clock went backwards by a substantial amount such
4294 that the microsecond fraction of "now" was more than the microsecond
4295 fraction of "then" (but the whole seconds number was less).
4297 38. Added support for the libradius Radius client library this is found on
4298 FreeBSD (previously only the radiusclient library was supported).
4304 1. When certain lookups returned multiple values in the form name=value, the
4305 quoting of the values was not always being done properly. Specifically:
4306 (a) If the value started with a double quote, but contained no whitespace,
4308 (b) If the value contained whitespace other than a space character (i.e.
4309 tabs or newlines or carriage returns) it was not quoted.
4310 This fix has been applied to the mysql and pgsql lookups by writing a
4311 separate quoting function and calling it from the lookup code. The fix
4312 should probably also be applied to nisplus, ibase and oracle lookups, but
4313 since I cannot test any of those, I have not disturbed their existing code.
4315 2. A hit in the callout cache for a specific address caused a log line with no
4316 reason for rejecting RCPT. Now it says "Previous (cached) callout
4317 verification failure".
4319 3. There was an off-by-one bug in the queryprogram router. An over-long
4320 return line was truncated at 256 instead of 255 characters, thereby
4321 overflowing its buffer with the terminating zero. As well as fixing this, I
4322 have increased the buffer size to 1024 (and made a note to document this).
4324 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4325 when Exim is waiting for an SMTP response from a remote server, Exim
4326 restarts its select() call on the socket, thereby resetting its timeout.
4327 This is not a problem when such interrupts are rare. Somebody set up a cron
4328 job to run exiwhat every 2 minutes, which is less than the normal select()
4329 timeout (5 or 10 minutes). This meant that the select() timeout never
4330 kicked in because it was always reset. I have fixed this by comparing the
4331 time when an interrupt arrives with the time at the start of the first call
4332 to select(). If more time than the timeout has elapsed, the interrupt is
4333 treated as a timeout.
4335 5. Some internal re-factoring in preparation for the addition of Sieve
4336 extensions (by MH). In particular, the "personal" test is moved to a
4337 separate function, and given an option for scanning Cc: and Bcc: (which is
4338 not set for Exim filters).
4340 6. When Exim created an email address using the login of the caller as the
4341 local part (e.g. when creating a From: or Sender: header line), it was not
4342 quoting the local part when it contained special characters such as @.
4344 7. Installed new OpenBSD configuration files.
4346 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4347 try to make them clearer.
4349 9. Callout options, other than the timeout value, were being ignored when
4350 verifying sender addresses in header lines. For example, when using
4352 verify = header_sender/callout=no_cache
4354 the cache was (incorrectly) being used.
4356 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4357 only the Cygwin environment.
4359 11. When return_path_on_delivery was set as a log selector, if different remote
4360 addresses in the same message used different return paths and parallel
4361 remote delivery occurred, the wrong values would sometimes be logged.
4362 (Whenever a remote delivery process finished, the return path value from
4363 the most recently started remote delivery process was logged.)
4365 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4366 lines when AUTH and/or TLS are in use. This is the "received protocol"
4367 field. Exim used to use "asmtp" for authenticated SMTP, without any
4368 indication (in the protocol name) for TLS use. Now it follows the RFC and
4369 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4370 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4371 names appear in log lines as well as in Received: header lines.
4373 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4374 extensions, and comparison tests, and to fix some bugs.
4376 14. Changes to the "personal" filter test:
4378 (1) The test was buggy in that it was just doing the equivalent of
4379 "contains" tests on header lines. For example, if a user's address was
4380 anne@some.where, the "personal" test would incorrectly be true for
4382 To: susanne@some.where
4384 This test is now done by extracting each address from the header in turn,
4385 and checking the entire address. Other tests that are part of "personal"
4386 are now done using regular expressions (for example, to check local parts
4387 of addresses in From: header lines).
4389 (2) The list of non-personal local parts in From: addresses has been
4390 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4391 taken from the Sieve specification recommendations.
4393 (3) If the message contains any header line starting with "List-" it is
4394 treated as non-personal.
4396 (4) The test for "circular" in the Subject: header line has been removed
4397 because it now seems ill-conceived.
4399 15. Minor typos in src/EDITME comments corrected.
4401 16. Installed latest exipick from John Jetmore.
4403 17. If headers_add on a router specified a text string that was too long for
4404 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4405 of string_sprintf() is now avoided.
4407 18. $message_body_size was not set (it was always zero) when running the DATA
4408 ACL and the local_scan() function.
4410 19. For the "mail" command in an Exim filter, no default was being set for
4411 the once_repeat time, causing a random time value to be used if "once" was
4412 specified. (If the value happened to be <= 0, no repeat happened.) The
4413 default is now 0s, meaning "never repeat". The "vacation" command was OK
4414 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4415 (I found it when inspecting the code).
4417 20. There is now an overall timeout for performing a callout verification. It
4418 defaults to 4 times the callout timeout, which applies to individual SMTP
4419 commands during the callout. The overall timeout applies when there is more
4420 than one host that can be tried. The timeout is checked before trying the
4421 next host. This prevents very long delays if there are a large number of
4422 hosts and all are timing out (e.g. when the network connections are timing
4423 out). The value of the overall timeout can be changed by specifying an
4424 additional sub-option for "callout", called "maxwait". For example:
4426 verify = sender/callout=5s,maxwait=20s
4428 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4429 to the end before writing, but this should make it even safer).
4431 22. Exim was forgetting that it had advertised PIPELINING for the second and
4432 subsequent messages on an SMTP connection. It was also not resetting its
4433 memory on STARTTLS and an internal HELO.
4435 23. When Exim logs an SMTP synchronization error within a session, it now
4436 records whether PIPELINING has been advertised or not.
4438 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4439 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4440 rather than long int.
4442 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4444 26. Added the never_mail option to autoreply.
4450 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4451 crash if the getsockname() call failed; this can happen if a connection is
4452 closed very soon after it is established. The problem was simply in the
4453 order in which certain operations were done, causing Exim to try to write
4454 to the SMTP stream before it had set up the file descriptor. The bug has
4455 been fixed by making things happen in the correct order.
4461 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4462 before the connection was closed, thus losing the rejection response.
4464 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4465 some early Solaris releases, but causes trouble in current releases where
4466 socklen_t is defined.
4468 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4471 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4474 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4475 that releasing the top of it at the end releases what was used for sub-
4476 expansions (unless the block got too big). However, discard this block if
4477 the first thing is a variable or header, so that we can use its block when
4478 it is dynamic (useful for very large $message_headers, for example).
4480 6. Lookups now cache *every* query, not just the most recent. A new, separate
4481 store pool is used for this. It can be recovered when all lookup caches are
4482 flushed. Lookups now release memory at the end of their result strings.
4483 This has involved some general refactoring of the lookup sources.
4485 7. Some code has been added to the store_xxx() functions to reduce the amount
4486 of flapping under certain conditions.
4488 8. log_incoming_interface used to affect only the <= reception log lines. Now
4489 it causes the local interface and port to be added to several more SMTP log
4490 lines, for example "SMTP connection from", and rejection lines.
4492 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4494 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4496 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4497 could be overwritten at the end of the current message (or the start of a
4498 new message if it was set in a HELO ACL). The value is now preserved for
4499 the duration of the SMTP connection.
4501 12. If a transport had a headers_rewrite setting, and a matching header line
4502 contained an unqualified address, that address was qualified, even if it
4503 did not match any rewriting rules. The underlying bug was that the values
4504 of the flags that permit the existence of unqualified sender and recipient
4505 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4506 non-local messages, and by -bnq for local messages) were not being
4507 preserved with the message after it was received.
4509 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4510 "next input=" as part of the text comprising the host identity instead of
4511 the correct text. The code was using the same buffer for two different
4512 strings. However, depending on which order the printing function evaluated
4513 its arguments, the bug did not always show up. Under Linux, for example, my
4514 test suite worked just fine.
4516 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4517 doesn't work with some older versions of Perl. It has been changed to "my",
4518 which in any case is probably the better facility to use.
4520 15. A really picky compiler found some instances of statements for creating
4521 error messages that either had too many or two few arguments for the format
4524 16. The size of the buffer for calls to the DNS resolver has been increased
4525 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4526 for addresses that have a lot of PTR records. This alleviates a problem; it
4527 does not fully solve it.
4529 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4530 buffer now truncates the list and logs the incident, which is the same
4531 action as happens when Exim is looking up a host name and its aliases.
4532 Previously in this situation something unpredictable would happen;
4533 sometimes it was "internal error: store_reset failed".
4535 18. If a server dropped the connection unexpectedly when an Exim client was
4536 using GnuTLS and trying to read a response, the client delivery process
4537 crashed while trying to generate an error log message.
4539 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4540 string, for example:
4542 warn message = H1: something\nH2: something
4544 the text was added as a single header line from Exim's point of view
4545 though it ended up OK in the delivered message. However, searching for the
4546 second and subsequent header lines using $h_h2: did not work. This has been
4547 fixed. Similarly, if a system filter added multiple headers in this way,
4548 the routers could not see them.
4550 20. Expanded the error message when iplsearch is called with an invalid key to
4551 suggest using net-iplsearch in a host list.
4553 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4554 ACLs are no longer actually imposed (and a message to that effect is
4557 22. If a "gecos" field in a passwd entry contained escaped characters, in
4558 particular, if it contained a \" sequence, Exim got it wrong when building
4559 a From: or a Sender: header from that name. A second bug also caused
4560 incorrect handling when an unquoted " was present following a character
4561 that needed quoting.
4563 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4564 was not being matched caselessly.
4566 24. Arranged for all hyphens in the exim.8 source to be escaped with
4569 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4570 properly. Recipient callout cache records were still being keyed to include
4571 the sender, even when use_sender was set false. This led to far more
4572 callouts that were necessary. The sender is no longer included in the key
4573 when use_sender is false.
4575 26. Added "control = submission" modifier to ACLs.
4577 27. Added the ${base62d: operator to decode base 62 numbers.
4579 28. dnsdb lookups can now access SRV records.
4581 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4582 the configuration file.
4584 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4585 (-v) mode. This makes the output for a verbose queue run more intelligible.
4587 31. Added a use_postmaster feature to recipient callouts.
4589 32. Added the $body_zerocount variable, containing the number of binary zero
4590 bytes in the message body.
4592 33. The time of last modification of the "new" subdirectory is now used as the
4593 "mailbox time last read" when there is a quota error for a maildir
4596 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4598 35. Added +ignore_unknown as a special item in host lists.
4600 36. Code for decoding IPv6 addresses in host lists is now included, even if
4601 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4602 address was recognized as an IP address, but was then not correctly decoded
4603 into binary, causing unexpected and incorrect effects when compared with
4610 1. Very minor rewording of debugging text in manualroute to say "list of
4611 hosts" instead of "hostlist".
4613 2. If verify=header_syntax was set, and a header line with an unqualified
4614 address (no domain) and a large number of spaces between the end of the
4615 name and the colon was received, the reception process suffered a buffer
4616 overflow, and (when I tested it) crashed. This was caused by some obsolete
4617 code that should have been removed. The fix is to remove it!
4619 3. When running in the test harness, delay a bit after writing a bounce
4620 message to get a bit more predictability in the log output.
4622 4. Added a call to search_tidyup() just before forking a reception process. In
4623 theory, someone could use a lookup in the expansion of smtp_accept_max_
4624 per_host which, without the tidyup, could leave open a database connection.
4626 5. Added the variables $recipient_data and $sender_data which get set from a
4627 lookup success in an ACL "recipients" or "senders" condition, or a router
4628 "senders" option, similar to $domain_data and $local_part_data.
4630 6. Moved the writing of debug_print from before to after the "senders" test
4633 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4634 problems for message scanning, either using a data ACL, or using
4635 local_scan() because the Received: header was not generated till after they
4636 were called (in order to set the time as the time of reception completion).
4637 I have revised the way this works. The header is now generated after the
4638 body is received, but before the ACL or local_scan() are called. After they
4639 are run, the timestamp in the header is updated.
4645 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4646 before starting a queue runner without re-exec. This happened only when
4647 deliver_drop_privilege was set or when the Exim user was set to root. The
4648 effect of the bug was that timeouts during subsequent deliveries caused
4649 crashes instead of being properly handled. The handler is now left at its
4650 default (and expected) setting.
4652 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4653 message, again when deliver_drop_privilege is set or Exim is run as root.
4654 The bug described in (1) was not present in this case, but the tidying up
4655 of the other signals was missing. I have made the two cases consistent.
4657 3. The ignore_target_hosts setting on a manualroute router was being ignored
4658 for hosts that were looked up using the /MX notation.
4660 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4663 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4664 operated on the sender address. After changing the $sender_address to <>
4665 for the sender address verify, Exim was re-instated it as the original
4666 (before rewriting) address, but remembering that it had rewritten it, so it
4667 wasn't rewriting it again. This bug also had the effect of breaking the
4668 sender address verification caching when the sender address was rewritten.
4670 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4671 This has been changed so that if the ip literal address matches
4672 ignore_target_hosts, the router declines.
4674 7. Added expansion conditions match_domain, match_address, and match_local_
4675 part (NOT match_host).
4677 8. The placeholder for the Received: header didn't have a length field set.
4679 9. Added code to Exim itself and to exim_lock to test for a specific race
4680 condition that could lead to file corruption when using MBX delivery. The
4681 issue is with the lockfile that is created in /tmp. If this file is removed
4682 after a process has opened it but before that process has acquired a lock,
4683 there is the potential for a second process to recreate the file and also
4684 acquire a lock. This could lead to two Exim processes writing to the file
4685 at the same time. The added code performs the same test as UW imapd; it
4686 checks after acquiring the lock that its file descriptor still refers to
4687 the same named file.
4689 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4690 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4693 11. Added the smtp_active_hostname option. If used, this will typically be made
4694 to depend on the incoming interface address. Because $interface_address is
4695 not set up until the daemon has forked a reception process, error responses
4696 that can happen earlier (such as "too many connections") no longer contain
4699 12. If an expansion in a condition on a "warn" statement fails because a lookup
4700 defers, the "warn" statement is abandoned, and the next ACL statement is
4701 processed. Previously this caused the whole ACL to be aborted.
4703 13. Added the iplsearch lookup type.
4705 14. Added ident_timeout as a log selector.
4707 15. Added tls_certificate_verified as a log selector.
4709 16. Added a global option tls_require_ciphers (compare the smtp transport
4710 option of the same name). This controls incoming TLS connections.
4712 17. I finally figured out how to make tls_require_ciphers do a similar thing
4713 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4714 before starting the TLS session.
4716 18. Tabs are now shown as \t in -bP output.
4718 19. If the log selector return_path_on_delivery was set, Exim crashed when
4719 bouncing a message because it had too many Received: header lines.
4721 20. If two routers both had headers_remove settings, and the first one included
4722 a superfluous trailing colon, the final name in the first list and the
4723 first name in the second list were incorrectly joined into one item (with a
4724 colon in the middle).
4730 1. Added -C and -D options to the exinext utility, mainly to make it easier
4731 to include in the automated testing, but these could be helpful when
4732 multiple configurations are in use.
4734 2. The exinext utility was not formatting the output nicely when there was
4735 an alternate port involved in the retry record key, nor when there was a
4736 message id as well (for retries that were specific to a specific message
4737 and a specific host). It was also confused by IPv6 addresses, because of
4738 the additional colons they contain. I have fixed the IPv4 problem, and
4739 patched it up to do a reasonable job for IPv6.
4741 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4742 delivery, the log line now contains "pipelined" if PIPELINING was used.
4744 4. An SMTP transport process used to panic and die if the bind() call to set
4745 an explicit outgoing interface failed. This has been changed; it is now
4746 treated in the same way as a connect() failure.
4748 5. A reference to $sender_host_name in the part of a conditional expansion
4749 that was being skipped was still causing a DNS lookup. This no longer
4752 6. The def: expansion condition was not recognizing references to header lines
4753 that used bh_ and bheader_.
4755 7. Added the _cache feature to named lists.
4757 8. The code for checking quota_filecount in the appendfile transport was
4758 allowing one more file than it should have been.
4760 9. For compatibility with Sendmail, the command line option
4768 and sets the incoming protocol and host name (for trusted callers). The
4769 host name and its colon can be omitted when only the protocol is to be set.
4770 Note the Exim already has two private options, -pd and -ps, that refer to
4771 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4772 "s", but I don't think that's a major issue.
4774 10. A number of refactoring changes to the code, none of which should affect
4777 (a) The number of logging options was getting close to filling up the
4778 32-bit word that was used as a bit map. I have split them into two classes:
4779 those that are passed in the argument to log_write(), and those that are
4780 only ever tested independently outside of that function. These are now in
4781 separate 32-bit words, so there is plenty of room for expansion again.
4782 There is no change in the user interface or the logging behaviour.
4784 (b) When building, for example, log lines, the code previously used a
4785 macro that called string_cat() twice, in order to add two strings. This is
4786 not really sufficiently general. Furthermore, there was one instance where
4787 it was actually wrong because one of the argument was used twice, and in
4788 one call a function was used. (As it happened, calling the function twice
4789 did not affect the overall behaviour.) The macro has been replaced by a
4790 function that can join an arbitrary number of extra strings onto a growing
4793 (c) The code for expansion conditions now uses a table and a binary chop
4794 instead of a serial search (which was left over from when there were very
4795 few conditions). Also, it now recognizes conditions like "pam" even when
4796 the relevant support is not compiled in: a suitably worded error message is
4797 given if an attempt is made to use such a condition.
4799 11. Added ${time_interval:xxxxx}.
4801 12. A bug was causing one of the ddress fields not to be passed back correctly
4802 from remote delivery subprocesses. The field in question was not being
4803 subsequently used, so this caused to problems in practice.
4805 13. Added new log selectors queue_time and deliver_time.
4807 14. Might have fixed a bug in maildirsizefile handling that threw up
4808 "unexpected character" debug warnings, and recalculated the data
4809 unnecessarily. In any case, I expanded the warning message to give more
4812 15. Added the message "Restricted characters in address" to the statements in
4813 the default ACL that block characters like @ and % in local parts.
4815 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4816 Three changes have been made:
4818 (a) There was a serious bug; a negative response to MAIL caused the whole
4819 recipient domain to be cached as invalid, thereby blocking all messages
4820 to all local parts at the same domain, from all senders. This bug has
4821 been fixed. The domain is no longer cached after a negative response to
4822 MAIL if the sender used is not empty.
4824 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4827 (c) A new callout option, "use_sender" has been added for people who want
4828 the modified behaviour.
4834 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4837 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4838 indeed breaks things for older releases.
4840 3. Added additional logging to the case where there is a problem reading data
4841 from a filter that is running in a subprocess using a pipe, in order to
4842 try to track down a specific problem.
4844 4. Testing facility fudge: when running in the test harness and attempting
4845 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4846 getting "No route to host". Convert this to a timeout.
4848 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4851 6. Some OS don't have socklen_t but use size_t instead. This affects the
4852 fifth argument of getsockopt() amongst other things. This is now
4853 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4854 can be set for individual OS. I have set it for SunOS5, OSF1, and
4855 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4856 some earlier ones do not.
4858 7. Change 4.30/15 was not doing the test caselessly.
4860 8. The standard form for an IPv6 address literal was being rejected by address
4861 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4862 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4863 this, as well as the form without the "IPv6" on the front (but only when
4864 address literals are enabled, of course).
4866 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4868 10. Exim crashed if a message with an empty sender address specified by -f
4869 encountered a router with an errors_to setting. This could be provoked only
4870 by a command such as
4874 where an empty string was supplied; "<>" did not hit this bug.
4876 11. Installed PCRE release 4.5.
4878 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4879 remained set. It is now erased.
4881 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4882 times from message ids (which are base 36 rather than the normal 62).
4884 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4885 were being counted as actual protocol errors, and logged if the log
4886 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4887 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4888 and DATA following a set of rejected RCPTs do not count as protocol errors.
4889 In other words, Exim assumes they were pipelined, though this may not
4890 actually be the case. Of course, in all cases the client gets an
4891 appropriate error code.
4893 15. If a lookup fails in an ACL condition, a message about the failure may
4894 be available; it is used if testing the ACL cannot continue, because most
4895 such messages specify what the cause of the deferral is. However, some
4896 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4897 that caused an old message to be retained and used if a later statement
4898 caused a defer, replacing the real cause of the deferral.
4900 16. If an IP address had so many PTR records that the DNS lookup buffer
4901 was not large enough to hold them, Exim could crash while trying to process
4902 the truncated data. It now detects and logs this case.
4904 17. Further to 4.21/58, another change has been made: if (and only if) the
4905 first line of a message (the first header line) ends with CRLF, a bare LF
4906 in a subsequent header line has a space inserted after it, so as not to
4907 terminate the header.
4909 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4910 unnecessarily, used atoi() instead of strtol(), and didn't check the
4911 termination when getting file sizes from file names by regex.
4913 19. Completely re-implemented the support for maildirsize files, in the light
4914 of a number of problems with the previous contributed implementation
4915 (4.30/29). In particular:
4917 . If the quota is zero, the maildirsize file is maintained, but no quota is
4920 . If the maildir directory does not exist, it is created before any attempt
4921 to write a maildirsize file.
4923 . The quota value in the file is just a cache; if the quota is changed in
4924 the transport, the new value overrides.
4926 . A regular expression is available for excluding directories from the
4929 20. The autoreply transport checks the characters in options that define the
4930 message's headers; it allows continued headers, but it was checking with
4931 isspace() after an embedded newline instead of explicitly looking for a
4934 21. If all the "regular" hosts to which an address was routed had passed their
4935 expiry times, and had not reached their retry times, the address was
4936 bounced, even if fallback hosts were defined. Now Exim should go on to try
4939 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4940 equivalent code in the SMTP transport. Some hosts send humungous responses
4941 to HELO/EHLO, more than 1024 it seems.
4943 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4944 gives compiler warnings in some environments. I've now done it "properly",
4947 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4948 (because of problems with the built-in one) was declared to return uschar *
4949 instead of char *, causing compiler failure.
4951 25. Fixed a file descriptor leak when processing alias/forward files.
4953 26. Fixed a minor format string issue in dbfn.c.
4955 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4957 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4958 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4961 29. When a sender address is verified, it is cached, to save repeating the test
4962 when there is more than one recipient in a message. However, when the
4963 verification involves a callout, it is possible for different callout
4964 options to be set for different recipients. It is too complicated to keep
4965 track of this in the cache, so now Exim always runs a verification when a
4966 callout is required, relying on the callout cache for the optimization.
4967 The overhead is duplication of the address routing, but this should not be
4970 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4971 to be verified with callout=postmaster, and the main callout worked but the
4972 postmaster check failed, the verification correctly failed. However, if a
4973 subsequent RCPT command asked for sender verification *without* the
4974 postmaster check, incorrect caching caused this verification also to fail,
4977 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4978 it was not caching the DNS options (qualify_single, search_parents) that
4979 were used when the lookup failed. A subsequent lookup with different
4980 options therefore always gave the same answer, though there were cases
4981 where it should not have. (Example: a "domains = !$mx_any" option on a
4982 dnslookup router: the "domains" option is always processed without any
4983 widening, but the router might have qualify_single set.) Now Exim uses the
4984 cached value only when the same options are set.
4986 32. Added John Jetmore's "exipick" utility to the distribution.
4988 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4989 than a timeout (e.g. a certificate is required, and is not provided), an
4990 Exim server now closes the connection immediately. Previously it waited for
4991 the client to close - but if the client is SSL, it seems that they each
4992 wait for each other, leading to a delay before one of them times out.
4994 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4995 maintained 0.8.x compatibility because I don't think many are using it, and
4996 it is clearly obsolete.
4998 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5001 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5002 client certificate was expired. A simple patch fixes this, though I don't
5003 understand the full logic of why the verify callback is called multiple
5006 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5007 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5008 which causes problems with some clients (such as the Certicom SSL Plus
5009 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5010 disables the coutermeasure allowing Eudora to connect."
5012 38. Exim was not checking that a write() to a log file succeeded. This could
5013 lead to Bad Things if a log got too big, in particular if it hit a file
5014 size limit. Exim now panics and dies if it cannot write to a log file, just
5015 as it does if it cannot open a log file.
5017 39. Modified OS/Makefile-Linux so that it now contains
5019 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5021 The two -D definitions ensure that Exim is compiled with large file
5022 support, which makes it possible to handle log files that are bigger than
5025 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5026 instance) a domain was checked against a named list that involved a lookup,
5027 causing $domain_data to be set, then another domain was checked against the
5028 same list, then the first domain was re-checked, the value of $domain_data
5029 after the final check could be wrong. In particular, if the second check
5030 failed, it could be set empty. This bug probably also applied to
5033 41. The strip_trailing_dot option was not being applied to the address given
5034 with the -f command-line option.
5036 42. The code for reading a message's header from the spool was incrementing
5037 $received_count, but never initializing it. This meant that the value was
5038 incorrect (doubled) while delivering a message in the same process in which
5039 it was received. In the most common configuration of Exim, this never
5040 happens - a fresh exec is done - but it can happen when
5041 deliver_drop_privilege is set.
5043 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5044 it now includes up to 150 characters of the unexpected data in the log
5047 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5048 and building data strings. The size of both of these buffers was 10 000
5049 bytes - far larger than anybody would *ever* want, thought I. Needless to
5050 say, somebody hit the limit. I have increased the maximum line length to
5051 20 000 and the maximum data length of concatenated lines to 100 000. I have
5052 also fixed two bugs, because there was no checking on these buffers. Tsk,
5053 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5054 buffer is too small.
5056 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5057 lsearch lookups. Now it does.
5059 46. When parsing a route_list item in a manualroute router, a fixed-length
5060 buffer was used for the list of hosts. I made this 1024 bytes long,
5061 thinking that nobody would ever have a list of hosts that long. Wrong.
5062 Somebody had a whole pile of complicated expansion conditions, and the
5063 string was silently truncated, leading to an expansion error. It turns out
5064 that it is easier to change to an unlimited length (owing to other changes
5065 that have happened since this code was originally written) than to build
5066 structure for giving a limitation error. The length of the item that
5067 expands into the list of hosts is now unlimited.
5069 47. The lsearch lookup could not handle data where the length of text line was
5070 more than 4095 characters. Such lines were truncated, leading to shortened
5071 data being returned. It should now handle lines of any length.
5073 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5074 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5077 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5078 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5079 no longer happens in comment lines. A list of replacements is now placed
5080 at the head of all of the source files, except those whose only change is
5081 to replace PERL_COMMAND in the very first #! line.
5083 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5084 messages on the queue, with a bottom-up merge sort, using code contributed
5085 by Michael Haardt. This should make operations like -bp somewhat faster on
5086 large queues. It won't affect queue runners, except when queue_run_in_order
5089 51. Installed eximstats 1.31 in the distribution.
5091 52. Added support for SRV lookups to the dnslookup router.
5093 53. If an ACL referred to $message_body or $message_body_end, the value was not
5094 reset for any messages that followed in the same SMTP session.
5096 54. The store-handling optimization for building very long strings was not
5097 differentiating between the different store pools. I don't think this
5098 actually made any difference in practice, but I've tidied it.
5100 55. While running the routers to verify a sender address, $sender_address
5101 was still set to the sender address. This is wrong, because when routing to
5102 send a bounce to the sender, it would be empty. Therefore, I have changed
5103 it so that, while verifying a sender address, $sender_address is set to <>.
5104 (There is no change to what happens when verifying a recipient address.)
5106 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5107 target A or AAAA records (if not already returned) without resetting the
5108 qualify_single or search_parents options of the DNS resolver. These are
5109 inappropriate in this case because the targets of MX and SRV records must
5110 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5111 target that, when qualified, matched something in the local domain. These
5112 two options are now turned off when doing these lookups.
5114 57. It seems that at least some releases of Reiserfs (which does not have the
5115 concept of a fixed number of inodes) returns zero and not -1 for the
5116 number of available inodes. This interacted badly with check_spool_inodes,
5117 which assumed that -1 was the "no such thing" setting. What I have done is
5118 to check that the total number of inodes is greater than zero before doing
5119 the test of how many are available.
5121 58. When a "warn" ACL statement has a log_message modifier, the message is
5122 remembered, and not repeated. This is to avoid a lot of repetition when a
5123 message has many recipients that cause the same warning to be written.
5124 Howewer, Exim was preserving the list of already written lines for an
5125 entire SMTP session, which doesn't seem right. The memory is now reset if a
5126 new message is started.
5128 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5129 debugging output unless log_rewrite was also set.
5131 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5132 of (int)(handle) when we know that handle contains (void *)(-1).
5134 61. The Exim daemon panic-logs an error return when it closes the incoming
5135 connection. However "connection reset by peer" seems to be common, and
5136 isn't really an error worthy of noting specially, so that particular error
5139 62. When Exim is trying to find all the local interfaces, it used to panic and
5140 die if the ioctl to get the interface flags failed. However, it seems that
5141 on at least one OS (Solaris 9) it is possible to have an interface that is
5142 included in the list of interfaces, but for which you get a failure error
5143 for this call. This happens when the interface is not "plumbed" into a
5144 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5145 failure of the "get flags" call assumes that the interface is down.
5147 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5148 makes life easier for people who are doing arithmetic on fields extracted
5149 from dates, where you often get leading zeros that should not be
5150 interpreted as octal.
5152 64. Added qualify_domain to the redirect router, to override the global
5155 65. If a pathologically long header line contained very many addresses (the
5156 report of this problem mentioned 10 000) and each of them was rewritten,
5157 Exim could use up a very large amount of memory. (It kept on making new
5158 copies of the header line as it rewrote, and never released the old ones.)
5159 At the expense of a bit more processing, the header rewriting function has
5160 been changed so that it no longer eats memory in this way.
5162 66. The generation of the Received: header has been moved from the time that a
5163 message starts to be received, to the time that it finishes. The timestamp
5164 in the Received: header should now be very close to that of the <= log
5165 line. There are two side-effects of this change:
5167 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5168 logged header lines no longer include the local Received: line, because
5169 it has not yet been created. The same applies to a copy of the message
5170 that is returned to a non-SMTP sender when a message is rejected.
5172 (b) When a filter file is tested using -bf, no additional Received: header
5173 is added to the test message. After some thought, I decided that this
5176 This change does not affect the value of $received_for. It is still set
5177 after address rewriting, but before local_scan() is called.
5179 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5181 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5182 gave an unhelpful panic error message, and a defer error. I have managed to
5183 change this behaviour so that it now rejects any supplied certificate,
5184 which seems right, as the list of acceptable certificates is empty.
5186 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5187 gave an unhelpful defer error. I have not managed to make this reject any
5188 supplied certificates, but the error message it gives is "no certificate
5189 supplied", which is not helpful.
5191 70. exigrep's output now also includes lines that are not associated with any
5192 message, but which match the given pattern. Implemented by a patch from
5193 Martin Sluka, which also tidied up the Perl a bit.
5195 71. Recipient callout verification, like sender verification, was using <> in
5196 the MAIL FROM command. This isn't really the right thing, since the actual
5197 sender may affect whether the remote host accepts the recipient or not. I
5198 have changed it to use the actual sender in the callout; this means that
5199 the cache record is now keyed on a recipient/sender pair, not just the
5200 recipient address. There doesn't seem to be a real danger of callout loops,
5201 since a callout by the remote host to check the sender would use <>.
5202 [SEE ABOVE: changed after hitting problems.]
5204 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5205 temporary errors. However, in the case of such a code being given after
5206 the end of a data transmission (i.e. after ".") Exim was failing to write
5207 a retry record for the message. (Yes, there was some broken host that was
5208 actually sending 8xx at this point.)
5210 73. An unknown lookup type in a host list could cause Exim to panic-die when
5211 the list was checked. (An example that provoked this was putting <; in the
5212 middle of a list instead of at the start.) If this happened during a DATA
5213 ACL check, a -D file could be left lying around. This kind of configuration
5214 error no longer causes Exim to die; instead it causes a defer errror. The
5215 incident is still logged to the main and panic logs.
5217 74. Buglet left over from Exim 3 conversion. The message "too many messages
5218 in one connection" was written to the rejectlog but not the mainlog, except
5219 when address rewriting (yes!) was being logged.
5221 75. Added write_rejectlog option.
5223 76. When a system filter was run not as root (that is, when system_filter_user
5224 was set), the values of the $n variables were not being returned to the
5225 main process; thus, they were not subsequently available in the $sn
5228 77. Added +return_path_on_delivery log selector.
5230 78. A connection timeout was being treated differently from recipients deferred
5231 when testing hosts_max_try with a message that was older than the host's
5232 retry timeout. (The host should not be counted, thus allowing all hosts to
5233 be tried at least once before bouncing.) This may have been the cause of an
5234 occasionally reported bug whereby a message would remain on the queue
5235 longer than the retry timeout, but would be bounced if a delivery was
5236 forced. I say "may" because I never totally pinned down the problem;
5237 setting up timeout/retry tests is difficult. See also the next item.
5239 79. The ultimate address timeout was not being applied to errors that involved
5240 a combination of host plus message (for example, a timeout on a MAIL
5241 command). When an address resolved to a number of possible hosts, and they
5242 were not all tried for each delivery (e.g. because of hosts_max_try), a
5243 message could remain on the queue longer than the retry timeout.
5245 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5248 81. Fixed an obscure SMTP outgoing bug which required at least the following
5249 conditions: (a) there was another message waiting for the same server;
5250 (b) the server returned 5xx to all RCPT commands in the first message so
5251 that the message was not completed; (c) the server dropped the connection
5252 or gave a negative response to the RSET that Exim sends to abort the
5253 transaction. The observed case was a dropped connection after DATA that had
5254 been sent in pipelining mode. That is, the server had advertised PIPELINING
5255 but was not implementing it correctly. The effect of the bug was incorrect
5256 behaviour, such as trying another host, and this could lead to a crash.
5262 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5263 and daemon.c were passed as pointers to ints; they should have been
5264 pointers to socklen_t variables (which are typically unsigned ints).
5266 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5269 3. Fixed a really odd bug that affected only the testing scheme; patching a
5270 certain fixed string in the binary changed the value of another string that
5271 happened to be identical to the end of the original first string.
5273 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5274 name", it returns that address as the IP address. On some operating
5275 systems (e.g. Solaris), it also passes back the IP address string as the
5276 "host name". However, on others (e.g. Linux), it passes back an empty
5277 string. Exim wasn't checking for this, and was changing the host name to an
5278 empty string, assuming it had been canonicized.
5280 5. Although rare, it is permitted to have more than one PTR record for a given
5281 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5282 all the names associated with an address, because they do in Solaris.
5283 However, it seems that they do not in Linux for data that comes from the
5284 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5285 I found this out when I moved to a new Linux workstation and tried to run
5286 the Exim test suite.
5288 To get round this problem I have changed the code so that it now does its
5289 own call to the DNS to look up PTR records when searching for a host name.
5290 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5291 addresses that are only in /etc/hosts are still found.
5293 This behaviour is, however, controlled by an option called host_lookup_
5294 order, which defaults to "bydns:byaddr". If people want to use the other
5295 order, or indeed, just use one or the other means of lookup, they can
5296 specify it in this variable.
5298 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5299 some operating systems, this comes back from gethostbyaddr() as an empty
5300 string, and this is what Exim used to test for. However, it seems that in
5301 other systems, "." is yielded. Exim now tests for this case too.
5303 7. The values of check_spool_space and check_log_space are now held internally
5304 as a number of kilobytes instead of an absolute number of bytes. If a
5305 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5306 kilobyte. This means that much larger values can be stored.
5308 8. Exim monitor: an attempt to get the action menu when not actually pointing
5309 at a message produces an empty menu entitled "No message selected". This
5310 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5311 no entries in it ("Shell widget menu has zero width and/or height"). So I
5312 have added a single, blank menu entry in this case.
5314 9. Added ${quote_local_part.
5316 10. MIME decoding is now applied to the contents of Subject: header lines when
5319 11. Now that a reference to $sender_host_address automatically causes a reverse
5320 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5321 host lookup before query-style lookups in lists that might use this
5322 variable. This has therefore been abolished, and the "net-" prefix is no
5323 longer necessary for query-style lookups.
5325 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5326 have been a typo for LFLAGS, so it has been changed.
5328 13. The install script calls Exim with "-C /dev/null" in order to find the
5329 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5330 to be output. Howeve, since Exim outputs its version number before the
5331 error, it didn't break the script. It just looked ugly. I fixed this by
5332 always allowing "-C /dev/null" if the caller is root.
5334 14. Ignore overlarge ACL variable number when reading spool file - insurance
5335 against a later release with more variables having written the file.
5337 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5338 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5339 this, as well as the form without the "IPv6" on the front.
5341 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5342 OS/Makefile-Darwin file.
5344 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5345 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5347 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5348 variable. However, some OS use ints for inodes. I've added cast to long int
5349 to get rid of the compiler warning.
5351 19. I had forgotten to lock out "/../" in configuration file names when
5352 ALT_CONFIG_PREFIX was set.
5354 20. Routers used for verification do not need to specify transports. However,
5355 if such a router generated a host list, and callout was configured, Exim
5356 crashed, because it could not find a port number from the (non-existent)
5357 transport. It now assumes port 25 in this circumstance.
5359 21. Added the -t option to exigrep.
5361 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5362 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5363 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5364 them is set, LOOKUP_LSEARCH is forced.
5366 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5368 24. Added sender and host information to the "rejected by local_scan()" log
5369 line; previously there was no indication of these.
5371 25. Added .include_if_exists.
5373 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5374 Linux. It turns out that not all file systems support this. Apparently some
5375 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5376 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5377 not supported on the file descriptor, is now ignored when Exim is trying to
5378 sync a directory. This applies only to Linux.
5380 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5382 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5383 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5386 29. Added support for maildirsize files from supplied patch (modified a bit).
5388 30. The use of :fail: followed by an empty string could lead Exim to respond to
5389 sender verification failures with (e.g.):
5391 550 Verification failed for <xxx>
5392 550 Sender verify failed
5394 where the first response line was missing the '-' that indicates it is not
5395 the final line of the response.
5397 31. The loop for finding the name of the user that called Exim had a hardwired
5398 limit of 10; it now uses the value of finduser_retries, which is used for
5399 all other user lookups.
5401 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5404 33. Exim was neglecting to zero errno before one call of strtol() when
5405 expanding a string and expecting an integer value. On some systems this
5406 resulted in spurious "integer overflow" errors. Also, it was casting the
5407 result into an int without checking.
5409 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5410 did not work. The code looks as if it has *never* worked, though it appears
5411 to have been documented since at least releast 1.62. I have made it work.
5413 35. The "timeout_DNS" error in retry rules, also documented since at least
5414 1.62, also never worked. As it isn't clear exactly what this means, and
5415 clearly it isn't a major issue, I have abolished the feature by treating it
5416 as "timeout", and writing a warning to the main and panic logs.
5418 36. The display of retry rules for -brt wasn't always showing the error code
5421 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5422 timeout_connect_A, timeout_connect_MX.
5424 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5425 to the empty sender.
5427 39. The daemon was not analysing the content of -oX till after it had closed
5428 stderr and disconnected from the controlling terminal. This meant that any
5429 syntax errors were only noted on the panic log, and the return code from
5430 the command was 0. By re-arranging the code a little, I've made the
5431 decoding happen first, so such errors now appear on stderr, and the return
5432 code is 1. However, the actual setting up of the sockets still happens in
5433 the disconnected process, so errors there are still only recorded on the
5436 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5437 connections (as happens on some IP stacks) was logged at start up time as
5438 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5439 it from "IPv6 and IPv4", which means that two separate sockets are being
5442 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5443 says whether AF_INET or AF_INET6 was passed as an argument.
5445 42. Exiwhat output was messed up when time zones were included in log
5448 43. Exiwhat now gives more information about the daemon's listening ports,
5449 and whether -tls-on-connect was used.
5451 44. The "port" option of the smtp transport is now expanded.
5453 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5454 silently ignored. Now an error message is written to the main and panic
5457 46. There's a new ACL modifier called "logwrite" which writes to a log file
5458 as soon as it is encountered.
5460 47. Added $local_user_uid and $local_user_gid at routing time.
5462 48. Exim crashed when trying to verify a sender address that was being
5465 49. Exim was recognizing only a space character after ".include". It now also
5466 recognizes a tab character.
5468 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5469 extracting the relevant information from the specification. The man page no
5470 longer contains scrambled data for the -d option, and I've added a section
5471 at the front about calling Exim under different names.
5473 51. Added "extra_headers" argument to the "mail" command in filter files.
5475 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5478 53. Installed eximstats 1.29.
5480 54. Added transport_filter_timeout as a generic transport option.
5482 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5483 Cc: header lines. This was required by RFC 822, but it not required by RFC
5486 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5487 incoming messages that did not have them. Now it does so only if the
5488 message originates locally, that is, if there is no associated remote host
5489 address. When Resent- header lines are present, this applies to the Resent-
5490 lines rather than the non-Resent- lines.
5492 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5493 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5495 58. Messages for configuration errors now include the name of the main
5496 configuration file - useful now that there may be more than one file in a
5497 list (.included file names were always shown).
5499 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5500 for those rare installations that do not start the daemon as root or run it
5501 setuid root. I've cut out the call to initgroups() if the daemon is not
5504 60. The Exim user and group can now be bound into the binary as text strings
5505 that are looked up at the start of Exim's processing.
5507 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5509 62. Added $mailstore_basename variable.
5511 63. Installed patch to sieve.c from Michael Haardt.
5513 64. When Exim failed to open the panic log after failing to open the main log,
5514 the original message it was trying to log was written to stderr and debug
5515 output, but if they were not available (the usual case in production), it
5516 was lost. Now it is written to syslog before the two lines that record the
5517 failures to open the logs.
5519 65. Users' Exim filters run in subprocesses under the user's uid. It is
5520 possible for a "deliver" command or an alias in a "personal" command to
5521 provoke an address rewrite. If logging of address rewriting is configured,
5522 this fails because the process is not running as root or exim. There may be
5523 a better way of dealing with this, but for the moment (because 4.30 needs
5524 to be released), I have disabled address rewrite logging when running a
5525 filter in a non-root, non-exim process.
5531 1. The buildconfig auxiliary program wasn't quoting the value set for
5532 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5533 not defined. This bug was masked in 4.22 by the effect that was fixed in
5536 2. Some messages that were rejected after a message id was allocated were
5537 shown as "incomplete" by exigrep. It no longer does this for messages that
5538 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5540 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5541 have allow_domain_literals set, the ID did not get logged in the <= line.
5542 Domain literals are now always recognized in Message-ID: header lines.
5544 4. The first argument for a ${extract expansion item is the key name or field
5545 number. Leading and trailing spaces in this item were not being ignored,
5546 causing some misleading effects.
5548 5. When deliver_drop_privilege was set, single queue runner processes started
5549 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5550 same command in the process it spins off) were not dropping privilege.
5552 6. When the daemon running as "exim" started a queue runner, it always
5553 re-executed Exim in the spun-off process. This is a waste of effort when
5554 deliver_drop_privilege is set. The new process now just calls the
5555 queue-runner function directly.
5561 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5564 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5565 ignored. Though the use of -oP was forcing the writing of a pid file, it
5566 was always written to the default place.
5568 3. If the message "no IP address found for host xxxx" is generated during
5569 incoming verification, it is now followed by identification of the incoming
5570 connection (so you can more easily find what provoked it).
5572 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5574 5. Added some features to "harden" Exim a bit more against certain attacks:
5576 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5577 be put in Local/Makefile. This is like the never_users runtime option,
5578 but it cannot be overridden. The default setting is "root".
5580 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5581 prefix string with which any file named in a -C command line option
5584 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5585 is retained for -C and -D only if the caller of Exim is root. Without
5586 it, the exim user may also use -C and -D and retain privilege.
5588 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5589 command line option is disabled.
5591 6. Macro names set by the -D option must start with an upper case letter, just
5592 like macro names defined in the configuration file.
5594 7. Added "dereference=" facility to LDAP.
5596 8. Two instances of the typo "uknown" in the source files are fixed.
5598 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5599 the Configure-Makefile script screwed up while processing it.
5601 10. Incorporated PCRE 4.4.
5603 11. The SMTP synchronization check was not operating right at the start of an
5604 SMTP session. For example, it could not catch a HELO sent before the client
5605 waited for the greeting. There is now a check for outstanding input at the
5606 point when the greeting is written. Because of the duplex, asynchronous
5607 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5608 way, but not yet received, when the check is performed.
5610 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5611 on TCP/IP sockets, because this apparently causes some broken clients to
5614 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5615 unchanged) from the Cygwin maintainer.
5617 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5618 was supported instead of testing for mbx. Effectively a typo.
5620 15. The spa authenticator server code was not checking that the input it
5621 received was valid base64.
5623 16. The debug output line for the "set" modifier in ACLs was not showing the
5624 name of the variable that was being set.
5626 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5628 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5629 lookup on its own. Something else was needed to trigger the lookup. For
5630 example, a match in host_lookup or the need for a host name in a host list.
5631 Now, if $sender_host_name is referenced and the host name has not yet been
5632 looked up, a lookup is performed. If the lookup fails, the variable remains
5633 empty, and $host_lookup_failed is set to "1".
5635 19. Added "eqi" as a case-independent comparison operator.
5637 20. The saslauthd authentication condition could segfault if neither service
5638 nor realm was specified.
5640 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5641 error message that was logged was misleading, and incoming SMTP
5642 connections were dropped. The message is now more accurate, and temporary
5643 errors are given to SMTP connections.
5645 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5646 (except RSET) with a 421 error, until QUIT is received. However, it was
5647 failing to send a response to QUIT.
5649 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5650 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5651 ACL is now run after the helo_try_verify_hosts code.
5653 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5654 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5655 case-independent, so other case variants are also recognized). Apparently
5656 some systems use these upper case variants.
5658 25. If more than two messages were waiting for the same host, and a transport
5659 filter was specified for the transport, Exim sent two messages over the
5660 same TCP/IP connection, and then failed with "socket operation on non-
5661 socket" when it tried to send the third.
5663 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5665 27. The extern definition of crypt16() in expand.c was not being excluded when
5666 the OS had its own crypt16() function.
5668 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5669 as a preferred synonym for return_size_limit, both as an option and as an
5672 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5674 30. Changed the default configuration ACL to relax the local part checking rule
5675 for addresses that are not in any local domains. For these addresses,
5676 slashes and pipe symbols are allowed within local parts, but the sequence
5677 /../ is explicitly forbidden.
5679 31. SPA server authentication was not clearing the challenge buffer before
5682 32. log_message in a "warn" ACL statement was writing to the reject log as
5683 well as to the main log, which contradicts the documentation and doesn't
5684 seem right (because no rejection is happening). So I have stopped it.
5686 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5687 However, I am unable to do any testing of this.
5689 34. Fixed an infelicity in the appendfile transport. When checking directories
5690 for a mailbox, to see if any needed to be created, it was accidentally
5691 using path names with one or more superfluous leading slashes; tracing
5692 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5694 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5695 contents are added to the log line that is written for every discarded
5696 recipient. (Previously a log_message setting was ignored.)
5698 36. The ${quote: operator now quotes the string if it is empty.
5700 37. The install script runs exim in order to find its version number. If for
5701 some reason other than non-existence or emptiness, which it checks, it
5702 could not run './exim', it was installing it with an empty version number,
5703 i.e. as "exim-". This error state is now caught, and the installation is
5706 38. An argument was missing from the function that creates an error message
5707 when Exim fails to connect to the socket for saslauthd authentication.
5708 This could cause Exim to crash, or give a corrupted message.
5710 39. Added isip, isip4, and isip6 to ${if conditions.
5712 40. The ACL variables $acl_xx are now saved with the message, and can be
5713 accessed later in routers, transports, and filters.
5715 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5716 strings in the file are not string-expanded.
5718 42. If a MAIL command specified a SIZE value that was too large to fit into an
5719 int variable, the check against message_size_limit failed. Such values are
5720 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5721 day this will have to be increased, but I don't think I want to be around
5722 when emails are that large.
5729 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5730 iconv() is not standard in FreeBSD.
5732 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5733 IPv6 enabled. The observed symptom was a segmentation fault on return from
5734 the function os_common_find_running_interfaces() in src/os.c.
5736 3. In the check_special_case() function in daemon.c I had used "errno" as an
5737 argument name, which causes warnings on some systems. This was basically a
5738 typo, since it was named "eno" in the comments!
5740 4. The code that waits for the clock to tick (at a resolution of some fraction
5741 of a second) so as to ensure message-id uniqueness was always waiting for
5742 at least one whole tick, when it could have waited for less. [This is
5743 almost certainly not relevant at current processor speeds, where it is
5744 unlikely to ever wait at all. But we try to future-proof.]
5746 5. The function that sleeps for a time interval that includes fractions of a
5747 second contained a race. It did not block SIGALRM between setting the
5748 timer, and suspending (a couple of lines later). If the interval was short
5749 and the sigsuspend() was delayed until after it had expired, the suspension
5750 never ended. On busy systems this could lead to processes getting stuck for
5753 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5754 process, before it forks any delivery processes. The open lookup caching
5755 mechanism meant that the open file or database connection was passed into
5756 the delivery process. The problem was that delivery processes always tidy
5757 up cached lookup data. This could cause a problem for the next delivery
5758 process started by the queue runner, because the external queue runner
5759 process does not know about the closure. So the next delivery process
5760 still has data in the lookup cache. In the case of a file lookup, there was
5761 no problem because closing a file descriptor in a subprocess doesn't affect
5762 the parent. However, if the lookup was caching a connection to a database,
5763 the connection was closed, and the second delivery process was likely to
5764 see errors such as "PGSQL: query failed: server closed the connection
5765 unexpectedly". The problem has been fixed by closing all cached lookups
5766 in a queue runner before running a delivery process.
5768 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5769 seem to have the "const" qualifier which it has on other OS. I've
5772 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5773 *of the same type* (client or server) with the same public name that an
5774 error should be diagnosed.
5776 9. When Exim looked up a host name for an IP address, but failed to find the
5777 original IP address when looking up the host name (a safety check), it
5778 output the message "<ip address> does not match any IP for NULL", which was
5779 confusing, to say the least. The bug was that the host name should have
5780 appeared instead of "NULL".
5782 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5783 user that is built into the binary, and the -C or -D options is used, root
5784 privilege is dropped before the configuration file is read. In addition,
5785 logging is switched to stderr instead of the normal log files. If the
5786 configuration then re-defines the Exim user, the unprivileged environment
5787 is probably not what is expected, so Exim logs a panic warning message (but
5790 However, if deliver_drop_privilege is set, the unprivileged state may well
5791 be exactly what is intended, so the warning has been cut out in that case,
5792 and Exim is allowed to try to write to its normal log files.
5798 1. smtp_return_error_details was not giving details for temporary sender
5799 or receiver verification errors.
5801 2. Diagnose a configuration error if two authenticators have the same public
5804 3. Exim used not to create the message log file for a message until the first
5805 delivery attempt. This could be confusing when incoming messages were held
5806 for policy or load reasons. The message log file is now created at the time
5807 the message is received, and an initial "Received" line is written to it.
5809 4. The automatically generated man page for command line options had a minor
5810 bug that caused no ill effects; however, a more serious problem was that
5811 the procedure for building the man page automatically didn't always
5812 operate. Consequently, release 4.20 contains an out-of-date version. This
5813 shouldn't happen again.
5815 5. When building Exim with embedded Perl support, the script that builds the
5816 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5817 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5819 6. The freeze_tell option was not being used for messages that were frozen on
5820 arrival, either by an ACL or by local_scan().
5822 7. Added the smtp_incomplete_transaction log selector.
5824 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5825 was accepting AUTH without a new EHLO.
5827 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5828 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5831 10. Logging of TCP/IP connections (when configured) now happens in the main
5832 daemon process instead of the child process, so that the TCP/IP connection
5833 count is more accurate (but it can never be perfect).
5835 11. The use of "drop" in a nested ACL was not being handled correctly in the
5836 outer ACL. Now, if condition failure induced by the nested "drop" causes
5837 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5838 or "require"), the connection is dropped.
5840 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5841 that yield "discard" can now be used with an "accept" or a "discard" verb,
5842 but an error is generated for any others (because I can't see a useful way
5843 to define what should happen).
5845 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5846 are now processed in the same way as lines in the Exim configuration file.
5847 In particular, continuation lines are supported.
5849 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5851 15. Added -ti meaning -t -i.
5853 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5854 domains, and warn by logging if others are found.
5856 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5857 returned from GIFCONF when Exim is trying to find the list of interfaces on
5858 a host. The code in os.c has been modified to copy each ifreq to an aligned
5859 structure in all cases.
5861 Also, in some cases, the returned ifreq's were being copied to a 'struct
5862 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5863 means the last couple of bytes of an IPv6 address could be chopped if the
5864 ifreq contained only a normal sockaddr (14 bytes storage).
5866 18. Named domain lists were not supported in the hosts_treat_as_local option.
5867 An entry such as +xxxx was not recognized, and was treated as a literal
5870 19. Ensure that header lines added by a DATA ACL are included in the reject log
5871 if the ACL subsequently rejects the message.
5873 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5874 MD5 (which is deprecated).
5876 21. When testing a filter file using -bf, Exim was writing a message when it
5877 took the sender from a "From " line in the message, but it was not doing so
5878 when it took $return_path from a Return-Path: header line. It now does.
5880 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5881 with a valid header line field name (a series of printing characters
5882 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5884 23. Changed "disc" in the source to "disk" to conform to the documentation and
5885 the book and for uniformity.
5887 24. Ignore Sendmail's -Ooption=value command line item.
5889 25. When execve() failed while trying to run a command in a pipe transport,
5890 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5891 could be confused with a return value of 69 from the command itself. This
5892 has been changed to 127, the value the shell returns if it is asked to run
5893 a non-existent command. The wording for the related log line suggests a
5894 non-existent command as the problem.
5896 26. If received_header_text expands to an empty string, do not add a Received:
5897 header line to the message. (Well, it adds a token one on the spool, but
5898 marks it "old" so that it doesn't get used or transmitted.)
5900 27. Installed eximstats 1.28 (addition of -nt option).
5902 28. There was no check for failure on the call to getsockname() in the daemon
5903 code. This can fail if there is a shortage of resources on the system, with
5904 ENOMEM, for example. A temporary error is now given on failure.
5906 29. Contrary to the C standard, it seems that in some environments, the
5907 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5908 program. Exim now does this explicitly; it affects the formatting of
5909 timestamps using strftime().
5911 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5912 complaints. I've now initialized all the variables, to avoid this.
5914 32. Header lines added by a system filter were not being "seen" during
5915 transport-time rewrites.
5917 33. The info_callback() function passed to OpenSSL is set up with type void
5918 (*)(SSL *, int, int), as described somewhere. However, when calling the
5919 function (actually a macro) that sets it up, the type void(*)() is
5920 expected. I've put in a cast to prevent warnings from picky compilers.
5922 34. If a DNS black list lookup found a CNAME record, but there were no A
5923 records associated with the domain it pointed at, Exim crashed.
5925 35. If a DNS black list lookup returned more than one A record, Exim ignored
5926 all but the first. It now scans all returned addresses if a particular IP
5927 value is being sought. In this situation, the contents of the
5928 $dnslist_value variable are a list of all the addresses, separated by a
5931 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5932 to accept a host name and all its aliases if the forward lookup for any of
5933 them yielded the IP address of the incoming connection. Now it accepts only
5934 those names whose forward lookup yields the correct IP address. Any other
5935 names are discarded. This closes a loophole whereby a rogue DNS
5936 administrator could create reverse DNS records to break through a
5937 wildcarded host restriction in an ACL.
5939 37. If a user filter or a system filter that ran in a subprocess used any of
5940 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5941 the wrong values were passed to the pipe command ($thisaddress had the
5942 value of $0, $0 had the value of $1, etc). This bug was introduced by
5943 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5945 38. Improved the line breaking for long SMTP error messages from ACLs.
5946 Previously, if there was no break point between 40 and 75 characters, Exim
5947 left the rest of the message alone. Two changes have been made: (a) I've
5948 reduced the minimum length to 35 characters; (b) if it can't find a break
5949 point between 35 and 75 characters, it looks ahead and uses the first one
5950 that it finds. This may give the occasional overlong line, but at least the
5951 remaining text gets split now.
5953 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5954 file descriptors might be low, and that setting 1000 would always raise it.
5955 It turns out that in some environments, the limit is already over 1000 and
5956 that lowering it causes trouble. So now Exim takes care not to decrease it.
5958 40. When delivering a message, the value of $return_path is set to $sender_
5959 address at the start of routing (routers may change the value). By an
5960 oversight, this default was not being set up when an address was tested by
5961 -bt or -bv, which affected the outcome if any router or filter referred to
5964 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5965 log only when "message" or "log_message" are set. However, if one of the
5966 conditions was an address verification, or a call to a nested ACL, the
5967 messages generated by the underlying test were being passed through. This
5968 no longer happens. The underlying message is available in $acl_verify_
5969 message for both "message" and "log_message" expansions, so it can be
5970 passed through if needed.
5972 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5973 new expansion $bh_ to give the encoded byte string without charset
5974 translation. Translation happens only if iconv() is available; HAVE_ICONV
5975 indicates this at build time. HEADERS_CHARSET gives the charset to
5976 translate to; headers_charset can change it in the configuration, and
5977 "headers charset" can change it in an individual filter file.
5979 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5980 that creates RFC 2047 encoded "words" labels them as that charset instead
5981 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5982 expansion operator; (ii) when Exim creates a From: line for a local
5983 message; (iii) when a header line is rewritten to include a "phrase" part.
5985 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5986 buggy, causing it to skip the first lines of messages whose message ID
5987 ended in 'D'. This would not have bitten before Exim release 4.14, because
5988 message IDs were unlikely to end in 'D' before then. The effect was to have
5989 incorrect size information for certain domains.
5991 45. #include "config.h" was missing at the start of the crypt16.c module. This
5992 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5995 46. If there was a timeout during a "random" callout check, Exim treated it as
5996 a failure of the random address, and carried on sending RSET and the real
5997 address. If the delay was just some slowness somewhere, the response to the
5998 original RCPT would be taken as a response to RSET and so on, causing
5999 mayhem of various kinds.
6001 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6002 when I implemented it. It didn't allow for the fact that some option values
6003 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6004 the right test for positive values.
6006 48. Domain names in DNS records are case-independent. Exim always looks them up
6007 in lower case. Some resolvers return domain names in exactly the case they
6008 appear in the zone file, that is, they may contain uppercase letters. Not
6009 all resolvers do this - some return always lower case. Exim was treating a
6010 change of case by a resolver as a change of domain, similar to a widening
6011 of a domain abbreviation. This triggered its re-routing code and so it was
6012 trying to route what was effectively the same domain again. This normally
6013 caused routing to fail (because the router wouldn't handle the domain
6014 twice). Now Exim checks for this case specially, and just changes the
6015 casing of the domain that it ultimately uses when it transmits the message
6018 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6021 50. If a filter generated a file delivery with a non-absolute name (possible if
6022 no home directory exists for the router), the forbid_file option was not
6025 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6026 the existing equality matching.
6028 52. Exim was using ints instead of ino_t variables in some places where it was
6029 dealing with inode numbers.
6031 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6032 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6033 called TMPDIR, and if it finds it is different, it changes its value.
6035 54. The smtp_printf() function is now made available to local_scan() so
6036 additional output lines can be written before returning. There is also an
6037 smtp_fflush() function to enable the detection of a dropped connection.
6038 The variables smtp_input and smtp_batched_input are exported to
6041 55. Changed the default runtime configuration: the message "Unknown user"
6042 has been removed from the ACL, and instead placed on the localuser router,
6043 using the cannot_route_message feature. This means that any verification
6044 failures that generate their own messages won't get overridden. Similarly,
6045 the "Unrouteable address" message that was in the ACL for unverifiable
6046 relay addresses has also been removed.
6048 56. Added hosts_avoid_esmtp to the smtp transport.
6050 57. The exicyclog script was not checking for the esoteric option
6051 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6052 will work only if exicyclog is run under the appropriate euid.
6054 58. Following a discussion on the list, the rules by which Exim recognises line
6055 endings on incoming messages have been changed. The -dropcr and drop_cr
6056 options are now no-ops, retained only for backwards compatibility. The
6057 following line terminators are recognized: LF CRLF CR. However, special
6058 processing applies to CR:
6060 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6061 nor a local message in the state where . is a terminator.
6063 (ii) If a bare CR is encountered in a header line, an extra space is added
6064 after the line terminator so as not to end the header. The reasoning
6065 behind this is that bare CRs in header lines are most likely either
6066 to be mistakes, or people trying to play silly games.
6068 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6069 was being incorrectly given as 18 bytes larger than it should have been.
6070 This is a VOB (very old bug).
6072 60. This may never have affected anything current, but just in case it has:
6073 When the local host is found other than at the start of a list of hosts,
6074 the local host, those with the same MX, and any that follow, are discarded.
6075 When the list in question was part of a longer list of hosts, the following
6076 hosts (not currently being processed) were also being discarded. This no
6077 longer happens. I'm not sure if this situation could ever has previously
6080 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6083 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6084 header. This is something that is recommended in a new Internet Draft, and
6085 is something that is documented as being done by Sendmail. There are two
6086 possible values. For messages generated by the autoreply transport, Exim
6089 Auto-Submitted: auto-replied
6091 whereas for all other generated messages (e.g. bounces) it adds
6093 Auto-Submitted: auto-generated
6095 63. The "personal" condition in filters now includes a test for the
6096 Auto-Submitted: header. If it contains the string "auto-" the message it
6097 not considered personal.
6099 64. Added rcpt_include_affixes as a generic transport option.
6101 65. Added queue_only_override (default true).
6103 66. Added the syslog_duplication option.
6105 67. If what should have been the first header line of a message consisted of
6106 a space followed by a colon, Exim was mis-interpreting it as a header line.
6107 It isn't of course - it is syntactically invalid and should therefore be
6108 treated as the start of the message body. The misbehaviour could have
6109 caused a number of strange effects, including loss of data in subsequent
6110 header lines, and spool format errors.
6112 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6113 client host had authenticated. This control can now be exercised by an ACL
6114 for more flexibility.
6116 69. By default, callouts do not happen when testing with -bh. There is now a
6117 variant, -bhc, which does actually run the callout code, including
6118 consulting and updating the callout cache.
6120 70. Added support for saslauthd authentication, courtesy of Alexander
6123 71. If statvfs() failed on the spool or log directories while checking their
6124 size for availability, Exim confusingly gave the error "space shortage".
6125 Furthermore, in debugging mode it crashed with a floating point exception.
6126 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6127 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6128 serious problem, Exim now writes to the main and panic logs when this
6129 happens, with details of the failure. It then refuses to accept the
6130 incoming message, giving the message "spool directory problem" or "log
6131 directory problem" with a 421 code for SMTP messages.
6133 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6134 0, 1, and 2 exist, because some OS complain for execs without them (see
6135 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6136 descriptors. However, the code omitted to check that the open succeeded,
6137 causing mysterious errors if for some reason the permissions on /dev/null
6138 got screwed. Now Exim writes a message to the main and panic logs, and
6139 bombs out if it can't open /dev/null.
6141 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6142 interact so that it is all more flexible. It is supposed to remain
6143 backwards compatible. Also added extra_local_interfaces.
6145 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6146 to bomb out with an assertion failure - to the client this appears as a
6147 connection drop. This problem occurs in the part of the code that was taken
6148 from the Samba project. Fortunately, the assertion is in a very simple
6149 function, so I have fixed this by reproducing the function inline in the
6150 one place where it is called, and arranging for authentication to fail
6151 instead of killing the process with assert().
6153 75. The SPA client code was not working when the server requested OEM rather
6154 than Unicode encoding.
6156 76. Added code to make require_files with a specific uid setting more usable in
6157 the case where statting the file as root fails - usually a non-root-mounted
6158 NFS file system. When this happens and the failure is EACCES, Exim now
6159 forks a subprocess and does the per-uid checking as the relevant uid.
6161 77. Added process_log_path.
6163 78. If log_file_path was not explicitly set, a setting of check_log_space or
6164 check_log_inodes was ignored.
6166 79. If a space check for the spool or log partitions fails, the incident is now
6167 logged. Of course, in the latter case the data may get lost...
6169 80. Added the %p formatting code to string_format() so that it can be used to
6170 print addresses in debug_print(). Adjusted all the address printing in the
6171 debugging in store.c to use %p rather than %d.
6173 81. There was a concern that a line of code in smtp_in.c could overflow a
6174 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6175 initially expressed, the concern was not well-founded, because trailing
6176 spaces are removed early. However, if the trailing spaces were followed by
6177 a NULL, they did not get removed, so the overflow was possible. Two fixes
6180 (a) I re-wrote the offending code in a cleaner fashion.
6181 (b) If an incoming SMTP command contains a NULL character, it is rejected
6184 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6185 runs initgroups(), so that if the Exim user is in any additional groups,
6186 they will be used during message reception.
6192 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob