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
13 TL/01 When running the test suite, the README says that variables such as
14 no_msglog_check are global and can be placed anywhere in a specific
15 test's script, however it was observed that placement needed to be near
16 the beginning for it to behave that way. Changed the runtest perl
17 script to read through the entire script once to detect and set these
18 variables, reset to the beginning of the script, and then run through
19 the script parsing/test process like normal.
21 TL/02 The BSD's have an arc4random API. One of the functions to induce
22 adding randomness was arc4random_stir(), but it has been removed in
23 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
24 function when detected.
26 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
27 cause callback expansion.
29 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
30 syntax errors in an expansion can be treated as a string instead of
31 logging or causing an error, due to the internal use of bool_lax
32 instead of bool when processing it.
34 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
35 server certificates when making smtp deliveries.
37 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
39 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
41 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
43 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
44 Merged patch from Sebastian Wiedenroth.
46 JH/05 Fix results-pipe from transport process. Several recipients, combined
47 with certificate use, exposed issues where response data items split
48 over buffer boundaries were not parsed properly. This eventually
49 resulted in duplicates being sent. This issue only became common enough
50 to notice due to the introduction of conection certificate information,
51 the item size being so much larger. Found and fixed by Wolfgang Breyha.
53 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
54 size buffer was used, resulting in syntax errors when an expansion
57 JH/07 Add support for directories of certificates when compiled with a GnuTLS
58 version 3.3.6 or later.
60 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
61 is EXPERIMENTAL_EVENT, the main-configuration and transport options
62 both become "event_action", the variables become $event_name, $event_data
63 and $event_defer_errno. There is a new variable $verify_mode, usable in
64 routers, transports and related events. The tls:cert event is now also
65 raised for inbound connections, if the main configuration event_action
68 TL/06 In test suite, disable OCSP for old versions of openssl which contained
69 early OCSP support, but no stapling (appears to be less than 1.0.0).
71 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
72 server certificate names available under the smtp transport option
73 "tls_verify_cert_hostname" now do not permit multi-component wildcard
76 JH/10 Time-related extraction expansions from certificates now use the main
77 option "timezone" setting for output formatting, and are consistent
78 between OpenSSL and GnuTLS compilations. Bug 1541.
80 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
81 encoded parameter in the incoming message. Bug 1558.
83 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
84 include certificate info, eximon was claiming there were spoolfile
87 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
89 JH/14 Log delivery-related information more consistently, using the sequence
90 "H=<name> [<ip>]" wherever possible.
92 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
93 are problematic for Debian distribution, omit them from the release
96 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
98 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
101 JH/17 Fix a null-indirection in certextract expansions when a nondefault
102 output list separator was used.
107 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
108 checkers that were complaining about end of non-void function with no
111 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
112 This was a regression intruduced in 4.83 by another bugfix.
114 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
116 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
117 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
123 TF/01 Correctly close the server side of TLS when forking for delivery.
125 When a message was received over SMTP with TLS, Exim failed to clear up
126 the incoming connection properly after forking off the child process to
127 deliver the message. In some situations the subsequent outgoing
128 delivery connection happened to have the same fd number as the incoming
129 connection previously had. Exim would try to use TLS and fail, logging
130 a "Bad file descriptor" error.
132 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
133 utilities have not been installed.
135 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
136 temporary space as the ACL may create new global variables.
138 TL/01 LDAP support uses per connection or global context settings, depending
139 upon the detected version of the libraries at build time.
141 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
142 to extract and use the src ip:port in logging and expansions as if it
143 were a direct connection from the outside internet. PPv2 support was
144 updated based on HAProxy spec change in May 2014.
146 JH/02 Add ${listextract {number}{list}{success}{fail}}.
148 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
149 Properly escape header and check for NULL return.
151 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
154 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
156 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
157 characters in header names, implemented as a verify condition.
158 Contributed by Michael Fischer v. Mollard.
160 TL/05 Rename SPF condition results err_perm and err_temp to standardized
161 results permerror and temperror. Previous values are deprecated but
162 still accepted. In a future release, err_perm and err_temp will be
163 completely removed, which will be a backward incompatibility if the
164 ACL tests for either of these two old results. Patch contributed by
165 user bes-internal on the mailing list.
167 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
169 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
170 selectors, in both main and reject logs.
172 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
175 JH/07 Add malware type "sock" for talking to simple daemon.
177 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
179 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
180 routers/transports under cutthrough routing.
182 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
183 numbers. Touch up "bool" conditional to keep the same definition.
185 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
187 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
189 JH/12 Expand items in router/transport headers_add or headers_remove lists
190 individually rather than the list as a whole. Bug 1452.
192 Required for reasonable handling of multiple headers_ options when
193 they may be empty; requires that headers_remove items with embedded
194 colons must have them doubled (or the list-separator changed).
196 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
197 view the policy declared in the DMARC record. Currently, $dmarc_status
198 is a combined value of both the record presence and the result of the
201 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
203 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
204 dnslookup router and the smtp transport (applying to the forward
207 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
208 of ldap servers used for a specific lookup. Patch provided by Heiko
211 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
212 New variable $lookup_dnssec_authenticated for observability.
214 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
215 Patch submitted by Lars Timman.
217 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
219 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
220 Requires trusted mode and valid format message id, aborts otherwise.
221 Patch contributed by Heiko Schlichting.
223 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
224 certextract with support for various fields. Bug 1358.
226 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
227 is requested by default, modifiable by smtp transport option
230 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
231 operate on certificate variables to give certificate fingerprints
232 Also new ${sha256:cert_variable}.
234 JH/23 The PRDR feature is moved from being Experimental into the mainline.
236 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
237 Christian Aistleitner.
239 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
241 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
242 file. Patch from Wolfgang Breyha.
244 JH/25 Expand the coverage of the delivery $host and $host_address to
245 client authenticators run in verify callout. Bug 1476.
247 JH/26 Port service names are now accepted for tls_on_connect_ports, to
248 align with daemon_smtp_ports. Bug 72.
250 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
251 support and error reporting did not work properly.
253 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
254 and is readable. Patch from Andrew Colin Kissa.
256 TL/14 Enhance documentation of ${run expansion and how it parses the
257 commandline after expansion, particularly in the case when an
258 unquoted variable expansion results in an empty value.
260 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
262 PP/02 Fix internal collision of T_APL on systems which support RFC3123
263 by renaming away from it. Addresses GH issue 15, reported by
266 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
268 TL/15 SECURITY: prevent double expansion in math comparison functions
269 (can expand unsanitized data). Not remotely exploitable.
276 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
278 PP/02 Make -n do something, by making it not do something.
279 When combined with -bP, the name of an option is not output.
281 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
284 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
285 $sender_host_name and config options to manage this, and basic check
288 PP/05 DSCP support for outbound connections and control modifier for inbound.
290 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
291 (Only plugin which currently uses this is kerberos4, which nobody should
292 be using, but we should make it available and other future plugins might
293 conceivably use it, even though it would break NAT; stuff *should* be
294 using channel bindings instead).
296 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
297 name; added for Sendmail compatibility; requires admin caller.
298 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
299 just ignore it); requires trusted caller.
300 Also parse but ignore: -Ac -Am -X<logfile>
303 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
305 TL/02 Add +smtp_confirmation as a default logging option.
307 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
308 Patch by Magnus Holmgren from 2007-02-20.
310 TL/04 Bugzilla 1281 - Spec typo.
311 Bugzilla 1283 - Spec typo.
312 Bugzilla 1290 - Spec grammar fixes.
314 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
316 TL/06 Add Experimental DMARC support using libopendmarc libraries.
318 TL/07 Fix an out of order global option causing a segfault. Reported to dev
319 mailing list by by Dmitry Isaikin.
321 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
323 JH/02 Support "G" suffix to numbers in ${if comparisons.
325 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
327 NM/01 Bugzilla 1197 - Spec typo
328 Bugzilla 1196 - Spec examples corrections
330 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
332 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
333 gnutls_enable_pkcs11, but renamed to more accurately indicate its
336 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
337 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
339 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
340 "acl {{name}{arg}...}", and optional args on acl condition
343 JH/05 Permit multiple router/transport headers_add/remove lines.
345 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
347 JH/07 Avoid using a waiting database for a single-message-only transport.
348 Performance patch from Paul Fisher. Bugzilla 1262.
350 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
353 JH/09 Add $headers_added variable, with content from use of ACL modifier
354 add_header (but not yet added to the message). Bugzilla 199.
356 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
357 Pulled from Bugzilla 817 by Wolfgang Breyha.
359 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
361 (nb: this is the same fix as in Exim 4.80.1)
363 JH/11 Add A= logging on delivery lines, and a client_set_id option on
366 JH/12 Add optional authenticated_sender logging to A= and a log_selector
369 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
371 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
372 advertise SMTP AUTH mechanism to us, instead of a generic
373 protocol violation error. Also, make Exim more robust to bad
374 data from the Dovecot auth socket.
376 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
378 When a queue runner is handling a message, Exim first routes the
379 recipient addresses, during which it prunes them based on the retry
380 hints database. After that it attempts to deliver the message to
381 any remaining recipients. It then updates the hints database using
384 So if a recipient address works intermittently, it can get repeatedly
385 deferred at routing time. The retry hints record remains fresh so the
386 address never reaches the final cutoff time.
388 This is a fairly common occurrence when a user is bumping up against
389 their storage quota. Exim had some logic in its local delivery code
390 to deal with this. However it did not apply to per-recipient defers
391 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
393 This change adds a proper retry rule check during routing so that the
394 final cutoff time is checked against the message's age. We only do
395 this check if there is an address retry record and there is not a
396 domain retry record; this implies that previous attempts to handle
397 the address had the retry_use_local_parts option turned on. We use
398 this as an approximation for the destination being like a local
399 delivery, as in LMTP.
401 I suspect this new check makes the old local delivery cutoff check
402 redundant, but I have not verified this so I left the code in place.
404 TF/02 Correct gecos expansion when From: is a prefix of the username.
406 Test 0254 submits a message to Exim with the header
410 When I ran the test suite under the user fanf2, Exim expanded
411 the header to contain my full name, whereas it should have added
412 a Resent-Sender: header. It erroneously treats any prefix of the
413 username as equal to the username.
415 This change corrects that bug.
417 GF/01 DCC debug and logging tidyup
418 Error conditions log to paniclog rather than rejectlog.
419 Debug lines prefixed by "DCC: " to remove any ambiguity.
421 TF/03 Avoid unnecessary rebuilds of lookup-related code.
423 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
424 Bug spotted by Jeremy Harris; was flawed since initial commit.
425 Would have resulted in OCSP responses post-SNI triggering an Exim
426 NULL dereference and crash.
428 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
430 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
431 Bug detection, analysis and fix by Samuel Thibault.
432 Bugzilla 1331, Debian bug #698092.
434 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
436 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
437 Server implementation by Todd Lyons, client by JH.
438 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
439 config variable "prdr_enable" controls whether the server
440 advertises the facility. If the client requests PRDR a new
441 acl_data_smtp_prdr ACL is called once for each recipient, after
442 the body content is received and before the acl_smtp_data ACL.
443 The client is controlled by bolth of: a hosts_try_prdr option
444 on the smtp transport, and the server advertisement.
445 Default client logging of deliveries and rejections involving
446 PRDR are flagged with the string "PRDR".
448 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
449 fclose(). Diagnosis by Todd Lyons.
451 PP/17 Update configure.default to handle IPv6 localhost better.
452 Patch by Alain Williams (plus minor tweaks).
455 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
456 This is now consistent with GnuTLS, and is now documented: the
457 previous undocumented portable approach to treating the option as
458 unset was to force an expansion failure. That still works, and
459 an empty string is now equivalent.
461 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
462 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
463 not performing validation itself.
465 PP/20 Added force_command boolean option to pipe transport.
466 Patch from Nick Koston, of cPanel Inc.
468 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
471 TF/04 Added udpsend ACL modifer and hexquote expansion operator
473 PP/21 Fix eximon continuous updating with timestamped log-files.
474 Broken in a format-string cleanup in 4.80, missed when I repaired the
475 other false fix of the same issue.
476 Report and fix from Heiko Schlichting.
479 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
480 Report from Prashanth Katuri.
482 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
483 It's SecureTransport, so affects any MacOS clients which use the
484 system-integrated TLS libraries, including email clients.
486 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
487 using a MIME ACL for non-SMTP local injection.
488 Report and assistance in diagnosis by Warren Baker.
490 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
492 JH/16 Fix comparisons for 64b. Bugzilla 1385.
494 TL/09 Add expansion variable $authenticated_fail_id to keep track of
495 last id that failed so it may be referenced in subsequent ACL's.
497 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
500 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
501 ldap library initialization, allowing self-signed CA's to be
502 used. Also properly sets require_cert option later in code by
503 using NULL (global ldap config) instead of ldap handle (per
504 session). Bug diagnosis and testing by alxgomz.
506 TL/12 Enhanced documentation in the ratelimit.pl script provided in
507 the src/util/ subdirectory.
509 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
510 renamed to Transport Post Delivery Action by Jeremy Harris, as
513 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
514 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
515 redis_servers = needs to be configured which will be used by the redis
516 lookup. Patch from Warren Baker, of The Packet Hub.
518 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
520 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
521 hostname or reverse DNS when processing a host list. Used suggestions
522 from multiple comments on this bug.
524 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
526 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
527 Missed a few lines, added it to make the runtest require no keyboard
530 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
531 contains upper case chars. Make router use caseful_local_part.
533 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
534 support when GnuTLS has been built with p11-kit.
540 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
542 This, or similar/improved, will also be change PP/11 of 4.82.
548 PP/01 Handle short writes when writing local log-files.
549 In practice, only affects FreeBSD (8 onwards).
550 Bugzilla 1053, with thanks to Dmitry Isaikin.
552 NM/01 Bugzilla 949 - Documentation tweak
554 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
557 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
559 PP/02 Implemented gsasl authenticator.
561 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
563 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
564 `pkg-config foo` for cflags/libs.
566 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
567 with rest of GSASL and with heimdal_gssapi.
569 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
570 `pkg-config foo` for cflags/libs for the TLS implementation.
572 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
573 properties get this fed in as external SSF. A number of robustness
574 and debugging improvements to the cyrus_sasl authenticator.
576 PP/08 cyrus_sasl server now expands the server_realm option.
578 PP/09 Bugzilla 1214 - Log authentication information in reject log.
579 Patch by Jeremy Harris.
581 PP/10 Added dbmjz lookup type.
583 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
585 PP/12 MAIL args handles TAB as well as SP, for better interop with
586 non-compliant senders.
587 Analysis and variant patch by Todd Lyons.
589 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
590 Bug report from Lars Müller <lars@samba.org> (via SUSE),
591 Patch from Dirk Mueller <dmueller@suse.com>
593 PP/13 tls_peerdn now print-escaped for spool files.
594 Observed some $tls_peerdn in wild which contained \n, which resulted
595 in spool file corruption.
597 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
598 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
599 or write after TLS renegotiation, which otherwise led to messages
602 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
603 as a tracking header (ie: a signed header comes before the signature).
604 Patch from Wolfgang Breyha.
606 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
607 comma-sep list; embedded commas doubled.
609 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
611 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
613 Report and patch from Dmitry Banschikov.
615 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
616 Removed SSL_clear() after SSL_new() which led to protocol negotiation
617 failures. We appear to now support TLS1.1+ with Exim.
619 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
620 lets Exim select keys and certificates based upon TLS SNI from client.
621 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
622 before an outbound SMTP session. New log_selector, +tls_sni.
624 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
625 NULL dereference. Report and patch from Alun Jones.
627 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
628 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
629 Not seeing resolver debug output on NetBSD, but suspect this is a
630 resolver implementation change.
632 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
633 Left warnings. Added "eximon gdb" invocation mode.
635 PP/21 Defaulting "accept_8bitmime" to true, not false.
637 PP/22 Added -bw for inetd wait mode support.
639 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
640 locate the relevant includes and libraries. Made this the default.
642 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
643 Bugzilla 1246, report and most of solution from Tomasz Kusy.
645 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
646 This may cause build issues on older platforms.
648 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
649 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
650 gnutls_require_mac & gnutls_require_protocols (no longer supported).
651 Added SNI support via GnuTLS too.
652 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
654 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
656 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
657 Applied second patch from Janne, implementing suggestion to default
658 multiple-strings-in-record handling to match SPF spec.
660 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
662 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
663 read-only, out of scope).
664 Patch from Wolfgang Breyha, report from Stuart Northfield.
666 PP/29 Fix three issues highlighted by clang analyser static analysis.
667 Only crash-plausible issue would require the Cambridge-specific
668 iplookup router and a misconfiguration.
669 Report from Marcin Mirosław.
671 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
673 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
674 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
675 As part of this, removing so much warning spew let me fix some minor
676 real issues in debug logging.
678 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
679 assignment on my part. Fixed.
681 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
682 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
683 Janne Snabb (who went above and beyond: thank you).
685 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
686 string otherwise requires a connection and a bunch more work and it's
687 relatively easy to get wrong. Should also expose TLS library linkage
690 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
691 64-bit ${eval} (JH/03).
693 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
694 GNU libc to support some of the 64-bit stuff, should not lead to
695 conflicts. Defined before os.h is pulled in, so if a given platform
696 needs to override this, it can.
698 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
699 protection layer was required, which is not implemented.
700 Bugzilla 1254, patch from Wolfgang Breyha.
702 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
703 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
704 tls_dhparam take prime identifiers. Also unbreak combination of
705 OpenSSL+DH_params+TLSSNI.
707 PP/39 Disable SSLv2 by default in OpenSSL support.
713 PP/01 Solaris build fix for Oracle's LDAP libraries.
714 Bugzilla 1109, patch from Stephen Usher.
716 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
718 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
721 TF/02 Fix a couple more cases where we did not log the error message
722 when unlink() failed. See also change 4.74-TF/03.
724 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
725 lock up or crash if it happened to be inside a call to libc when it
726 got a SIGUSR1 from exiwhat.
728 The SIGUSR1 handler appends the current process status to the process
729 log which is later printed by exiwhat. It used to use the general
730 purpose logging code to do this, but several functions it calls are
731 not safe for signals.
733 The new output code in the SIGUSR1 handler is specific to the process
734 log, and simple enough that it's easy to inspect for signal safety.
735 Removing some special cases also simplifies the general logging code.
736 Removing the spurious timestamps from the process log simplifies
739 TF/04 Improved ratelimit ACL condition.
741 The /noupdate option has been deprecated in favour of /readonly which
742 has clearer semantics. The /leaky, /strict, and /readonly update modes
743 are mutually exclusive. The update mode is no longer included in the
744 database key; it just determines when the database is updated. (This
745 means that when you upgrde Exim will forget old rate measurements.)
747 Exim now checks that the per_* options are used with an update mode that
748 makes sense for the current ACL. For example, when Exim is processing a
749 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
750 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
751 must specify per_mail/readonly. If you omit the update mode it defaults to
752 /leaky where that makes sense (as before) or /readonly where required.
754 The /noupdate option is now undocumented but still supported for
755 backwards compatibility. It is equivalent to /readonly except that in
756 ACLs where /readonly is required you may specify /leaky/noupdate or
757 /strict/noupdate which are treated the same as /readonly.
759 A useful new feature is the /count= option. This is a generalization
760 of the per_byte option, so that you can measure the throughput of other
761 aggregate values. For example, the per_byte option is now equivalent
762 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
764 The per_rcpt option has been generalized using the /count= mechanism
765 (though it's more complicated than the per_byte equivalence). When it is
766 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
767 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
768 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
769 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
770 Note that using per_rcpt with a non-readonly update mode in more than
771 one ACL will cause the recipients to be double-counted. (The per_mail
772 and per_byte options don't have this problem.)
774 The handling of very low rates has changed slightly. If the computed rate
775 is less than the event's count (usually one) then this event is the first
776 after a long gap. In this case the rate is set to the same as this event's
777 count, so that the first message of a spam run is counted properly.
779 The major new feature is a mechanism for counting the rate of unique
780 events. The new per_addr option counts the number of different
781 recipients that someone has sent messages to in the last time period. It
782 behaves like per_rcpt if all the recipient addresses are different, but
783 duplicate recipient addresses do not increase the measured rate. Like
784 the /count= option this is a general mechanism, so the per_addr option
785 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
786 example, measure the rate that a client uses different sender addresses
787 with the options per_mail/unique=$sender_address. There are further
788 details in the main documentation.
790 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
792 TF/06 Removed a few PCRE remnants.
794 TF/07 Automatically extract Exim's version number from tags in the git
795 repository when doing development or release builds.
797 PP/02 Raise smtp_cmd_buffer_size to 16kB.
798 Bugzilla 879. Patch from Paul Fisher.
800 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
801 Heavily based on revision 40f9a89a from Simon Arlott's tree.
804 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
806 PP/05 Variable $av_failed, true if the AV scanner deferred.
807 Bugzilla 1078. Patch from John Horne.
809 PP/06 Stop make process more reliably on build failure.
810 Bugzilla 1087. Patch from Heiko Schlittermann.
812 PP/07 Make maildir_use_size_file an _expandable_ boolean.
813 Bugzilla 1089. Patch from Heiko Schlittermann.
815 PP/08 Handle ${run} returning more data than OS pipe buffer size.
816 Bugzilla 1131. Patch from Holger Weiß.
818 PP/09 Handle IPv6 addresses with SPF.
819 Bugzilla 860. Patch from Wolfgang Breyha.
821 PP/10 GnuTLS: support TLS 1.2 & 1.1.
823 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
826 PP/11 match_* no longer expand right-hand-side by default.
827 New compile-time build option, EXPAND_LISTMATCH_RHS.
828 New expansion conditions, "inlist", "inlisti".
830 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
832 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
834 PP/14 fix log_write() format string regression from TF/03.
835 Bugzilla 1152. Patch from Dmitry Isaikin.
841 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
843 PP/02 Harmonised TLS library version reporting; only show if debugging.
844 Layout now matches that introduced for other libraries in 4.74 PP/03.
846 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
848 PP/04 New "dns_use_edns0" global option.
850 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
853 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
854 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
856 TK/01 Updated PolarSSL code to 0.14.2.
857 Bugzilla 1097. Patch from Andreas Metzler.
859 PP/07 Catch divide-by-zero in ${eval:...}.
862 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
865 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
866 format-string attack -- SECURITY: remote arbitrary code execution.
868 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
869 time unintentionally subject to list matching rules, letting the header
870 cause arbitrary Exim lookups (of items which can occur in lists, *not*
871 arbitrary string expansion). This allowed for information disclosure.
873 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
874 INT_MIN/-1 -- value coerced to INT_MAX.
880 NM/01 Workround for PCRE version dependency in version reporting
883 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
884 This fixes portability to compilers other than gcc, notably
885 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
887 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
888 makefiles for portability to HP-UX and POSIX correctness.
890 PP/01 Permit LOOKUP_foo enabling on the make command-line.
891 Also via indented variable definition in the Makefile.
892 (Debugging by Oliver Heesakkers).
894 PP/02 Restore caching of spamd results with expanded spamd_address.
895 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
897 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
898 Improves build reliability. Fix from: Frank Elsner
900 NM/02 Fix wide character breakage in the rfc2047 coding
901 Fixes bug 1064. Patch from Andrey N. Oktyabrski
903 NM/03 Allow underscore in dnslist lookups
904 Fixes bug 1026. Patch from Graeme Fowler
906 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
907 Code patches from Adam Ciarcinski of NetBSD.
909 NM/04 Fixed exiqgrep to cope with mailq missing size issue
912 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
913 is logged, to avoid truncation. Patch from John Horne.
915 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
916 Patch from Jakob Hirsch.
918 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
919 SQL string expansion failure details.
920 Patch from Andrey Oktyabrski.
922 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
923 Patch from Simon Arlott.
925 PP/09 New lookups functionality failed to compile on old gcc which rejects
926 extern declarations in function scope.
927 Patch from Oliver Fleischmann
929 PP/10 Use sig_atomic_t for flags set from signal handlers.
930 Check getgroups() return and improve debugging.
931 Fixed developed for diagnosis in bug 927 (which turned out to be
934 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
935 Patch from Mark Zealey.
937 PP/12 Bugzilla 1056: Improved spamd server selection.
938 Patch from Mark Zealey.
940 PP/13 Bugzilla 1086: Deal with maildir quota file races.
941 Based on patch from Heiko Schlittermann.
943 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
944 Patch from Uwe Doering, sign-off by Michael Haardt.
946 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
947 variable declaration deep within a block. Bug and patch from
950 PP/15 lookups-Makefile IRIX compatibilty coercion.
952 PP/16 Make DISABLE_DKIM build knob functional.
954 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
955 Patch by Simon Arlott
957 TF/03 Fix valgrind.h portability to C89 compilers that do not support
958 variable argument macros. Our copy now differs from upstream.
964 TF/01 Failure to get a lock on a hints database can have serious
965 consequences so log it to the panic log.
967 TF/02 Log LMTP confirmation messages in the same way as SMTP,
968 controlled using the smtp_confirmation log selector.
970 TF/03 Include the error message when we fail to unlink a spool file.
972 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
973 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
974 for maintaining out-of-tree patches for some time.
976 PP/01 Bugzilla 139: Documentation and portability issues.
977 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
978 Handle per-OS dynamic-module compilation flags.
980 PP/02 Let /dev/null have normal permissions.
981 The 4.73 fixes were a little too stringent and complained about the
982 permissions on /dev/null. Exempt it from some checks.
983 Reported by Andreas M. Kirchwitz.
985 PP/03 Report version information for many libraries, including
986 Exim version information for dynamically loaded libraries. Created
987 version.h, now support a version extension string for distributors
988 who patch heavily. Dynamic module ABI change.
990 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
991 privilege escalation vulnerability whereby the Exim run-time user
992 can cause root to append content of the attacker's choosing to
995 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
998 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
999 If dropping privileges for untrusted macros, we disabled normal logging
1000 on the basis that it would fail; for the Exim run-time user, this is not
1001 the case, and it resulted in successful deliveries going unlogged.
1002 Fixed. Reported by Andreas Metzler.
1008 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1009 only prepend for the Resent-* case. Fixes regression introduced in
1010 Exim 4.70 by NM/22 for Bugzilla 607.
1012 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1013 increasing numbers of administrators be bitten by this.
1015 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1017 PP/03 Bugzilla 994: added openssl_options main configuration option.
1019 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1021 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1023 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1025 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1026 without a peer certificate, leading to a segfault because of an
1027 assumption that peers always have certificates. Be a little more
1028 paranoid. Problem reported by Martin Tscholak.
1030 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1031 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1032 NB: ClamAV planning to remove STREAM in "middle of 2010".
1033 CL also introduces -bmalware, various -d+acl logging additions and
1034 more caution in buffer sizes.
1036 PP/09 Implemented reverse_ip expansion operator.
1038 PP/10 Bugzilla 937: provide a "debug" ACL control.
1040 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1042 PP/12 Bugzilla 973: Implement --version.
1044 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1046 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1048 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1050 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1051 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1052 ignore trailing whitespace.
1054 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1056 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1059 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1060 Notification from Dr Andrew Aitchison.
1062 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1063 ExtendedDetectionInfo response format.
1064 Notification from John Horne.
1066 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1069 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1070 XSL and documented dependency on system catalogs, with examples of how
1073 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1076 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1077 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1078 configuration file which is writeable by the Exim user or group.
1080 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1081 of configuration files to cover files specified with the -C option if
1082 they are going to be used with root privileges, not just the default
1085 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1086 option (effectively making it always true).
1088 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1089 files to be used while preserving root privileges.
1091 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1092 that rogue child processes cannot use them.
1094 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1095 run-time user, instead of root.
1097 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1098 Exim run-time user without dropping privileges.
1100 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1101 result string, instead of calling string_vformat() twice with the same
1104 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1105 for other users. Others should always drop root privileges if they use
1106 -C on the command line, even for a whitelisted configure file.
1108 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1110 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1116 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1117 $data_path, and $header_path variables; fixed documentation bugs and
1120 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1121 exipick to access non-standard spools, including the "frozen" queue
1124 NM/01 Bugzilla 965: Support mysql stored procedures.
1125 Patch from Alain Williams
1127 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1129 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1130 Patch from Andreas Metzler
1132 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1133 Patch from Kirill Miazine
1135 NM/05 Bugzilla 671: Added umask to procmail example.
1137 JJ/03 installed exipick 20100323.0, fixing doc bug
1139 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1140 directory. Notification and patch from Dan Rosenberg.
1142 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1144 TK/02 Improve log output when DKIM signing operation fails.
1146 MH/01 Treat the transport option dkim_domain as a colon separated
1147 list, not as a single string, and sign the message with each element,
1148 omitting multiple occurences of the same signer.
1150 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1151 Bugzilla 985, 986. Patch by Simon Arlott
1153 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1154 Patch by Simon Arlott
1156 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1157 MBX locking. Notification from Dan Rosenberg.
1163 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1165 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1167 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1169 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1171 NM/04 Bugzilla 915: Flag broken perl installation during build.
1177 TK/01 Added patch by Johannes Berg that expands the main option
1178 "spamd_address" if it starts with a dollar sign.
1180 TK/02 Write list of recipients to X-Envelope-Sender header when building
1181 the mbox-format spool file for content scanning (suggested by Jakob
1184 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1185 (http://www.dcc-servers.net/) support via dccifd. Activated by
1186 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1188 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1189 by Mark Daniel Reidel <mr@df.eu>.
1191 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1192 When building exim an external PCRE library is now needed -
1193 PCRE is a system library on the majority of modern systems.
1194 See entry on PCRE_LIBS in EDITME file.
1196 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1197 conversation. Added nologin parameter to request.
1198 Patch contributed by Kirill Miazine.
1200 TF/01 Do not log submission mode rewrites if they do not change the address.
1202 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1204 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1205 log files in place. Contributed by Roberto Lima.
1207 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1209 TF/03 Bugzilla 615: When checking the local_parts router precondition
1210 after a local_part_suffix or local_part_prefix option, Exim now
1211 does not use the address's named list lookup cache, since this
1212 contains cached lookups for the whole local part.
1214 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1215 Robert Millan. Documentation is in experimental-spec.txt.
1217 TF/04 Bugzilla 668: Fix parallel build (make -j).
1219 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1221 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1222 Patch provided by Jan Srzednicki.
1224 TF/05 Leading white space used to be stripped from $spam_report which
1225 wrecked the formatting. Now it is preserved.
1227 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1228 that they are available at delivery time.
1230 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1232 TF/08 TLS error reporting now respects the incoming_interface and
1233 incoming_port log selectors.
1235 TF/09 Produce a more useful error message if an SMTP transport's hosts
1236 setting expands to an empty string.
1238 NM/06 Bugzilla 744: EXPN did not work under TLS.
1239 Patch provided by Phil Pennock.
1241 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1242 Patch provided by Richard Godbee.
1244 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1245 acl_smtp_notquit, added index entry.
1247 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1248 Patch provided by Eugene Bujak.
1250 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1251 accept(). Patch provided by Maxim Dounin.
1253 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1254 Patch provided by Phil Pennock.
1256 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1258 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1259 Patch provided by Brad "anomie" Jorsch.
1261 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1262 Patch provided by Dean Brooks.
1264 TK/05 Add native DKIM support (does not depend on external libraries).
1266 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1267 Patch provided by Graeme Fowler.
1269 NM/16 Bugzilla 851: Documentation example syntax fix.
1271 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1273 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1276 NM/19 Bugzilla 745: TLS version reporting.
1277 Patch provided by Phil Pennock.
1279 NM/20 Bugzilla 167: bool: condition support.
1280 Patch provided by Phil Pennock.
1282 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1283 clients. Patch provided by Phil Pennock.
1285 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1286 Patch provided by Brad "anomie" Jorsch.
1288 NM/23 Bugzilla 687: Fix misparses in eximstats.
1289 Patch provided by Heiko Schlittermann.
1291 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1292 Patch provided by Heiko Schlittermann.
1294 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1295 plus update to original patch.
1297 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1299 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1300 Patch provided by David Brownlee.
1302 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1304 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1306 NM/30 Bugzilla 888: TLS documentation bugfixes.
1308 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1310 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1311 Unlike the original bugzilla I have changed all shell scripts in src tree.
1313 NM/33 Bugzilla 898: Transport filter timeout fix.
1314 Patch by Todd Rinaldo.
1316 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1317 Patch by Serge Demonchaux.
1319 NM/35 Bugzilla 39: Base64 decode bug fixes.
1320 Patch by Jakob Hirsch.
1322 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1324 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1326 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1328 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1334 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1335 ALT-N's libdkim that I have put here:
1336 http://duncanthrax.net/exim-experimental/
1338 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1339 were called 'true' and it seems that C99 defines that as a reserved
1340 keyword to be used with 'bool' variable types. That means you could
1341 not include C99-style headers which use bools without triggering
1342 build errors in sieve.c.
1344 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1345 as mailq or other aliases. Changed the --help handling significantly
1346 to do whats expected. exim_usage() emits usage/help information.
1348 SC/01 Added the -bylocaldomain option to eximstats.
1350 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1352 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1354 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1360 PH/01 Another patch from the Sieve maintainer.
1362 PH/02 When an IPv6 address is converted to a string for single-key lookup
1363 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1364 dots are used instead of colons so that keys in lsearch files need not
1365 contain colons. This was done some time before quoting was made available
1366 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1367 (notated using the quote facility) so as to distinguish them from IPv4
1368 keys. This meant that lookups for IP addresses in host lists did not work
1369 for iplsearch lookups.
1371 This has been fixed by arranging for IPv6 addresses to be expressed with
1372 colons if the lookup type is iplsearch. This is not incompatible, because
1373 previously such lookups could never work.
1375 The situation is now rather anomolous, since one *can* have colons in
1376 ordinary lsearch keys. However, making the change in all cases is
1377 incompatible and would probably break a number of configurations.
1379 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1382 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1383 conversion specification without a maximum field width, thereby enabling
1384 a rogue spamd server to cause a buffer overflow. While nobody in their
1385 right mind would setup Exim to query an untrusted spamd server, an
1386 attacker that gains access to a server running spamd could potentially
1387 exploit this vulnerability to run arbitrary code as the Exim user.
1389 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1390 $primary_hostname instead of what libspf2 thinks the hosts name is.
1392 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1393 a directory entry by the name of the lookup key. Previously, if a
1394 symlink pointed to a non-existing file or a file in a directory that
1395 Exim lacked permissions to read, a lookup for a key matching that
1396 symlink would fail. Now it is enough that a matching directory entry
1397 exists, symlink or not. (Bugzilla 503.)
1399 PH/03 The body_linecount and body_zerocount variables are now exported in the
1402 PH/04 Added the $dnslist_matched variable.
1404 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1405 This means they are set thereafter only if the connection becomes
1408 PH/06 Added the client_condition to authenticators so that some can be skipped
1409 by clients under certain conditions.
1411 PH/07 The error message for a badly-placed control=no_multiline_responses left
1412 "_responses" off the end of the name.
1414 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1416 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1417 (without spaces) instead of just copying the configuration text.
1419 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1421 PH/11 Added $max_received_linelength.
1423 PH/12 Added +ignore_defer and +include_defer to host lists.
1425 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1426 way in which PCRE > 7.0 is built.
1428 PH/14 Implemented queue_only_load_latch.
1430 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1431 MAIL command. The effect was to mangle the value on 64-bit systems.
1433 PH/16 Another patch from the Sieve maintainer.
1435 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1437 PH/18 If a system quota error occurred while trying to create the file for
1438 a maildir delivery, the message "Mailbox is full" was not appended to the
1439 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1440 only to a quota excession during the actual writing of the file.
1442 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1443 characters?) which causes problems in log lines. The DN values are now
1444 passed through string_printing() before being added to log lines.
1446 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1447 and InterBase are left for another time.)
1449 PH/21 Added message_body_newlines option.
1451 PH/22 Guard against possible overflow in moan_check_errorcopy().
1453 PH/23 POSIX allows open() to be a macro; guard against that.
1455 PH/24 If the recipient of an error message contained an @ in the local part
1456 (suitably quoted, of course), incorrect values were put in $domain and
1457 $local_part during the evaluation of errors_copy.
1463 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1464 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1467 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1468 issue a MAIL command.
1470 PH/02 In an ACL statement such as
1472 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1474 if a client was not listed at all, or was listed with a value other than
1475 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1476 the condition was not true (as it should be), so access was not denied.
1477 The bug was that the ! inversion was incorrectly passed on to the second
1478 item. This has been fixed.
1480 PH/03 Added additional dnslists conditions == and =& which are different from
1481 = and & when the dns lookup returns more than one IP address.
1483 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1484 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1486 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1487 FSYNC, which compiles an option called disable_fsync that allows for
1488 bypassing fsync(). The documentation is heavily laced with warnings.
1490 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1492 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1493 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1494 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1495 including adding "make clean"; (3) Added -fPIC when compiling the test
1496 dynamically loaded module, to get rid of a warning.
1498 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1499 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1500 = 0s. The bug is otherwise harmless.
1502 PH/07 There was a bug in the dovecot authenticator such that the value of
1503 $auth1 could be overwritten, and so not correctly preserved, after a
1504 successful authentication. This usually meant that the value preserved by
1505 the server_setid option was incorrect.
1507 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1509 PH/09 Installed PCRE release 7.0.
1511 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1512 run for batched SMTP input. It is now run at the start of every message
1513 in the batch. While fixing this I discovered that the process information
1514 (output by running exiwhat) was not always getting set for -bs and -bS
1515 input. This is fixed, and it now also says "batched" for BSMTP.
1517 PH/11 Added control=no_pipelining.
1519 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1520 patch, slightly modified), and move the expansion of helo_data till after
1521 the connection is made in the smtp transport (so it can use these
1524 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1526 PH/14 Added log_selector = +pid.
1528 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1530 PH/16 Add ${if forany and ${if forall.
1532 PH/17 Added dsn_from option to vary the From: line in DSNs.
1534 PH/18 Flush SMTP output before performing a callout, unless control =
1535 no_callout_flush is set.
1537 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1538 was true (the default) a successful delivery failed to delete the retry
1539 item, thus causing premature timeout of the address. The bug is now
1542 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1544 PH/21 Long custom messages for fakedefer and fakereject are now split up
1545 into multiline reponses in the same way that messages for "deny" and
1546 other ACL rejections are.
1548 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1549 with slight modification.
1551 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1552 draft, changing the syntax and factoring some duplicate code".
1554 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1555 for deliveries of the second and subsequent messages over the same SMTP
1558 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1559 ${reduce, with only minor "tidies".
1561 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1563 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1564 expansion side effects.
1566 PH/27 When a message times out after an over-quota error from an Exim-imposed
1567 quota, the bounce message says "mailbox is full". This message was not
1568 being given when it was a system quota that was exceeded. It now should
1571 MH/03 Made $recipients available in local_scan(). local_scan() already has
1572 better access to the recipient list through recipients_list[], but
1573 $recipients can be useful in postmaster-provided expansion strings.
1575 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1576 in the case of a MAIL command with additional options following the
1577 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1578 were accidentally chopped off.
1580 PH/29 SMTP synchronization checks are implemented when a command is read -
1581 there is a check that no more input is waiting when there shouldn't be
1582 any. However, for some commands, a delay in an ACL can mean that it is
1583 some time before the response is written. In this time, more input might
1584 arrive, invalidly. So now there are extra checks after an ACL has run for
1585 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1586 pipelining has not been advertised.
1588 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1590 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1591 returned in the SMTP response when smtp_return_error_details was false.
1592 This has been fixed.
1594 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1595 instead of the C I/O that was originally supplied, because problems were
1596 reported on Solaris.
1598 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1599 Exim which did not show up earlier: it was assuming that a call to
1600 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1601 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1602 was a macro that became an assignment, so it seemed to work. This has
1603 changed to a proper function call with a void return, hence the compile
1604 error. Exim's code has been fixed.
1606 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1609 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1611 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1613 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1614 because it turns out to be too big or there is a timeout) there is no
1615 "Completed" line in the log. When some messages of this type were
1616 selected by exigrep, they were listed as "not completed". Others were
1617 picked up by some special patterns. I have improved the selection
1618 criteria to be more general.
1620 PH/38 The host_find_failed option in the manualroute router can now be set
1621 to "ignore", to completely ignore a host whose IP address cannot be
1622 found. If all hosts are ignored, the behaviour is controlled by the new
1623 host_all_ignored option.
1625 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1626 homing or because of multiple MX records with /mx) generated more than
1627 one IP address, and the following item turned out to be the local host,
1628 all the secondary addresses of the first item were incorrectly removed
1629 from the list, along with the local host and any following hosts (which
1630 is what is supposed to happen).
1632 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1633 whoever called Exim into the -H file. In the case of the daemon it was
1634 behaving confusingly. When first started, it used values for whoever
1635 started the daemon, but after a SIGHUP it used the Exim user (because it
1636 calls itself on a restart). I have changed the code so that it now always
1639 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1640 message are rejected with the same error (e.g. no authentication or bad
1641 sender address), and a DATA command is nevertheless sent (as can happen
1642 with PIPELINING or a stupid MUA), the error message that was given to the
1643 RCPT commands is included in the rejection of the DATA command. This is
1644 intended to be helpful for MUAs that show only the final error to their
1647 PH/42 Another patch from the Sieve maintainer.
1649 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1650 Eximstats - Fixed some broken HTML links and added missing column headers
1652 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1653 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1655 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1657 PH/43 Yet another patch from the Sieve maintainer.
1659 PH/44 I found a way to check for a TCP/IP connection going away before sending
1660 the response to the final '.' that terminates a message, but only in the
1661 case where the client has not sent further data following the '.'
1662 (unfortunately, this is allowed). However, in many cases there won't be
1663 any further data because there won't be any more messages to send. A call
1664 to select() can be used: if it shows that the input is "ready", there is
1665 either input waiting, or the socket has been closed. An attempt to read
1666 the next input character can distinguish the two cases. Previously, Exim
1667 would have sent an OK response which the client would never have see.
1668 This could lead to message repetition. This fix should cure that, at
1669 least in a lot of common cases.
1671 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1672 advertised in response to EHLO.
1678 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1679 fixed by 4.65/MH/01 (is this a record?) are fixed:
1681 (i) An empty string was always treated as zero by the numeric comparison
1682 operators. This behaviour has been restored.
1684 (ii) It is documented that the numeric comparison operators always treat
1685 their arguments as decimal numbers. This was broken in that numbers
1686 starting with 0 were being interpreted as octal.
1688 While fixing these problems I realized that there was another issue that
1689 hadn't been noticed. Values of message_size_limit (both the global option
1690 and the transport option) were treated as octal if they started with 0.
1691 The documentation was vague. These values are now always treated as
1692 decimal, and I will make that clear in the documentation.
1698 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1699 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1702 MH/01 Don't check that the operands of numeric comparison operators are
1703 integers when their expansion is in "skipping" mode (fixes bug
1704 introduced by 4.64-PH/07).
1706 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1707 child addresses, Exim now panics and dies. Previously, because the count
1708 is held in a short int, deliveries were likely to be lost. As such a
1709 large number of recipients for a single message is ridiculous
1710 (performance will be very, very poor), I have chosen to impose a limit
1711 rather than extend the field.
1717 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1718 leftover -K file (the existence of which was triggered by #402).
1719 While we were at it, introduced process PID as part of the -K
1720 filename. This should rule out race conditions when creating
1723 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1724 processing considerably. Previous code took too long for large mails,
1725 triggering a timeout which in turn triggers #401.
1727 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1728 in the DK code in transports.c. sendfile() is not really portable,
1729 hence the _LINUX specificness.
1731 TF/01 In the add_headers option to the mail command in an Exim filter,
1732 there was a bug that Exim would claim a syntax error in any
1733 header after the first one which had an odd number of characters
1736 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1737 callout verification, Exim cached a "reject" for the entire domain. This
1738 is correct for most verifications, but it is not correct for a recipient
1739 verification with use_sender or use_postmaster set, because in that case
1740 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1741 case of MAIL FROM:<> rejection from other early rejections (e.g.
1742 rejection of HELO). When verifying a recipient using a non-null MAIL
1743 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1744 Whatever the result of the callout, the value of the domain cache is
1745 left unchanged (for any other kind of callout, getting as far as trying
1746 RCPT means that the domain itself is ok).
1748 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1751 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1752 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1753 session, the daemon ignores ECONNECT errors and logs others; it now
1754 ignores EPIPE as well.
1756 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1757 (quoted-printable decoding).
1759 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1760 later the small subsequent patch to fix an introduced bug.
1762 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1764 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1766 PH/08 An error is now given if message_size_limit is specified negative.
1768 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1769 to be given (somewhat) arbitrary names.
1771 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1774 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1775 miscellaneous code fixes
1777 PH/10 Added the log_reject_target ACL modifier to specify where to log
1780 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1781 hostname. This is wrong, because it relates to the incoming message (and
1782 probably the interface on which it is arriving) and not to the outgoing
1783 callout (which could be using a different interface). This has been
1784 changed to use the value of the helo_data option from the smtp transport
1785 instead - this is what is used when a message is actually being sent. If
1786 there is no remote transport (possible with a router that sets up host
1787 addresses), $smtp_active_hostname is used.
1789 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1790 tweaks were necessary in order to get it to work (see also 21 below):
1791 (a) The code assumed that strncpy() returns a negative number on buffer
1792 overflow, which isn't the case. Replaced with Exim's string_format()
1794 (b) There were several signed/unsigned issues. I just did the minimum
1795 hacking in of casts. There is scope for a larger refactoring.
1796 (c) The code used strcasecmp() which is not a standard C function.
1797 Replaced with Exim's strcmpic() function.
1798 (d) The code set only $1; it now sets $auth1 as well.
1799 (e) A simple test gave the error "authentication client didn't specify
1800 service in request". It would seem that Dovecot has changed its
1801 interface. Fortunately there's a specification; I followed it and
1802 changed what the client sends and it appears to be working now.
1804 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1807 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1808 address A is aliased to B and C, where B exists and C does not. Without
1809 -v the output is "A verified" because verification stops after a
1810 successful redirection if more than one address is generated. However,
1811 with -v the child addresses are also verified. Exim was outputting "A
1812 failed to verify" and then showing the successful verification for C,
1813 with its parentage. It now outputs "B failed to verify", showing B's
1814 parentage before showing the successful verification of C.
1816 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1817 look up a TXT record in a specific list after matching in a combined
1820 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1821 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1822 they consult the DNS. I had assumed they would set it the way they
1823 wanted; and indeed my experiments on Linux seem to show that in some
1824 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1825 To be on the safe side, however, I have now made the interface to
1826 host_find_byname() similar to host_find_bydns(), with an argument
1827 containing the DNS resolver options. The host_find_byname() function now
1828 sets these options at its start, just as host_find_bydns() does. The smtp
1829 transport options dns_qualify_single and dns_search_parents are passed to
1830 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1831 of host_find_byname() use the default settings of RES_DEFNAMES
1832 (qualify_single) but not RES_DNSRCH (search_parents).
1834 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1835 spool_read_header() do less string testing, by means of a preliminary
1836 switch on the second character of optional "-foo" lines. (This is
1837 overdue, caused by the large number of possibilities that now exist.
1838 Originally there were few.) While I was there, I also converted the
1839 str(n)cmp tests so they don't re-test the leading "-" and the first
1840 character, in the hope this might squeeze out yet more improvement.
1842 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1843 flag allowing group syntax was set by the header_syntax check but not
1844 turned off, possible causing trouble later; (2) The flag was not being
1845 set at all for the header_verify test, causing "group"-style headers to
1846 be rejected. I have now set it in this case, and also caused header_
1847 verify to ignore an empty address taken from a group. While doing this, I
1848 came across some other cases where the code for allowing group syntax
1849 while scanning a header line wasn't quite right (mostly, not resetting
1850 the flag correctly in the right place). These bugs could have caused
1851 trouble for malformed header lines. I hope it is now all correct.
1853 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1854 with the "reply" argument non-NULL. The code, however (which originally
1855 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1856 but it didn't always do it. This confused somebody who was copying the
1857 code for some other use. I have removed all the tests.
1859 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1860 feature that was used to support insecure browsers during the U.S. crypto
1861 embargo. It requires special client support, and Exim is probably the
1862 only MTA that supported it -- and would never use it because real RSA is
1863 always available. This code has been removed, because it had the bad
1864 effect of slowing Exim down by computing (never used) parameters for the
1865 RSA_EXPORT functionality.
1867 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1868 authenticator to fail if there's a tab character in the incoming data
1869 (there should never be unless someone is messing about, as it's supposed
1870 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1871 if the connection is using TLS or if the remote IP is the same as the
1872 local IP, and the "valid-client-cert option" if a client certificate has
1875 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1876 authenticators. This can be used for authorization after authentication
1877 succeeds. (In the case of plaintext, it servers for both authentication
1880 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1881 if any retry times were supplied.
1883 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1884 connection, where there is no client IP address to check. In this
1885 situation, the verify now always succeeds.
1887 PH/25 Applied John Jetmore's -Mset patch.
1889 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1891 PH/27 In a string expansion for a processed (not raw) header when multiple
1892 headers of the same name were present, leading whitespace was being
1893 removed from all of them, but trailing whitespace was being removed only
1894 from the last one. Now trailing whitespace is removed from each header
1895 before concatenation. Completely empty headers in a concatenation (as
1896 before) are ignored.
1898 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1899 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1901 PH/29 [Removed. This was a change that I later backed out, and forgot to
1902 correct the ChangeLog entry (that I had efficiently created) before
1903 committing the later change.]
1905 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1906 address errors (4xx response to RCPT) over the same connection as other
1907 messages routed to the same hosts. Such deliveries are always "forced",
1908 so retry times are not inspected. This resulted in far too many retries
1909 for the affected addresses. The effect occurred only when there were more
1910 hosts than the hosts_max_try setting in the smtp transport when it had
1911 the 4xx errors. Those hosts that it had tried were not added to the list
1912 of hosts for which the message was waiting, so if all were tried, there
1913 was no problem. Two fixes have been applied:
1915 (i) If there are any address or message errors in an SMTP delivery, none
1916 of the hosts (tried or untried) are now added to the list of hosts
1917 for which the message is waiting, so the message should not be a
1918 candidate for sending over the same connection that was used for a
1919 successful delivery of some other message. This seems entirely
1920 reasonable: after all the message is NOT "waiting for some host".
1921 This is so "obvious" that I'm not sure why it wasn't done
1922 previously. Hope I haven't missed anything, but it can't do any
1923 harm, as the worst effect is to miss an optimization.
1925 (ii) If, despite (i), such a delivery is accidentally attempted, the
1926 routing retry time is respected, so at least it doesn't keep
1927 hammering the server.
1929 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1930 in ${readsocket because some servers need this prod.
1932 PH/32 Added some extra debug output when updating a wait-xxx database.
1934 PH/33 The hint "could be header name not terminated by colon", which has been
1935 given for certain expansion errors for a long time, was not being given
1936 for the ${if def:h_colon_omitted{... case.
1938 PH/34 The spec says: "With one important exception, whenever a domain list is
1939 being scanned, $domain contains the subject domain." There was at least
1940 one case where this was not true.
1942 PH/35 The error "getsockname() failed: connection reset by peer" was being
1943 written to the panic log as well as the main log, but it isn't really
1944 panic-worthy as it just means the connection died rather early on. I have
1945 removed the panic log writing for the ECONNRESET error when getsockname()
1948 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1949 runs only) independently of the message's sender address. This meant
1950 that, if the 4xx error was in fact related to the sender, a different
1951 message to the same recipient with a different sender could confuse
1952 things. In particualar, this can happen when sending to a greylisting
1953 server, but other circumstances could also provoke similar problems.
1954 I have changed the default so that the retry time for these errors is now
1955 based a combination of the sender and recipient addresses. This change
1956 can be overridden by setting address_retry_include_sender=false in the
1959 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1960 remote server are returned as part of bounce messages. This was not
1961 happening for LMTP over a pipe (the lmtp transport), but now it is the
1962 same for both kinds of LMTP.
1964 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1965 in header lines that were in fact CNAMEs. This is no longer the case.
1967 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1968 and queue runs started by the daemon processed all messages. This has
1969 been fixed so that -R and -S can now usefully be given with -q<time>.
1971 PH/40 Import PCRE release 6.7 (fixes some bugs).
1973 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1975 PH/42 Give an error if -q is specified more than once.
1977 PH/43 Renamed the variables $interface_address and $interface_port as
1978 $received_ip_address and $received_port, to make it clear that these
1979 values apply to message reception, and not to the outgoing interface when
1980 a message is delivered. (The old names remain recognized, of course.)
1982 PH/44 There was no timeout on the connect() call when using a Unix domain
1983 socket in the ${readsocket expansion. There now is.
1985 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1986 be meaningful with "accept".
1988 SC/01 Eximstats V1.43
1989 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1991 SC/02 Eximstats V1.44
1992 Use a glob alias rather than an array ref in the generated
1993 parser. This improves both readability and performance.
1995 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1996 Collect SpamAssassin and rejection statistics.
1997 Don't display local sender or destination tables unless
1998 there is data to show.
1999 Added average volumes into the top table text output.
2001 SC/04 Eximstats V1.46
2002 Collect data on the number of addresses (recipients)
2003 as well as the number of messages.
2005 SC/05 Eximstats V1.47
2006 Added 'Message too big' to the list of mail rejection
2007 reasons (thanks to Marco Gaiarin).
2009 SC/06 Eximstats V1.48
2010 Mainlog lines which have GMT offsets and are too short to
2011 have a flag are now skipped.
2013 SC/07 Eximstats V1.49 (Alain Williams)
2014 Added the -emptyok flag.
2016 SC/08 Eximstats V1.50
2017 Fixes for obtaining the IP address from reject messages.
2019 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2020 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2021 whitesspace changes from 4.64-PH/27
2023 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2026 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2027 are found, allow negative numbers in numeric criteria)
2029 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2031 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2032 to match changes made in 4.64-PH/43
2034 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2036 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2037 "domain" is actually an IP address, and force a failure. This locks out
2038 those revolvers/nameservers that support "A-for-A" lookups, in
2039 contravention of the specifications.
2041 PH/48 When a host name was looked up from an IP address, and the subsequent
2042 forward lookup of the name timed out, the host name was left in
2043 $sender_host_name, contrary to the specification.
2045 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2046 restricted to single-key lookups, Exim was not diagnosing an error if
2047 * or *@ was used with a query-style lookup.
2049 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2051 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2052 long ago, but noone interested enough thought of it. Let's just say that
2053 the "1.1" means that there are some new functions that weren't there at
2054 some point in the past.
2056 PH/51 Error processing for expansion failure of helo_data from an smtp
2057 transport during callout processing was broken.
2059 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2060 tested/used via the -bh/-bhc/-bs options.
2062 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2063 bug, fixed in subsequent PCRE releases).
2065 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2066 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2068 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2074 SC/01 Use a glob alias rather than an array ref in eximstats generated
2075 parser. This improves both readability and performance.
2077 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2078 Don't display local sender or destination tables in eximstats unless
2079 there is data to show.
2080 Added average volumes into the eximstats top table text output.
2082 SC/03 Collect data on the number of addresses (recipients) as well
2083 as the number of messages in eximstats.
2085 TF/01 Correct an error in the documentation for the redirect router. Exim
2086 does (usually) call initgroups() when daemonizing.
2088 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2089 with consistent privilege compared to when running as a daemon.
2091 TF/03 Note in the spec that $authenticated_id is not set for local
2092 submissions from trusted users.
2094 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2095 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2097 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2098 by adding some example configuration directives to the default
2099 configuration file. A little bit of work is required to uncomment the
2100 directives and define how usernames and passwords are checked, but
2101 there is now a framework to start from.
2103 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2104 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2105 without this. I don't know how relevant this is to other LDAP libraries.
2107 PH/02 Add the verb name to the "unknown ACL verb" error.
2109 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2111 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2113 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2114 directory not expanded when it should be if an expanded home directory
2115 was set for the address (which is overridden by the transport).
2117 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2120 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2121 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2122 because it is too late at that time, and has no effect.
2124 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2125 security issue with \' (bugzilla #107). I could not use the
2126 PQescapeStringConn() function, because it needs a PGconn value as one of
2129 PH/08 When testing addresses using -bt, indicate those final addresses that
2130 are duplicates that would not cause an additional delivery. At least one
2131 person was confused, thinking that -bt output corresponded to deliveries.
2132 (Suppressing duplicates isn't a good idea as you lose the information
2133 about possibly different redirections that led to the duplicates.)
2135 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2136 systems where poll() doesn't work, in particular OS X.
2138 PH/10 Added more information to debugging output for retry time not reached.
2140 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2141 operations in malware.c.
2143 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2146 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2147 syslog_duplication set false, log lines that would normally be written
2148 both the the main log and to the reject log were not written to syslog at
2151 PH/14 In the default configuration, change the use of "message" in ACL warn
2152 statements to "add_header".
2154 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2155 not followed by a command (e.g. "seen endif").
2157 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2158 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2161 PH/17 Added extra conditions to the default value of delay_warning_condition
2165 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2166 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2167 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2170 The Auto-Submitted: and various List- headers are standardised, whereas I
2171 don't think Precedence: ever was.
2173 PH/18 Refactored debugging code in route_finduser() to show more information,
2174 in particular, the error code if getpwnam() issues one.
2176 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2177 This is apparently needed in addition to the PH/07 change above to avoid
2178 any possible encoding problems.
2180 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2181 but not after initializing Perl.
2183 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2184 output them only if debugging. By default they are written stderr,
2185 apparently, which is not desirable.
2187 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2190 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2193 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2195 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2196 authenticated or an ident call has been made. Suppress the default
2197 values for $authenticated_id and $authenticated_sender (but permit -oMai
2198 and -oMas) when testing with -bh.
2200 PH/24 Re-jigged the order of the tests in the default configuration so that the
2201 tests for valid domains and recipients precede the DNS black list and CSA
2202 tests, on the grounds that those ones are more expensive.
2204 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2205 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2206 This bug exists in every version of Exim that I still have, right back to
2209 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2210 However, an attempt to turn on case-sensitivity in a regex key by
2211 including (?-i) didn't work because the subject string was already
2212 lowercased, and the effects were non-intuitive. It turns out that a
2213 one-line patch can be used to allow (?-i) to work as expected.
2219 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2220 other effects) broke the use of negated acl sub-conditions.
2222 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2225 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2226 "Deny" causes Exim to reject the incoming connection with a 554 error.
2227 Unfortunately, if there is a major crisis, such as a disk failure,
2228 tcp-wrappers gives "deny", whereas what one would like would be some
2229 kind of temporary error. A kludge has been added to help with this.
2230 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2231 554 error is used if errno is still zero or contains ENOENT (which occurs
2232 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2235 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2237 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2238 errors. Otherwise a message that provokes a temporary error (when other
2239 messages do not) can cause a whole host to time out.
2241 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2242 addresses were routed directly to files or pipes from a redirect router.
2243 File deliveries just didn't batch; pipe deliveries might have suffered
2246 PH/06 A failure to get a lock for a hints database would erroneously always say
2247 "Failed to get write lock", even when it was really a read lock.
2249 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2250 of 0600. This has been changed to use the value of the lockfile_mode
2251 option (which defaults to 0600).
2253 PH/08 Applied small patch from the Sieve maintainer.
2255 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2256 folder from quota calculations, a direct delivery into this folder messed
2257 up the contents of the maildirsize file. This was because the regex was
2258 used only to exclude .Trash (or whatever) when the size of the mailbox
2259 was calculated. There was no check that a delivery was happening into an
2260 excluded directory. This bug has been fixed by ignoring all quota
2261 processing for deliveries into excluded directories.
2263 PH/10 Added the maildirfolder_create_regex option to appendfile.
2269 PH/01 The code for finding all the local interface addresses on a FreeBSD
2270 system running IPv6 was broken. This may well have applied to all BSD
2271 systems, as well as to others that have similar system calls. The broken
2272 code found IPv4 interfaces correctly, but gave incorrect values for the
2273 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2274 that it would not match correctly against @[] and not recognize the IPv6
2277 PH/02 The ipliteral router was not recognizing addresses of the form user@
2278 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2280 PH/03 Added disable_ipv6.
2282 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2283 decoded form, because it is most often used to construct To: headers
2284 lines in autoreplies, and the decoded form may well be syntactically
2285 invalid. However, $reply_address has leading white space removed, and all
2286 newlines turned into spaces so that the autoreply transport does not
2289 PH/05 If group was specified without a user on a router, and no group or user
2290 was specified on a transport, the group from the router was ignored.
2292 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2293 for visible compile-time settings that can be used to change these
2294 numbers, for those that want even more. Backwards compatibility with old
2295 spool files has been maintained. However, going back to a previous Exim
2296 release will lost any variables that are in spool files.
2298 PH/07 Two small changes when running in the test harness: increase delay when
2299 passing a TCP/IP connection to a new process, in case the original
2300 process has to generate a bounce, and remove special handling of
2301 127.0.0.2 (sic), which is no longer necessary.
2303 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2304 be the same on different OS.
2306 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2309 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2310 whether --show-vars was specified or not
2312 JJ/02 exipick: Added support for new ACL variable spool format introduced
2315 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2316 syntactically invalid From: or Reply-to: line, and a filter used this to
2317 generate an autoreply, and therefore failed to obtain an address for the
2318 autoreply, Exim could try to deliver to a non-existent relative file
2319 name, causing unrelated and misleading errors. What now happens is that
2320 it logs this as a hard delivery error, but does not attempt to create a
2323 PH/11 The exinext utility has a -C option for testing purposes, but although
2324 the given file was scanned by exinext itself; it wasn't being passed on
2325 when Exim was called.
2327 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2328 an end-of-file indication when reading a command response.
2330 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2331 compiled. In many other places in Exim, IPv6 addresses are always
2332 recognized, so I have changed this. It also means that IPv4 domain
2333 literals of the form [IPV4:n.n.n.n] are now always recognized.
2335 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2336 used if the router is not running as root, for example, when verifying at
2337 ACL time, or when using -bh. The debugging output from this situation was
2338 non-existent - all you got was a failure to exec. I have made two
2341 (a) Failures to set uid/gid, the current directory, or a process leader
2342 in a subprocess such as that created by queryprogram now generate
2343 suitable debugging ouput when -d is set.
2345 (b) The queryprogram router detects when it is not running as root,
2346 outputs suitable debugging information if -d is set, and then runs
2347 the subprocess without attempting to change uid/gid.
2349 PH/15 Minor change to Makefile for building test_host (undocumented testing
2352 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2353 additional section of a DNS packet that returns MX or SRV records.
2354 Instead, it always explicitly searches for A/AAAA records. This avoids
2355 major problems that occur when a DNS server includes only records of one
2356 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2357 fixed another bug: if SRV records were looked up and the corresponding
2358 address records were *not* found in the additional section, the port
2359 values from the SRV records were lost.
2361 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2362 using the correct key (the original address) when searching the retry
2363 rules in order to find which one to use for generating the retry hint.
2365 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2366 adding the default one. Similarly, if it contains a Reply-To: header, the
2367 errors_reply_to option, if set, is not used.
2369 PH/19 When calculating a retry time, Exim used to measure the "time since
2370 failure" by looking at the "first failed" field in the retry record. Now
2371 it does not use this if it is later than than the arrival time of the
2372 message. Instead it uses the arrival time. This makes for better
2373 behaviour in cases where some deliveries succeed, thus re-setting the
2374 "first failed" field. An example is a quota failure for a huge message
2375 when small messages continue to be delivered. Without this change, the
2376 "time since failure" will always be short, possible causing more frequent
2377 delivery attempts for the huge message than are intended.
2378 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2380 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2381 $1, $2, $3) because the numerical variables can be reset during some
2382 expansion items (e.g. "match"), thereby losing the authentication data.
2384 PH/21 Make -bV show the size of off_t variables so that the test suite can
2385 decide whether to run tests for quotas > 2G.
2387 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2388 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2389 filecount value is greater than 2G or if a quota value is greater than 2G
2390 on a system where the size of off_t is not greater than 4, a panic error
2393 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2394 never match. The debug and -bh output now contains an explicit error
2395 message indicating a malformed IPv4 address or mask.
2397 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2398 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2399 PH/23 above applies.
2401 PH/25 Do not write to syslog when running in the test harness. The only
2402 occasion when this arises is a failure to open the main or panic logs
2403 (for which there is an explicit test).
2405 PH/26 Added the /no_tell option to "control=freeze".
2407 PH/27 If a host name lookup failed very early in a connection, for example, if
2408 the IP address matched host_lookup and the reverse lookup yielded a name
2409 that did not have a forward lookup, an error message of the form "no IP
2410 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2411 could be logged. Now it outputs the IP address instead of "NULL".
2413 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2414 allows the sender and the authenticated sender to be set when
2415 submitting a message from within Exim. Since child_open_exim() is
2416 documented for local_scan(), the new function should be too.
2418 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2419 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2420 results in an empty string is now treated as unset.
2422 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2424 PH/31 Added sender_verify_fail logging option.
2426 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2427 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2428 tidied the source and removed it altogether.
2430 PH/33 When a queue run was abandoned because the load average was too high, a
2431 log line was always written; now it is written only if the queue_run log
2432 selector is set. In addition, the log line for abandonment now contains
2433 information about the queue run such as the pid. This is always present
2434 in "start" and "stop" lines but was omitted from the "abandon" line.
2436 PH/34 Omit spaces between a header name and the colon in the error message that
2437 is given when verify = headers_syntax fails (if there are lots of them,
2438 the message gets confusing).
2440 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2441 names, as there are now some PTR records that contain slashes. This check
2442 is only to protect against broken name servers that fall over on strange
2443 characters, so the fact that it applies to all lookups doesn't matter.
2445 PH/36 Now that the new test suite is complete, we can remove some of the
2446 special code in Exim that was needed for the old test suite. For example,
2447 sorting DNS records because real resolvers return them in an arbitrary
2448 order. The new test suite's fake resolver always returns records in the
2451 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2452 bounces) except when queue_only is set, to avoid logging races between
2453 the different processes.
2455 PH/38 Panic-die if .include specifies a non-absolute path.
2457 PH/39 A tweak to the "H" retry rule from its user.
2459 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2460 a label. They prevented compilation on older perls.
2462 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2463 a warning to be raised on newish perls.
2465 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2466 on queue. Changes to match documented behaviour of showing count of
2467 messages matching specified criteria.
2469 PH/40 Changed the default ident timeout from 30s to 5s.
2471 PH/41 Added support for the use of login_cap features, on those BSD systems
2472 that have them, for controlling the resources used by pipe deliveries.
2474 PH/42 The content-scanning code uses fopen() to create files in which to put
2475 message data. Previously it was not paying any attention to the mode of
2476 the files. Exim runs with umask(0) because the rest of the code creates
2477 files with open(), and sets the required mode explicitly. Thus, these
2478 files were ending up world-writeable. This was not a big issue, because,
2479 being within the spool directory, they were not world-accessible. I have
2480 created a function called modefopen, which takes an additional mode
2481 argument. It sets umask(777), creates the file, chmods it to the required
2482 mode, then resets the umask. All the relevant calls to fopen() in the
2483 content scanning code have been changed to use this function.
2485 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2486 to 24 hours. This avoids potential overflow problems when processing G
2487 and H retry rules. I suspect nobody ever tinkers with this value.
2489 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2491 PH/45 When the plaintext authenticator is running as a client, the server's
2492 challenges are checked to ensure they are valid base64 strings. By
2493 default, the authentication attempt is cancelled if an invalid string is
2494 received. Setting client_ignore_invalid_base64 true ignores these errors.
2495 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2496 they are received. Thus, the responses can be made to depend on the
2497 challenges. If an invalid string is ignored, an empty string is placed in
2500 PH/46 Messages that are created by the autoreply transport now contains a
2501 References: header, in accordance with RFCs 2822 and 3834.
2503 PH/47 Added authenticated_sender_force to the smtp transport.
2505 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2507 PH/49 Installed latest patch from the Sieve maintainer.
2509 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2510 was also set, the appendfile transport was unnecessarily scanning a
2511 directory of message files (e.g. for maildir delivery) to find the count
2512 of files (along with the size), even though it did not need this
2513 information. It now does the scan only if it needs to find either the
2514 size of the count of files.
2516 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2518 PH/52 Two bugs concerned with error handling when the smtp transport is
2521 (i) Exim was not creating retry information for temporary errors given
2522 for individual recipients after the DATA command when the smtp transport
2523 was used in LMTP mode. This meant that they could be retried too
2524 frequently, and not timed out correctly.
2526 (ii) Exim was setting the flag that allows error details to be returned
2527 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2528 recipients that were returned after the DATA command.
2530 PH/53 This is related to PH/52, but is more general: for any failing address,
2531 when detailed error information was permitted to be returned to the
2532 sender, but the error was temporary, then after the final timeout, only
2533 "retry timeout exceeded" was returned. Now it returns the full error as
2534 well as "retry timeout exceeded".
2536 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2537 do this, and (what is worse) MTAs that accept it.
2539 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2540 will now be deprecated.
2542 PH/56 New os.c-cygwin from the Cygwin maintainer.
2544 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2545 formats (previously only available in exim formats via -bpr, -bpru,
2546 and -bpra. Now also available in native and exiqgrep formats)
2548 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2549 with very large, slow to parse queues
2551 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2553 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2555 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2556 responses to authentication challenges, though it was showing the
2557 challenges; (ii) I've removed the CR characters from the debug output for
2560 PH/58 Allow for the insertion of a newline as well as a space when a string
2561 is turned into more than one encoded-word during RFC 2047 encoding. The
2562 Sieve code now uses this.
2564 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2565 data_4xx, lost_connection, tls_required.
2567 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2568 message was being sent as an SMTP response.
2570 PH/61 Add -l and -k options to exicyclog.
2572 PH/62 When verifying, if an address was redirected to one new address, so that
2573 verification continued, and the new address failed or deferred after
2574 having set something in $address_data, the value of $address_data was not
2575 passed back to the ACL. This was different to the case when no
2576 redirection occurred. The value is now passed back in both cases.
2578 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2579 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2580 use login_cap.h, so on its own it isn't the distinguishing feature. The
2581 new name refers directly to the setclassresources() function.
2583 PH/65 Added configuration files for NetBSD3.
2585 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2587 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2588 is preferred over IPv4.
2590 PH/68 The bounce_return_message and bounce_return_body options were not being
2591 honoured for bounces generated during the reception of non-SMTP messages.
2592 In particular, this applied to messages rejected by the ACL. This bug has
2593 been fixed. However, if bounce_return_message is true and bounce_return_
2594 body is false, the headers that are returned for a non-SMTP message
2595 include only those that have been read before the error was detected.
2596 (In the case of an ACL rejection, they have all been read.)
2598 PH/69 The HTML version of the specification is now built in a directory called
2599 spec_html instead of spec.html, because the latter looks like a path with
2600 a MIME-type, and this confuses some software.
2602 PH/70 Catch two compiler warnings in sieve.c.
2604 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2605 function verify_get_ident() calls ip_connect() to connect a socket, but
2606 if the "connect()" function timed out, ip_connect() used to close the
2607 socket. However, verify_get_ident() also closes the socket later, and in
2608 between Exim writes to the log, which may get opened at this point. When
2609 the socket was closed in ip_connect(), the log could get the same file
2610 descriptor number as the socket. This naturally causes chaos. The fix is
2611 not to close the socket in ip_connect(); the socket should be closed by
2612 the function that creates it. There was only one place in the code where
2613 this was missing, in the iplookup router, which I don't think anybody now
2614 uses, but I've fixed it anyway.
2616 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2617 well as to direct DNS lookups. Otherwise the handling of names in host
2618 lists is inconsistent and therefore confusing.
2624 PH/01 Two changes to the default runtime configuration:
2626 (1) Move the checks for relay_from_hosts and authenticated clients from
2627 after to before the (commented out) DNS black list checks.
2629 (2) Add control=submission to the relay_from_hosts and authenticated
2630 clients checks, on the grounds that messages accepted by these
2631 statements are most likely to be submissions.
2633 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2635 (1) Generate an error if the third argument for the ${prvs expansion is
2638 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2641 (3) Reset the variables that are obtained from the first argument of
2642 ${prvscheck and used in the second argument before leaving the code,
2643 because their memory is reclaimed, so using them afterwards may do
2646 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2647 one (it's much easier than Tom thought :-).
2649 (5) Because of (4), we can now allow for the use of $prvscheck_result
2650 inside the third argument.
2652 PH/03 For some reason, the default setting of PATH when running a command from
2653 a pipe transport was just "/usr/bin". I have changed it to
2656 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2657 anything to be listed in the output from -bV.
2659 PH/05 When a filter generated an autoreply, the entire To: header line was
2660 quoted in the delivery log line, like this:
2662 => >A.N.Other <ano@some.domain> <original@ddress> ...
2664 This has been changed so that it extracts the operative address. There
2665 may be more than one such address. If so, they are comma-separated, like
2668 => >ano@some.domain,ona@other.domain <original@ddress> ...
2670 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2671 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2672 not being looked up in the rDNS to get a host name, Exim was showing the
2673 IP address twice in Received: lines, even though the IP addresses were
2674 identical. For example:
2676 Received: from [1.2.3.4] (helo=[1.2.3.4])
2678 However, if the real host name was known, it was omitting the HELO data
2679 if it matched the actual IP address. This has been tidied up so that it
2680 doesn't show the same IP address twice.
2682 PH/07 When both +timestamp and +memory debugging was on, the value given by
2683 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2684 called by the debug printing, thereby overwriting the timestamp buffer.
2685 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2687 PH/08 When the original message was included in an autoreply transport, it
2688 always said "this is a copy of the message, including all the headers",
2689 even if body_only or headers_only was set. It now gives an appropriate
2692 PH/09 Applied a patch from the Sieve maintainer which:
2694 o fixes some comments
2695 o adds the (disabled) notify extension core
2696 o adds some debug output for the result of if/elsif tests
2697 o points to the current vacation draft in the documentation
2698 and documents the missing references header update
2702 o fixes a bug in processing the envelope test (when testing
2703 multiple envelope elements, the last element determinted the
2706 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2707 Electronic Mail") by including:
2709 Auto-submitted: auto-generated
2711 in the messages that it generates (bounce messages and others, such as
2712 warnings). In the case of bounce messages for non-SMTP mesages, there was
2713 also a typo: it was using "Auto_submitted" (underscore instead of
2714 hyphen). Since every message generated by Exim is necessarily in response
2715 to another message, thes have all been changed to:
2717 Auto-Submitted: auto-replied
2719 in accordance with these statements in the RFC:
2721 The auto-replied keyword:
2723 - SHOULD be used on messages sent in direct response to another
2724 message by an automatic process,
2726 - MUST NOT be used on manually-generated messages,
2728 - MAY be used on Delivery Status Notifications (DSNs) and Message
2729 Disposition Notifications (MDNs),
2731 - MUST NOT be used on messages generated by automatic or periodic
2732 processes, except for messages which are automatic responses to
2735 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2736 to the default Received: header definition.
2738 PH/12 Added log selector acl_warn_skipped (default on).
2740 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2741 variables because (a) they are in the wrong storage pool and (b) even if
2742 they were copied, it wouldn't work properly because of the caching.
2744 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2745 checking when decoding. Apparently there are clients that generate
2746 overlong encoded strings. Why am I not surprised?
2748 PH/15 If the first argument of "${if match_address" was not empty, but did not
2749 contain an "@" character, Exim crashed. Now it writes a panic log message
2750 and treats the condition as false.
2752 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2754 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2755 extension "envelope-auth". The code is finished and in agreement with
2756 other implementations, but there is no documentation so far and in fact,
2757 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2758 not changing the active code.
2760 Print executed "if" and "elsif" statements when debugging is used. This
2761 helps a great deal to understand what a filter does.
2763 Document more things not specified clearly in RFC3028. I had all this
2764 sorted out, when out of a sudden new issues came to my mind. Oops."
2766 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2769 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2770 canonical form (as documented). However, after a host name lookup from
2771 the IP address, check_host() was doing a simple string comparison with
2772 addresses acquired from the DNS when checking that the found name did
2773 have the original IP as one of its addresses. Since any found IPv6
2774 addresses are likely to be in abbreviated form, the comparison could
2775 fail. Luckily, there already exists a function for doing the comparison
2776 by converting both addresses to binary, so now that is used instead of
2777 the text comparison.
2779 PH/20 There was another similar case to PH/19, when a complete host name was
2780 given in a host list; looking up its IP address could give an abbreviated
2781 form, whereas the current host's name might or might not be abbreviated.
2782 The same fix has been applied.
2788 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2789 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2792 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2793 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2795 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2797 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2798 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2799 Also, for "Auto-Submitted", treat anything other than "no" as
2800 non-personal, in accordance with RFC 3834. (Previously it treated
2801 anything starting "auto-" as non-personal.)
2803 TF/01 The control=submission/name=... option had a problem with syntax
2804 errors if the name included a slash character. The /name= option
2805 now slurps the rest of the string, so it can include any characters
2806 but it must come last in the list of options (after /sender_retain
2809 PH/05 Some modifications to the interface to the fake nameserver for the new
2817 TK/01 Added the "success_on_redirect" address verification option. See
2818 NewStuff for rationale and an example.
2820 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2822 PH/02 Patch to exigrep to allow it to work on syslog lines.
2824 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2825 fread() to skip over the body file's header line, because in Cygwin the
2826 header line is locked and is inaccessible.
2828 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2829 co-exist for some time) to make it clear that it is the Exim ID that is
2830 referenced, not the Message-ID: header line.
2832 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2833 string_format() function, because snprintf() does not exist on all
2836 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2837 ${stat: expansion item.
2839 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2840 protocol synchronization error", to keep the pedants happy.
2842 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2843 well as for IRIX systems, when gcc is being used. See the host.c source
2846 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2848 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2851 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2852 transport and to the smtp transport in LMTP mode.
2854 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2856 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2857 run a filter in a subprocess. This could lead to confusion in subsequent
2858 lookups in the parent process. There should also be a search_tidyup() at
2859 the end of the subprocess.
2861 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2862 only if the host matched helo_try_verify_hosts, which caused the
2863 verification to occur when the EHLO/HELO command was issued. The ACL just
2864 tested the remembered result. Now, if a previous verification attempt has
2865 not happened, "verify = helo" does it there and then.
2867 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2869 TK/03 Fix log output including CR from clamd.
2871 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2872 exist provoked a memory error which could cause a segfault.
2874 PH/15 Installed PCRE 6.2
2876 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2878 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2879 of the problem. Specifically, suggested +O2 rather than +O1 for the
2882 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2884 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2885 record, the port was not correctly logged when the outgoing_port log
2886 selector was set (it logged the transort's default port).
2888 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2889 fallback_hosts, and "hosts" in the smtp transport.
2891 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2892 host errors such as "Connection refused".
2894 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2895 authentication with radiusclient 0.4.9:
2897 - Error returned from rc_read_config was caught wrongly
2898 - Username/password not passed on to radius server due to wrong length.
2900 The presumption is that some radiusclient API changes for 4.51/PH/17
2901 were not taken care of correctly. The code is still untested by me (my
2902 Linux distribution still has 0.3.2 of radiusclient), but it was
2903 contributed by a Radius user.
2905 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2906 expanding the "port" option of the smtp transport.
2908 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2909 while reading a MIME header. Thanks to Tom Hughes for a patch.
2911 PH/24 Include config.h inside local_scan.h so that configuration settings are
2914 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2915 that in an ACL for RCPT (for example), you can examine exactly what was
2918 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2919 commands, but it was not correctly comparing the address with the actual
2920 client host address. Thus, it would show the EHLO address in Received:
2921 header lines when this was not necessary.
2923 PH/27 Added the % operator to ${eval:}.
2925 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2926 it should be ignoring failures (because with -C, for example, it has lost
2927 privilege). It wasn't ignoring creation failures other than "already
2930 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2933 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2934 because an input line was too long, either on its own, or by virtue of
2935 too many continuations, the temporary file was not being removed, and the
2936 return code was incorrect.
2938 PH/31 Missing "BOOL" in function definition in filtertest.c.
2940 PH/32 Applied Sieve patches from the maintainer.
2942 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2944 PH/33 Added "verify = not_blind".
2946 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2947 Local/Makefile (with some defaults set). These are used in built scripts
2948 such as exicyclog, but they have never been used in the exim_install
2949 script (though there are many overriding facilities there). I have
2950 arranged that the exim_install script now takes note of these two
2953 PH/35 Installed configuration files for Dragonfly.
2955 PH/36 When a locally submitted message by a trusted user did not contain a
2956 From: header, and the sender address was obtained from -f or from an SMTP
2957 MAIL command, and the trusted user did not use -F to supply a sender
2958 name, $originator_name was incorrectly used when constructing a From:
2959 header. Furthermore, $originator_name was used for submission mode
2960 messages from external hosts without From: headers in a similar way,
2961 which is clearly wrong.
2963 PH/37 Added control=suppress_local_fixups.
2965 PH/38 When log_selector = +received_sender was set, and the addition of the
2966 sender made the log line's construction buffer exactly full, or one byte
2967 less than full, an overflow happened when the terminating "\n" was
2970 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2971 when the result of a list match is failure because a DNS lookup failed.
2973 PH/40 RM_COMMAND is now used in the building process.
2975 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2976 the "build-* directories that it finds.
2978 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2979 address in a domain literal was a prefix of an interface address.
2981 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2982 when verifying a sender address, unless rewrite_headers is false.
2984 PH/44 Wrote a long comment about why errors_to addresses are verified as
2985 recipients, not senders.
2987 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2988 the ratelimit ACL was added.
2990 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2992 PH/46 Added extra information about PostgreSQL errors to the error string.
2994 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2995 avoiding the need to install special zones in a real server. This is
2996 backwards compatible; if it can't find the fake resolver, it drops back.
2997 Thus, both old and new test suites can be run.
2999 TF/02 Added util/ratelimit.pl
3001 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3002 clock is set back in time.
3004 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3005 Candler <B.Candler@pobox.com>.
3007 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3008 OK for addresses that are the result of redirections.
3010 PH/48 A number of further additions for the benefit of the new test suite,
3011 including a fake gethostbyname() that interfaces to the fake DNS resolver
3014 TF/06 The fix for widen_domains has also been applied to qualify_single and
3015 search_parents which are the other dnslookup options that can cause
3018 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3021 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3023 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3024 service" if it encountered an MX record with an empty target hostname.
3025 The message is now "an MX or SRV record indicated no SMTP service".
3027 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3028 if the DNS of the sending site is misconfigured. This is quite a
3029 common situation. This change restores the behaviour of treating a
3030 helo verification defer as a failure.
3032 PH/51 If self=fail was set on a router, the bounce message did not include the
3033 actual error message.
3039 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3041 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3042 command itself ended in error, the underlying message about the transport
3043 filter timeout was being overwritten with the pipe command error. Now the
3044 underlying error message should be appended to the second error message.
3046 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3048 PH/02 Reduce the amount of output that "make" produces by default. Full output
3049 can still be requested.
3051 PH/03 The warning log line about a condition test deferring for a "warn" verb
3052 was being output only once per connection, rather than after each
3053 occurrence (because it was using the same function as for successful
3054 "warn" verbs). This seems wrong, so I have changed it.
3056 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3057 it should not have, which might have caused a crash in the right
3058 circumstances, but probably never did.
3060 PH/04 Installed a modified version of Tony Finch's patch to make submission
3061 mode fix the return path as well as the Sender: header line, and to
3062 add a /name= option so that you can make the user's friendly name appear
3065 TF/03 Added the control = fakedefer ACL modifier.
3067 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3068 Mark Lowes for thorough testing.
3070 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3072 TK/03 Merged latest SRS patch from Miles Wilton.
3074 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3075 with the definition in sysexits.h (which is #included earlier).
3076 Fortunately, Exim does not actually use EX_OK. The code used to try to
3077 preserve the sysexits.h value, by assumimg that macro definitions were
3078 scanned for macro replacements. I have been disabused of this notion,
3079 so now the code just undefines EX_OK before #including unistd.h.
3081 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3082 in the smtp transport. When a block could not be written in a single
3083 write() function, the timeout was being re-applied to each part-write.
3084 This seems wrong - if the receiver was accepting one byte at a time it
3085 would take for ever. The timeout is now adjusted when this happens. It
3086 doesn't have to be particularly precise.
3088 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3089 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3091 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3092 without a domain if the check to <postmaster@domain> fails.
3094 SC/01 Eximstats: added -xls and the ability to specify output files
3095 (patch written by Frank Heydlauf).
3097 SC/02 Eximstats: use FileHandles for outputing results.
3099 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3101 SC/04 Eximstats: fixed display of large numbers with -nvr option
3103 SC/05 Eximstats: fixed merging of reports with empty tables.
3105 SC/06 Eximstats: added the -include_original_destination flag
3107 SC/07 Eximstats: removed tabs and trailing whitespace.
3109 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3111 TK/06 MBOX spool code: Add real "From " MBOX separator line
3112 so the .eml file is really in mbox format (even though
3113 most programs do not really care). Patch from Alex Miller.
3115 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3116 The latter is generated from $received_to and is only set if the
3117 message has one envelope recipient. SA can use these headers,
3118 obviously out-of-the-box. Patch from Alex Miller.
3120 PH/08 The ${def test on a variable was returning false if the variable's
3121 value was "0", contrary to what the specification has always said!
3122 The result should be true unless the variable is empty.
3124 PH/09 The syntax error of a character other than { following "${if
3125 def:variable_name" (after optional whitespace) was not being diagnosed.
3126 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3127 accidental colon was present, for example, could give incorrect results.
3129 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3130 result is used (not including appendfile, where other changes are about
3133 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3134 This involved changing a lot of size variables from int to off_t. It
3135 should work with maildirs and everything.
3137 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3138 spamd dying while we are connected to it.
3140 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3143 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3144 function for BDB 4.3.
3146 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3148 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3149 the sources for PCRE have been re-arranged and more files are now
3152 PH/15 The code I had for printing potentially long long variables in PH/11
3153 above was not the best (it lost precision). The length of off_t variables
3154 is now inspected at build time, and an appropriate printing format (%ld
3155 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3156 to be "long long int" or "long int". This is needed for the internal
3157 formatting function string_vformat().
3159 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3160 the configuration file to be ":syslog", then the script "guesses" where
3161 the logs files are, rather than using the compiled in default. In our
3162 case the guess is not the same as the compiled default, so the script
3163 suddenly stopped working when I started to use syslog. The patch checks
3164 to see if log_file_path is "". If so, it attempts to read it from exim
3165 with no configuration file to get the compiled in version, before it
3166 falls back to the previous guessing code."
3168 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3169 implementing BATV in an Exim configuration. See NewStuff for the gory
3172 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3173 Makefile that are specific to HP-UX.
3175 PH/18 If the "use_postmaster" option was set for a recipient callout together
3176 with the "random" option, the postmaster address was used as the MAIL
3177 FROM address for the random test, but not for the subsequent recipient
3178 test. It is now used for both.
3180 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3181 patch removes a few documentation additions to RFC 3028, because the
3182 latest draft now contains them. It adds the new en;ascii-case comparator
3183 and a new error check for 8bit text in MIME parts. Comparator and
3184 require names are now matched exactly. I enabled the subaddress
3185 extension, but it is not well tested yet (read: it works for me)."
3187 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3188 rework some of the code of TK/09 above to avoid the hardwired use of
3189 "%lld" and "long long". Replaced the call to snprintf() with a call to
3192 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3193 records point to non-existent hosts", "retry timeout exceeded", and
3194 "retry time not reached for any host after a long failure period".
3196 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3197 experimental DomainKeys support:
3199 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3200 (2) On an error such as an illegally used "control", the wrong name for
3201 the control was given.
3203 These problems did NOT occur unless DomainKeys support was compiled.
3205 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3207 PH/24 Added ${if match_ip condition.
3209 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3210 databases so that it will be absolutely obvious if a crash occurs in the
3211 DB library. This is a regular occurrence (often caused by mis-matched
3214 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3215 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3216 picked up on a user's system that detects such things. There doesn't seem
3217 to be a gcc warning option for this - only an attribute that has to be
3218 put on the function's prototype. It seems that in Fedora Core 4 they have
3219 set this on a number of new functions. No doubt there will be more in due
3222 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3223 specify a transport. However, if an address that was verified by such a
3224 router was the subject of a callout, Exim crashed because it tried to
3225 read the rcpt_include_affixes from the non-existent transport. Now it
3226 just assumes that the setting of that option is false. This bug was
3227 introduced by 4.51/PH/31.
3229 PH/28 Changed -d+all to exclude +memory, because that information is very
3230 rarely of interest, but it makes the output a lot bigger. People tend to
3231 do -d+all out of habit.
3233 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3234 code in os-type was giving problems when libc.so lives in lib64, like on
3237 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3238 aren't the modern standard, and it seems that some systems' include files
3239 don't always have them. Exim was already checking for some of the newer
3240 ones like T_AAAA, and defining it itself. I've added checks for all the
3241 record types that Exim uses.
3243 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3244 not automatically generating a new one, as it is supposed to. This
3245 prevented TLS from working. If the file did exist, but contained invalid
3246 data, a new version was generated, as expected. It was only the case of a
3247 non-existent file that was broken.
3249 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3250 with a change in libdomainkeys > 0.64.
3252 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3253 from DNS. If the selector record carries the flag, it now has
3254 precedence over the domain-wide flag.
3256 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3258 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3259 the use of a port name that isn't defined in /etc/services) occurred, the
3260 message was deferred as in a normal delivery, and thus remained on the
3261 spool, instead of being failed because of the mua_wrapper setting. This
3262 is now fixed, and I tidied up some of the mua_wrapper messages at the
3265 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3266 the messages in a hash of arrays rather than using individual hashes.
3267 This is a bit cleaner and results in dramatic memory savings, albeit
3268 at a slight CPU cost.
3270 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3271 as requested by Marc Sherman.
3273 SC/10 Eximstats: added histograms for user specified patterns as requested
3276 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3278 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3279 fopen() in the content-scanning modules that did not already have it.
3285 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3286 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3288 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3290 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3292 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3293 to test Sieve filters that use "vacation".
3295 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3296 that changes the way the GnuTLS parameters are stored in the cache file.
3297 The new format can be generated externally. For backward compatibility,
3298 if the data in the cache doesn't make sense, Exim assumes it has read an
3299 old-format file, and it generates new data and writes a new file. This
3300 means that you can't go back to an older release without removing the
3303 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3304 work if there are any delivery delays because "one_time" forces the
3305 parent to be marked "delivered", so its unseen clone is never tried
3306 again. For this reason, Exim now forbids the simultaneous setting of
3309 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3310 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3311 entry if you want to know the details. The fix, however, neglected to
3312 consider the case when local delivery batching is involved. The test for
3313 "previously delivered" was not happening when checking to see if an
3314 address could be batched with a previous (undelivered) one; under
3315 certain circumstances this could lead to multiple deliveries to the same
3318 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3319 in its include files, and this causes problems building Exim.
3321 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3322 header_syntax) but Exim was just ignoring anything given after a slash.
3323 In particular, this caused confusion with an attempt to use "verify =
3324 reverse_host_lookup/defer_ok". An error is now given when options are
3325 supplied for verify items that do not have them. (Maybe reverse_host_
3326 lookup should have a defer_ok option, but that's a different point.)
3328 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3329 defined by RFC 821) to 2048, because there were problems with some AUTH
3330 commands, and RFC 1869 says the size should be increased for extended
3331 SMTP commands that take arguments.
3333 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3336 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3337 "unknown" error; now it says that the functionality isn't in the binary.
3339 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3340 an address' error message when a string expansion fails (syntax or
3341 whatever). Otherwise the password may appear in the log. Following change
3342 PH/42 below, there is no longer a chance of it appearing in a bounce
3345 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3347 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3348 of its addresses was ever tried. (Bugzilla bug #2.)
3350 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3351 the result incorrectly in the debug output. (It correctly added a newline
3352 to what was transported.)
3354 TF/01 Added $received_time.
3356 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3357 commented out examples of how to interface to a virus scanner and to
3358 SpamAssassin. Also added commented examples of av_scanner and
3359 spamd_address settings.
3361 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3362 and controls are allowed in which ACLs. There were a couple of minor
3363 errors. Some of the entries in the conditions table (which is a table of
3364 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3365 negation of where the condition IS allowed.
3367 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3369 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3370 header file does not have a version number, so I've had to invent a new
3371 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3372 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3373 radiusclient), but it was contributed by a Radius user.
3375 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3376 files or directories, for OpenSSL.
3378 PH/19 When an Exim process that is running as root has to create an Exim log
3379 file, it does so in a subprocess that runs as exim:exim so as to get the
3380 ownership right at creation (otherwise, other Exim processes might see
3381 the file with the wrong ownership). There was no test for failure of this
3382 fork() call, which would lead to the process getting stuck as it waited
3383 for a non-existent subprocess. Forks do occasionally fail when resources
3384 run out. I reviewed all the other calls to fork(); they all seem to check
3387 PH/20 When checking for unexpected SMTP input at connect time (before writing
3388 the banner), Exim was not dealing correctly with a non-positive return
3389 from the read() function. If the client had disconnected by this time,
3390 the result was a log entry for a synchronization error with an empty
3391 string after "input=" when read() returned zero. If read() returned -1
3392 (an event I could not check), uninitialized data bytes were printed.
3393 There were reports of junk text (parts of files, etc) appearing after
3396 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3398 PH/22 Added support for macro redefinition, and (re)definition in between
3399 driver and ACL definitions.
3401 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3402 forgetting to use the resulting value; it was using the unexpanded value.
3404 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3405 hadn't been configured. The fix is from Juergen Kreileder, who
3406 understands it better than I do:
3408 "Here's what I see happening with three configured cyrus_sasl
3409 authenticators configured (plain, login, cram-md5):
3411 On startup auth_cyrus_sasl_init() gets called for each of these.
3412 This means three calls to sasl_listmech() without a specified mech_list.
3413 => SASL tests which mechs of all available mechs actually work
3414 => three warnings about OTP not working
3415 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3417 With the patch, sasl_listmech() also gets called three times. But now
3418 SASL's mech_list option is set to the server_mech specified in the the
3419 authenticator. Or in other words, the answer from sasl_listmech()
3420 gets limited to just the mech you're testing for (which is different
3422 => the return list contains just 'plain' or 'login', 'cram-md5' or
3423 nothing depending on the value of ob->server_mech.
3425 I've just tested the patch: Authentication still works fine,
3426 unavailable mechs specified in the exim configuration are still
3427 caught, and the auth.log warnings about OTP are gone."
3429 PH/25 When debugging is enabled, the contents of the command line are added
3430 to the debugging output, even when log_selector=+arguments is not
3433 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3434 answer is "GNU", and only if the return is "GNU/something" is the answer
3437 PH/27 $acl_verify_message is now set immediately after the failure of a
3438 verification in an ACL, and so is available in subsequent modifiers. In
3439 particular, the message can be preserved by coding like this:
3441 warn !verify = sender
3442 set acl_m0 = $acl_verify_message
3444 Previously, $acl_verify_message was set only while expanding "message"
3445 and "log_message" when a very denied access.
3447 PH/28 Modified OS/os.c-Linux with
3449 -#ifndef OS_LOAD_AVERAGE
3450 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3452 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3453 nomenclature these days.)
3455 PH/29 Installed patch from the Sieve maintainer that adds the options
3456 sieve_useraddress and sieve_subaddress to the redirect router.
3458 PH/30 In these circumstances:
3459 . Two addresses routed to the same list of hosts;
3460 . First host does not offer TLS;
3461 . First host accepts first address;
3462 . First host gives temporary error to second address;
3463 . Second host offers TLS and a TLS session is established;
3464 . Second host accepts second address.
3465 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3466 and peerdn, if requested) that were in fact used only for the second
3469 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3470 attention to any local part prefix or suffix that was matched by the
3471 router that accepted the address. It now behaves in the same way as it
3472 does for delivery: the affixes are removed from the local part unless
3473 rcpt_include_affixes is set on the transport.
3475 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3476 timeout during the DATA phase of an incoming message.
3478 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3479 applied a patch sanctioned by the Sieve maintainer.
3481 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3482 the uid or gid is negative. A case of a negative gid caused this to be
3483 noticed. The fix allows for either to be negative.
3485 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3486 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3489 PH/36 The change PH/12 above was broken. Fixed it.
3491 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3492 the grounds that routing the same address twice would always produce the
3493 same answer. This might have been true once, but it is certainly no
3494 longer true now. Routing a child address may depend on the previous
3495 routing that produced that child. Some complicated redirection strategies
3496 went wrong when messages had multiple recipients, and made Exim's
3497 behaviour dependent on the order in which the addresses were given.
3499 I have moved the duplicate checking until after the routing is complete.
3500 Exim scans the addresses that are assigned to local and remote
3501 transports, and removes any duplicates. This means that more work will be
3502 done, as duplicates will always all be routed, but duplicates are
3503 presumably rare, so I don't expect this is of any significance.
3505 For deliveries to pipes, files, and autoreplies, the duplicate checking
3506 still happens during the routing process, since they are not going to be
3509 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3510 It corrects a timeout issue with spamd. This is Ian's comment: "The
3511 background is that sometimes spamd either never reads data from a
3512 connection it has accepted, or it never writes response data. The exiscan
3513 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3514 blindly assumes that writes won't block so it may never time out."
3516 PH/39 Allow G after quota size as well as K and M.
3518 PH/40 The value set for $authenticated_id in an authenticator may not contain
3519 binary zeroes or newlines because the value is written to log lines and
3520 to spool files. There was no check on this. Now the value is run through
3521 the string_printing() function so that such characters are converted to
3522 printable escape sequences.
3524 PH/41 $message_linecount is a new variable that contains the total number of
3525 lines in the message. Compare $body_linecount, which is the count for the
3528 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3529 bounce and delay warning messages, except in certain special cases, which
3532 (a) An SMTP error message from a remote host;
3533 (b) A message specified in a :fail: redirection;
3534 (c) A message specified in a "fail" command in a system filter;
3535 (d) A message specified in a FAIL return from the queryprogram router;
3536 (e) A message specified by the cannot_route_message router option.
3538 In these cases only, Exim does include the error details in bounce and
3539 warning messages. There are also a few cases where bland messages such
3540 as "unrouteable address" or "local delivery error" are given.
3542 PH/43 $value is now also set for the "else" part of a ${run expansion.
3544 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3545 being worked on, but at least Exim now implements the latest version to
3548 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3549 process to complete was treated as a delivery failure, a timeout while
3550 writing the message to the pipe was logged, but erroneously treated as a
3551 successful delivery. Such timeouts include transport filter timeouts. For
3552 consistency with the overall process timeout, these timeouts are now
3553 treated as errors, giving rise to delivery failures by default. However,
3554 there is now a new Boolean option for the pipe transport called
3555 timeout_defer, which, if set TRUE, converts the failures into defers for
3556 both kinds of timeout. A transport filter timeout is now identified in
3559 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3560 systems where "make" and "gmake" are different, calling "gmake" at top
3561 level broke things. I've arranged for the value of $(MAKE) to be passed
3562 from the Makefile to this script so that it can call the same version of
3566 A note about Exim versions 4.44 and 4.50
3567 ----------------------------------------
3569 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3570 changes of various kinds. As a consequence, a big documentation update was
3571 needed. This delayed the release for rather longer than seemed good, especially
3572 in the light of a couple of (minor) security issues. Therefore, the changes
3573 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3574 release. So 4.44 and 4.50 are in effect two different branches that both start
3577 I have left the 4.50 change log unchanged; it contains all the changes since
3578 4.43. The change log for 4.44 is below; many of its items are identical to
3579 those for 4.50. This seems to be the most sensible way to preserve the
3580 historical information.
3586 1. Minor wording change to the doc/README.SIEVE file.
3588 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3589 computation of the current number of files was incorrect.
3591 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3592 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3595 4. Give more explanation in the error message when the command for a transport
3596 filter fails to execute.
3598 5. There are several places where Exim runs a non-Exim command in a
3599 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3600 was being done only for the command run by the queryprogram router. It is
3601 now done for all such subprocesses. The other cases are: ${run, transport
3602 filters, and the commands run by the lmtp and pipe transports.
3604 6. Added CONFIGURE_GROUP build-time option.
3606 7. Some older OS have a limit of 256 on the maximum number of file
3607 descriptors. Exim was using setrlimit() to set 1000 as a large value
3608 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3609 systems. I've change it so that if it can't get 1000, it tries for 256.
3611 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3612 was an oversight, and furthermore, ever since the addition of extra
3613 controls (e.g. 4.43/32), the checks on when to allow different forms of
3614 "control" were broken. There should now be diagnostics for all cases when a
3615 control that does not make sense is encountered.
3617 9. Added the /retain_sender option to "control=submission".
3619 10. $recipients is now available in the predata ACL (oversight).
3621 11. Tidy the search cache before the fork to do a delivery from a message
3622 received from the command line. Otherwise the child will trigger a lookup
3623 failure and thereby defer the delivery if it tries to use (for example) a
3624 cached ldap connection that the parent has called unbind on.
3626 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3627 of $address_data from the recipient verification was clobbered by the
3628 sender verification.
3630 13. The value of address_data from a sender verification is now available in
3631 $sender_address_data in subsequent conditions in the ACL statement.
3633 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3635 15. Added a new option "connect=<time>" to callout options, to set a different
3638 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3639 was its contents. (It was OK if the option was not defined at all.)
3641 17. A "Completed" log line is now written for messages that are removed from
3642 the spool by the -Mrm option.
3644 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3645 information about exactly what failed.
3647 19. Added -dd to debug only the daemon process.
3649 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3650 handles timeouts, both on the server side and network timeouts. Renamed the
3651 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3653 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3654 It is now set to "smtps".
3656 22. $host_address is now set to the target address during the checking of
3657 ignore_target_hosts.
3659 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3660 being passed; this would have caused $sender_host_name to have been used if
3661 matching the list had actually called for a host name (not very likely,
3662 since this list is usually IP addresses). A host name is now passed as
3665 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3666 code that specifies a non-restarting handler (typically sigaction() in
3667 modern systems) in an attempt to fix a rare and obscure crash bug.
3669 25. Narrowed the window for a race in the daemon that could cause it to ignore
3670 SIGCHLD signals. This is not a major problem, because they are used only to
3671 wake it up if nothing else does.
3673 26. A malformed maildirsize file could cause Exim to calculate negative values
3674 for the mailbox size or file count. Odd effects could occur as a result.
3675 The maildirsize information is now recalculated if the size or filecount
3678 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3679 support for a long time. Removed HAVE_SYS_VFS_H.
3681 28. Installed the latest version of exipick from John Jetmore.
3683 29. In an address list, if the pattern was not a regular expression, an empty
3684 subject address (from a bounce message) matched only if the pattern was an
3685 empty string. Non-empty patterns were not even tested. This was the wrong
3686 because it is perfectly reasonable to use an empty address as part of a
3687 database query. An empty address is now tested by patterns that are
3688 lookups. However, all the other forms of pattern expect the subject to
3689 contain a local part and a domain, and therefore, for them, an empty
3690 address still always fails if the pattern is not itself empty.
3692 30. Exim went into a mad DNS loop when attempting to do a callout where the
3693 host was specified on an smtp transport, and looking it up yielded more
3694 than one IP address.
3696 31. Re-factored the code for checking spool and log partition space into a
3697 function that finds that data and another that does the check. The former
3698 is then used to implement four new variables: $spool_space, $log_space,
3699 $spool_inodes, and $log_inodes.
3701 32. The RFC2047 encoding function was originally intended for short strings
3702 such as real names; it was not keeping to the 75-character limit for
3703 encoded words that the RFC imposes. It now respects the limit, and
3704 generates multiple encoded words if necessary. To be on the safe side, I
3705 have increased the buffer size for the ${rfc2047: expansion operator from
3708 33. It is now permitted to omit both strings after an "if" condition; if the
3709 condition is true, the result is "true". As before, when the second string
3710 is omitted, a false condition yields an empty string. This makes it less
3711 cumbersome to write custom ACL and router conditions.
3713 34. Failure to deliver a bounce message always caused it to be frozen, even if
3714 there was an errors_to setting on the router. The errors_to setting is now
3717 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3718 canonical form (fully expanded) before being placed in
3719 $sender_host_address.
3721 36. The table in the code that translates DNS record types into text (T_A to
3722 "A" for instance) was missing entries for NS and CNAME. It is just possible
3723 that this could have caused confusion if both these types were looked up
3724 for the same domain, because the text type is used as part of Exim's
3725 per-process caching. But the chance of anyone hitting this buglet seems
3728 37. The dnsdb lookup has been extended in a number of ways.
3730 (1) There is a new type, "zns", which walks up the domain tree until it
3731 finds some nameserver records. It should be used with care.
3733 (2) There is a new type, "mxh", which is like "mx" except that it returns
3734 just the host names, not the priorities.
3736 (3) It is now possible to give a list of domains (or IP addresses) to be
3737 looked up. The behaviour when one of the lookups defers can be
3738 controlled by a keyword.
3740 (4) It is now possible to specify the separator character for use when
3741 multiple records are returned.
3743 38. The dnslists ACL condition has been extended: it is now possible to supply
3744 a list of IP addresses and/or domains to be looked up in a particular DNS
3747 39. Added log_selector=+queue_time_overall.
3749 40. When running the queue in the test harness, wait just a tad after forking a
3750 delivery process, to get repeatability of debugging output.
3752 41. Include certificate and key file names in error message when GnuTLS fails
3753 to set them up, because the GnuTLS error message doesn't include the name
3754 of the failing file when there is a problem reading it.
3756 42. Allow both -bf and -bF in the same test run.
3758 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3760 44. The "Exiscan patch" is now merged into the mainline Exim source.
3762 45. Sometimes the final signoff response after QUIT could fail to get
3763 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3764 < 0 before doing a fflush(). This bug looks as though it goes back to the
3765 introduction of TLS in release 3.20, but "sometimes" must have been rare
3766 because the tests only now provoked it.
3768 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3769 (this can affect the format of dates).
3771 47. exim_tidydb, when checking for the continued existence of a message for
3772 which it has found a message-specific retry record, was not finding
3773 messages that were in split spool directories. Consequently, it was
3774 deleting retry records that should have stayed in existence.
3776 48. Steve fixed some bugs in eximstats.
3778 49. The SPA authentication driver was not abandoning authentication and moving
3779 on to the next authenticator when an expansion was forced to fail,
3780 contradicting the general specification for all authenticators. Instead it
3781 was generating a temporary error. It now behaves as specified.
3783 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3784 (the order specifies the preference for clients). The order is now AES256,
3785 AES128, 3DES, ARCFOUR128.
3787 51. Small patch to Sieve code - explicitly set From: when generating an
3790 52. Exim crashed if a remote delivery caused a very long error message to be
3791 recorded - for instance if somebody sent an entire SpamAssassin report back
3792 as a large number of 550 error lines. This bug was coincidentally fixed by
3793 increasing the size of one of Exim's internal buffers (big_buffer) that
3794 happened as part of the Exiscan merge. However, to be on the safe side, I
3795 have made the code more robust (and fixed the comments that describe what
3798 53. Now that there can be additional text after "Completed" in log lines (if
3799 the queue_time_overall log selector is set), a one-byte patch to exigrep
3800 was needed to allow it to recognize "Completed" as not the last thing in
3803 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3804 patch that reportedly fixes this has been added. I am not expert enough to
3805 create a test for it. This is what the patch creator wrote:
3807 "I found a little strange behaviour of ldap code when working with
3808 Windows 2003 AD Domain, where users was placed in more than one
3809 Organization Units. When I tried to give exim partial DN, the exit code
3810 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3811 But simultaneously result of request was absolutely normal ldap result,
3812 so I produce this patch..."
3814 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3815 REFERENCE, so I have modified the code to exclude the patch when that macro
3818 55. Some experimental protocols are using DNS PTR records for new purposes. The
3819 keys for these records are domain names, not reversed IP addresses. The
3820 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3821 leaves it alone. Component reversal etc. now happens only for IP addresses.
3824 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3826 57. Double the size of the debug message buffer (to 2048) so that more of very
3827 long debug lines gets shown.
3829 58. The exicyclog utility now does better if the number of log files to keep
3830 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3832 59. Two changes related to the smtp_active_hostname option:
3834 (1) $smtp_active_hostname is now available as a variable.
3835 (2) The default for smtp_banner uses $smtp_active_hostname instead
3836 of $primary_hostname.
3838 60. The host_aton() function is supposed to be passed a string that is known
3839 to be a valid IP address. However, in the case of IPv6 addresses, it was
3840 not checking this. This is a hostage to fortune. Exim now panics and dies
3841 if the condition is not met. A case was found where this could be provoked
3842 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3843 components; fortuitously, this particular loophole had already been fixed
3844 by change 4.50/55 above.
3846 If there are any other similar loopholes, the new check in host_aton()
3847 itself should stop them being exploited. The report I received stated that
3848 data on the command line could provoke the exploit when Exim was running as
3849 exim, but did not say which command line option was involved. All I could
3850 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3851 running as the user.
3854 61. There was a buffer overflow vulnerability in the SPA authentication code
3855 (which came originally from the Samba project). I have added a test to the
3856 spa_base64_to_bits() function which I hope fixes it.
3859 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3860 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3862 63. The daemon start-up calls getloadavg() while still root for those OS that
3863 need the first call to be done as root, but it missed one case: when
3864 deliver_queue_load_max is set with deliver_drop_privilege. This is
3865 necessary for the benefit of the queue runner, because there is no re-exec
3866 when deliver_drop_privilege is set.
3868 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3869 This has been fixed.
3871 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3872 was in use, was not putting the data itself into the right store pool;
3873 consequently, it could be overwritten for a subsequent message in the same
3874 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3877 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3879 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3880 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3881 the function were treating the return as a boolean value, which happened to
3882 work because 0=false and not-0=true, but is not correct code.
3884 68. The host_aton() function was not handling scoped IPv6 addresses (those
3885 with, for example, "%eth0" on the end) correctly.
3887 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3888 negated items (that is, ~something) in unsigned ints. Some compilers
3889 apparently mutter when there is no cast.
3891 70. If an address verification called from an ACL failed, and did not produce a
3892 user-specific message (i.e. there was only a "system" message), nothing was
3893 put in $acl_verify_message. In this situation, it now puts the system
3896 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3897 session; change 4.31/43 added the unwanted input to the log line - except
3898 that it did not do this in the start of session case. It now does.
3900 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3901 This is wrong and can cause the other end to generate a synchronization
3902 error if it is another Exim or anything else that does the synchronization
3903 check. A QUIT command is no longer sent after a timeout.
3905 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3906 during host lookups.
3908 74. The defer_ok option of callout verification was not working if it was used
3909 when verifying addresses in header lines, that is, for this case:
3911 verify = header_sender/callout=defer_ok
3913 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3914 those file descriptors could be used for SMTP connections. If anything
3915 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3916 could be sent to the SMTP client, causing chaos. The daemon now opens
3917 stdin, stdout, and stderr to /dev/null when it puts itself into the
3920 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3921 log by default. The user can override this with suitable Perl magic.
3923 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3924 the log message when discard triggers, was not working for the DATA ACL or
3925 for the non-SMTP ACL.
3927 78. Error message wording change in sieve.c.
3929 79. If smtp_accept_max_per_host was set, the number of connections could be
3930 restricted to fewer than expected, because the daemon was trying to set up
3931 a new connection before checking whether the processes handling previous
3932 connections had finished. The check for completed processes is now done
3933 earlier. On busy systems, this bug wouldn't be noticed because something
3934 else would have woken the daemon, and it would have reaped the completed
3937 80. If a message was submitted locally by a user whose login name contained one
3938 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3939 It caused a spool format error. I have fixed the spool reading code. A
3940 related problem was that the "from" clause in the Received: line became
3941 illegal because of the space(s). It is now covered by ${quote_local_part.
3943 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3946 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3948 83. Patch from Sieve maintainer to fix unterminated string problem in
3949 "vacation" handling.
3951 84. Some minor changes to the Linux configuration files to help with other
3952 OS variants using glibc.
3954 85. One more patch for Sieve to update vacation handling to latest spec.
3957 ----------------------------------------------------
3958 See the note above about the 4.44 and 4.50 releases.
3959 ----------------------------------------------------
3965 1. Change 4.43/35 introduced a bug that caused file counts to be
3966 incorrectly computed when quota_filecount was set in an appendfile
3969 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3970 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3973 3. Give more explanation in the error message when the command for a transport
3974 filter fails to execute.
3976 4. There are several places where Exim runs a non-Exim command in a
3977 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3978 was being done only for the command run by the queryprogram router. It is
3979 now done for all such subprocesses. The other cases are: ${run, transport
3980 filters, and the commands run by the lmtp and pipe transports.
3982 5. Some older OS have a limit of 256 on the maximum number of file
3983 descriptors. Exim was using setrlimit() to set 1000 as a large value
3984 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3985 systems. I've change it so that if it can't get 1000, it tries for 256.
3987 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3988 was an oversight, and furthermore, ever since the addition of extra
3989 controls (e.g. 4.43/32), the checks on when to allow different forms of
3990 "control" were broken. There should now be diagnostics for all cases when a
3991 control that does not make sense is encountered.
3993 7. $recipients is now available in the predata ACL (oversight).
3995 8. Tidy the search cache before the fork to do a delivery from a message
3996 received from the command line. Otherwise the child will trigger a lookup
3997 failure and thereby defer the delivery if it tries to use (for example) a
3998 cached ldap connection that the parent has called unbind on.
4000 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4001 of $address_data from the recipient verification was clobbered by the
4002 sender verification.
4004 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4005 was its contents. (It was OK if the option was not defined at all.)
4007 11. A "Completed" log line is now written for messages that are removed from
4008 the spool by the -Mrm option.
4010 12. $host_address is now set to the target address during the checking of
4011 ignore_target_hosts.
4013 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4014 being passed; this would have caused $sender_host_name to have been used if
4015 matching the list had actually called for a host name (not very likely,
4016 since this list is usually IP addresses). A host name is now passed as
4019 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4020 code that specifies a non-restarting handler (typically sigaction() in
4021 modern systems) in an attempt to fix a rare and obscure crash bug.
4023 15. Narrowed the window for a race in the daemon that could cause it to ignore
4024 SIGCHLD signals. This is not a major problem, because they are used only to
4025 wake it up if nothing else does.
4027 16. A malformed maildirsize file could cause Exim to calculate negative values
4028 for the mailbox size or file count. Odd effects could occur as a result.
4029 The maildirsize information is now recalculated if the size or filecount
4032 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4033 support for a long time. Removed HAVE_SYS_VFS_H.
4035 18. Updated exipick to current release from John Jetmore.
4037 19. Allow an empty sender to be matched against a lookup in an address list.
4038 Previously the only cases considered were a regular expression, or an
4041 20. Exim went into a mad DNS lookup loop when doing a callout where the
4042 host was specified on the transport, if the DNS lookup yielded more than
4045 21. The RFC2047 encoding function was originally intended for short strings
4046 such as real names; it was not keeping to the 75-character limit for
4047 encoded words that the RFC imposes. It now respects the limit, and
4048 generates multiple encoded words if necessary. To be on the safe side, I
4049 have increased the buffer size for the ${rfc2047: expansion operator from
4052 22. Failure to deliver a bounce message always caused it to be frozen, even if
4053 there was an errors_to setting on the router. The errors_to setting is now
4056 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4057 canonical form (fully expanded) before being placed in
4058 $sender_host_address.
4060 24. Updated eximstats to version 1.33
4062 25. Include certificate and key file names in error message when GnuTLS fails
4063 to set them up, because the GnuTLS error message doesn't include the name
4064 of the failing file when there is a problem reading it.
4066 26. Expand error message when OpenSSL has problems setting up cert/key files.
4069 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4070 (this can affect the format of dates).
4072 28. exim_tidydb, when checking for the continued existence of a message for
4073 which it has found a message-specific retry record, was not finding
4074 messages that were in split spool directories. Consequently, it was
4075 deleting retry records that should have stayed in existence.
4077 29. eximstats updated to version 1.35
4078 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4079 1.35 - bugfix such that pie charts by volume are generated correctly
4081 30. The SPA authentication driver was not abandoning authentication and moving
4082 on to the next authenticator when an expansion was forced to fail,
4083 contradicting the general specification for all authenticators. Instead it
4084 was generating a temporary error. It now behaves as specified.
4086 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4087 (the order specifies the preference for clients). The order is now AES256,
4088 AES128, 3DES, ARCFOUR128.
4090 31. Small patch to Sieve code - explicitly set From: when generating an
4093 32. Exim crashed if a remote delivery caused a very long error message to be
4094 recorded - for instance if somebody sent an entire SpamAssassin report back
4095 as a large number of 550 error lines. This bug was coincidentally fixed by
4096 increasing the size of one of Exim's internal buffers (big_buffer) that
4097 happened as part of the Exiscan merge. However, to be on the safe side, I
4098 have made the code more robust (and fixed the comments that describe what
4101 33. Some experimental protocols are using DNS PTR records for new purposes. The
4102 keys for these records are domain names, not reversed IP addresses. The
4103 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4104 leaves it alone. Component reversal etc. now happens only for IP addresses.
4107 34. The host_aton() function is supposed to be passed a string that is known
4108 to be a valid IP address. However, in the case of IPv6 addresses, it was
4109 not checking this. This is a hostage to fortune. Exim now panics and dies
4110 if the condition is not met. A case was found where this could be provoked
4111 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4112 components; fortuitously, this particular loophole had already been fixed
4113 by change 4.50/55 or 4.44/33 above.
4115 If there are any other similar loopholes, the new check in host_aton()
4116 itself should stop them being exploited. The report I received stated that
4117 data on the command line could provoke the exploit when Exim was running as
4118 exim, but did not say which command line option was involved. All I could
4119 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4120 running as the user.
4123 35. There was a buffer overflow vulnerability in the SPA authentication code
4124 (which came originally from the Samba project). I have added a test to the
4125 spa_base64_to_bits() function which I hope fixes it.
4128 36. The daemon start-up calls getloadavg() while still root for those OS that
4129 need the first call to be done as root, but it missed one case: when
4130 deliver_queue_load_max is set with deliver_drop_privilege. This is
4131 necessary for the benefit of the queue runner, because there is no re-exec
4132 when deliver_drop_privilege is set.
4134 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4135 was in use, was not putting the data itself into the right store pool;
4136 consequently, it could be overwritten for a subsequent message in the same
4137 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4140 38. Sometimes the final signoff response after QUIT could fail to get
4141 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4142 < 0 before doing a fflush(). This bug looks as though it goes back to the
4143 introduction of TLS in release 3.20, but "sometimes" must have been rare
4144 because the tests only now provoked it.
4150 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4151 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4152 Now it returns an integer. A number of calls were still expecting a T/F
4153 return. Fortuitously, in all cases, the tests worked in OK situations,
4154 which is the norm. However, things would have gone wrong on any write
4155 failures on the smtp file descriptor. This function is used when sending
4156 messages over SMTP and also when doing verify callouts.
4158 2. When Exim is called to do synchronous delivery of a locally submitted
4159 message (the -odf or -odi options), it no longer closes stderr before doing
4162 3. Implemented the mua_wrapper option.
4164 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4166 5. Implemented the functions header_remove(), header_testname(),
4167 header_add_at_position(), and receive_remove_recipient(), and exported them
4170 6. If an ACL "warn" statement specified the addition of headers, Exim already
4171 inserted X-ACL-Warn: at the start if there was no header name. However, it
4172 was not making this test for the second and subsequent header lines if
4173 there were newlines in the string. This meant that an invalid header could
4174 be inserted if Exim was badly configured.
4176 7. Allow an ACL "warn" statement to add header lines at the start or after all
4177 the Received: headers, as well as at the end.
4179 8. Added the rcpt_4xx retry error code.
4181 9. Added postmaster_mailfrom=xxx to callout verification option.
4183 10. Added mailfrom=xxxx to the callout verification option, for verify=
4186 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4187 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4189 12. Inserted some casts to stop certain compilers warning when using pointer
4190 differences as field lengths or precisions in printf-type calls (mostly
4191 affecting debugging statements).
4193 13. Added optional readline() support for -be (dynamically loaded).
4195 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4196 same clock tick as a message's arrival, so that its received time was the
4197 same as the "first fail" time on the retry record, and that message
4198 remained on the queue past the ultimate address timeout, every queue runner
4199 would try a delivery (because it was past the ultimate address timeout) but
4200 after another failure, the ultimate address timeout, which should have then
4201 bounced the address, did not kick in. This was a "< instead of <=" error;
4202 in most cases the first failure would have been in the next clock tick
4203 after the received time, and all would be well.
4205 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4206 being recognized when the domain list was tested by the match_domain
4207 condition in an expansion string.
4209 16. Added the ${str2b64: operator.
4211 17. Exim was always calling setrlimit() to set a large limit for the number of
4212 processes, without checking whether the existing limit was already
4213 adequate. (It did check for the limit on file descriptors.) Furthermore,
4214 errors from getrlimit() and setrlimit() were being ignored. Now they are
4215 logged to the main and panic logs, but Exim does carry on, to try to do its
4216 job under whatever limits there are.
4218 18. Imported PCRE 5.0.
4220 19. Trivial typo in log message " temporarily refused connection" (the leading
4223 20. If the log selector return_path_on_delivery was set and an address was
4224 redirected to /dev/null, the delivery process crashed because it assumed
4225 that a return path would always be set for a "successful" delivery. In this
4226 case, the whole delivery is bypassed as an optimization, and therefore no
4229 21. Internal re-arrangement: the function for sending a challenge and reading
4230 a response while authentication was assuming a zero-terminated challenge
4231 string. It's now changed to take a pointer and a length, to allow for
4232 binary data in such strings.
4234 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4236 23. Exim was not respecting finduser_retries when seeking the login of the
4237 uid under which it was called; it was always trying 10 times. (The default
4238 setting of finduser_retries is zero.) Also, it was sleeping after the final
4239 failure, which is pointless.
4241 24. Implemented tls_on_connect_ports.
4243 25. Implemented acl_smtp_predata.
4245 26. If the domain in control=submission is set empty, Exim assumes that the
4246 authenticated id is a complete email address when it generates From: or
4247 Sender: header lines.
4249 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4250 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4251 chown and chgrp in /bin and hostname in /usr/bin.
4253 28. Exim was keeping the "process log" file open after each use, just as it
4254 does for the main log. This opens the possibility of it remaining open for
4255 long periods when the USR1 signal hits a daemon. Occasional processlog
4256 errors were reported, that could have been caused by this. Anyway, it seems
4257 much more sensible not to leave this file open at all, so that is what now
4260 29. The long-running daemon process does not normally write to the log once it
4261 has entered its main loop, and it closes the log before doing so. This is
4262 so that log files can straightforwardly be renamed and moved. However,
4263 there are a couple of unusual error situations where the daemon does write
4264 log entries, and I had neglected to close the log afterwards.
4266 30. The text of an SMTP error response that was received during a remote
4267 delivery was being truncated at 512 bytes. This is too short for some of
4268 the long messages that one sometimes sees. I've increased the limit to
4271 31. It is now possible to make retry rules that apply only when a message has a
4272 specific sender, in particular, an empty sender.
4274 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4275 it possible to be selective about when SMTP synchronization is enforced.
4277 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4279 32. Implemented hosts_connection_nolog.
4281 33. Added an ACL for QUIT.
4283 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4286 35. Added mailbox_size and mailbox_filecount to appendfile.
4288 36. Added control = no_multiline_responses to ACLs.
4290 37. There was a bug in the logic of the code that waits for the clock to tick
4291 in the case where the clock went backwards by a substantial amount such
4292 that the microsecond fraction of "now" was more than the microsecond
4293 fraction of "then" (but the whole seconds number was less).
4295 38. Added support for the libradius Radius client library this is found on
4296 FreeBSD (previously only the radiusclient library was supported).
4302 1. When certain lookups returned multiple values in the form name=value, the
4303 quoting of the values was not always being done properly. Specifically:
4304 (a) If the value started with a double quote, but contained no whitespace,
4306 (b) If the value contained whitespace other than a space character (i.e.
4307 tabs or newlines or carriage returns) it was not quoted.
4308 This fix has been applied to the mysql and pgsql lookups by writing a
4309 separate quoting function and calling it from the lookup code. The fix
4310 should probably also be applied to nisplus, ibase and oracle lookups, but
4311 since I cannot test any of those, I have not disturbed their existing code.
4313 2. A hit in the callout cache for a specific address caused a log line with no
4314 reason for rejecting RCPT. Now it says "Previous (cached) callout
4315 verification failure".
4317 3. There was an off-by-one bug in the queryprogram router. An over-long
4318 return line was truncated at 256 instead of 255 characters, thereby
4319 overflowing its buffer with the terminating zero. As well as fixing this, I
4320 have increased the buffer size to 1024 (and made a note to document this).
4322 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4323 when Exim is waiting for an SMTP response from a remote server, Exim
4324 restarts its select() call on the socket, thereby resetting its timeout.
4325 This is not a problem when such interrupts are rare. Somebody set up a cron
4326 job to run exiwhat every 2 minutes, which is less than the normal select()
4327 timeout (5 or 10 minutes). This meant that the select() timeout never
4328 kicked in because it was always reset. I have fixed this by comparing the
4329 time when an interrupt arrives with the time at the start of the first call
4330 to select(). If more time than the timeout has elapsed, the interrupt is
4331 treated as a timeout.
4333 5. Some internal re-factoring in preparation for the addition of Sieve
4334 extensions (by MH). In particular, the "personal" test is moved to a
4335 separate function, and given an option for scanning Cc: and Bcc: (which is
4336 not set for Exim filters).
4338 6. When Exim created an email address using the login of the caller as the
4339 local part (e.g. when creating a From: or Sender: header line), it was not
4340 quoting the local part when it contained special characters such as @.
4342 7. Installed new OpenBSD configuration files.
4344 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4345 try to make them clearer.
4347 9. Callout options, other than the timeout value, were being ignored when
4348 verifying sender addresses in header lines. For example, when using
4350 verify = header_sender/callout=no_cache
4352 the cache was (incorrectly) being used.
4354 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4355 only the Cygwin environment.
4357 11. When return_path_on_delivery was set as a log selector, if different remote
4358 addresses in the same message used different return paths and parallel
4359 remote delivery occurred, the wrong values would sometimes be logged.
4360 (Whenever a remote delivery process finished, the return path value from
4361 the most recently started remote delivery process was logged.)
4363 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4364 lines when AUTH and/or TLS are in use. This is the "received protocol"
4365 field. Exim used to use "asmtp" for authenticated SMTP, without any
4366 indication (in the protocol name) for TLS use. Now it follows the RFC and
4367 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4368 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4369 names appear in log lines as well as in Received: header lines.
4371 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4372 extensions, and comparison tests, and to fix some bugs.
4374 14. Changes to the "personal" filter test:
4376 (1) The test was buggy in that it was just doing the equivalent of
4377 "contains" tests on header lines. For example, if a user's address was
4378 anne@some.where, the "personal" test would incorrectly be true for
4380 To: susanne@some.where
4382 This test is now done by extracting each address from the header in turn,
4383 and checking the entire address. Other tests that are part of "personal"
4384 are now done using regular expressions (for example, to check local parts
4385 of addresses in From: header lines).
4387 (2) The list of non-personal local parts in From: addresses has been
4388 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4389 taken from the Sieve specification recommendations.
4391 (3) If the message contains any header line starting with "List-" it is
4392 treated as non-personal.
4394 (4) The test for "circular" in the Subject: header line has been removed
4395 because it now seems ill-conceived.
4397 15. Minor typos in src/EDITME comments corrected.
4399 16. Installed latest exipick from John Jetmore.
4401 17. If headers_add on a router specified a text string that was too long for
4402 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4403 of string_sprintf() is now avoided.
4405 18. $message_body_size was not set (it was always zero) when running the DATA
4406 ACL and the local_scan() function.
4408 19. For the "mail" command in an Exim filter, no default was being set for
4409 the once_repeat time, causing a random time value to be used if "once" was
4410 specified. (If the value happened to be <= 0, no repeat happened.) The
4411 default is now 0s, meaning "never repeat". The "vacation" command was OK
4412 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4413 (I found it when inspecting the code).
4415 20. There is now an overall timeout for performing a callout verification. It
4416 defaults to 4 times the callout timeout, which applies to individual SMTP
4417 commands during the callout. The overall timeout applies when there is more
4418 than one host that can be tried. The timeout is checked before trying the
4419 next host. This prevents very long delays if there are a large number of
4420 hosts and all are timing out (e.g. when the network connections are timing
4421 out). The value of the overall timeout can be changed by specifying an
4422 additional sub-option for "callout", called "maxwait". For example:
4424 verify = sender/callout=5s,maxwait=20s
4426 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4427 to the end before writing, but this should make it even safer).
4429 22. Exim was forgetting that it had advertised PIPELINING for the second and
4430 subsequent messages on an SMTP connection. It was also not resetting its
4431 memory on STARTTLS and an internal HELO.
4433 23. When Exim logs an SMTP synchronization error within a session, it now
4434 records whether PIPELINING has been advertised or not.
4436 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4437 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4438 rather than long int.
4440 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4442 26. Added the never_mail option to autoreply.
4448 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4449 crash if the getsockname() call failed; this can happen if a connection is
4450 closed very soon after it is established. The problem was simply in the
4451 order in which certain operations were done, causing Exim to try to write
4452 to the SMTP stream before it had set up the file descriptor. The bug has
4453 been fixed by making things happen in the correct order.
4459 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4460 before the connection was closed, thus losing the rejection response.
4462 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4463 some early Solaris releases, but causes trouble in current releases where
4464 socklen_t is defined.
4466 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4469 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4472 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4473 that releasing the top of it at the end releases what was used for sub-
4474 expansions (unless the block got too big). However, discard this block if
4475 the first thing is a variable or header, so that we can use its block when
4476 it is dynamic (useful for very large $message_headers, for example).
4478 6. Lookups now cache *every* query, not just the most recent. A new, separate
4479 store pool is used for this. It can be recovered when all lookup caches are
4480 flushed. Lookups now release memory at the end of their result strings.
4481 This has involved some general refactoring of the lookup sources.
4483 7. Some code has been added to the store_xxx() functions to reduce the amount
4484 of flapping under certain conditions.
4486 8. log_incoming_interface used to affect only the <= reception log lines. Now
4487 it causes the local interface and port to be added to several more SMTP log
4488 lines, for example "SMTP connection from", and rejection lines.
4490 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4492 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4494 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4495 could be overwritten at the end of the current message (or the start of a
4496 new message if it was set in a HELO ACL). The value is now preserved for
4497 the duration of the SMTP connection.
4499 12. If a transport had a headers_rewrite setting, and a matching header line
4500 contained an unqualified address, that address was qualified, even if it
4501 did not match any rewriting rules. The underlying bug was that the values
4502 of the flags that permit the existence of unqualified sender and recipient
4503 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4504 non-local messages, and by -bnq for local messages) were not being
4505 preserved with the message after it was received.
4507 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4508 "next input=" as part of the text comprising the host identity instead of
4509 the correct text. The code was using the same buffer for two different
4510 strings. However, depending on which order the printing function evaluated
4511 its arguments, the bug did not always show up. Under Linux, for example, my
4512 test suite worked just fine.
4514 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4515 doesn't work with some older versions of Perl. It has been changed to "my",
4516 which in any case is probably the better facility to use.
4518 15. A really picky compiler found some instances of statements for creating
4519 error messages that either had too many or two few arguments for the format
4522 16. The size of the buffer for calls to the DNS resolver has been increased
4523 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4524 for addresses that have a lot of PTR records. This alleviates a problem; it
4525 does not fully solve it.
4527 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4528 buffer now truncates the list and logs the incident, which is the same
4529 action as happens when Exim is looking up a host name and its aliases.
4530 Previously in this situation something unpredictable would happen;
4531 sometimes it was "internal error: store_reset failed".
4533 18. If a server dropped the connection unexpectedly when an Exim client was
4534 using GnuTLS and trying to read a response, the client delivery process
4535 crashed while trying to generate an error log message.
4537 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4538 string, for example:
4540 warn message = H1: something\nH2: something
4542 the text was added as a single header line from Exim's point of view
4543 though it ended up OK in the delivered message. However, searching for the
4544 second and subsequent header lines using $h_h2: did not work. This has been
4545 fixed. Similarly, if a system filter added multiple headers in this way,
4546 the routers could not see them.
4548 20. Expanded the error message when iplsearch is called with an invalid key to
4549 suggest using net-iplsearch in a host list.
4551 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4552 ACLs are no longer actually imposed (and a message to that effect is
4555 22. If a "gecos" field in a passwd entry contained escaped characters, in
4556 particular, if it contained a \" sequence, Exim got it wrong when building
4557 a From: or a Sender: header from that name. A second bug also caused
4558 incorrect handling when an unquoted " was present following a character
4559 that needed quoting.
4561 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4562 was not being matched caselessly.
4564 24. Arranged for all hyphens in the exim.8 source to be escaped with
4567 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4568 properly. Recipient callout cache records were still being keyed to include
4569 the sender, even when use_sender was set false. This led to far more
4570 callouts that were necessary. The sender is no longer included in the key
4571 when use_sender is false.
4573 26. Added "control = submission" modifier to ACLs.
4575 27. Added the ${base62d: operator to decode base 62 numbers.
4577 28. dnsdb lookups can now access SRV records.
4579 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4580 the configuration file.
4582 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4583 (-v) mode. This makes the output for a verbose queue run more intelligible.
4585 31. Added a use_postmaster feature to recipient callouts.
4587 32. Added the $body_zerocount variable, containing the number of binary zero
4588 bytes in the message body.
4590 33. The time of last modification of the "new" subdirectory is now used as the
4591 "mailbox time last read" when there is a quota error for a maildir
4594 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4596 35. Added +ignore_unknown as a special item in host lists.
4598 36. Code for decoding IPv6 addresses in host lists is now included, even if
4599 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4600 address was recognized as an IP address, but was then not correctly decoded
4601 into binary, causing unexpected and incorrect effects when compared with
4608 1. Very minor rewording of debugging text in manualroute to say "list of
4609 hosts" instead of "hostlist".
4611 2. If verify=header_syntax was set, and a header line with an unqualified
4612 address (no domain) and a large number of spaces between the end of the
4613 name and the colon was received, the reception process suffered a buffer
4614 overflow, and (when I tested it) crashed. This was caused by some obsolete
4615 code that should have been removed. The fix is to remove it!
4617 3. When running in the test harness, delay a bit after writing a bounce
4618 message to get a bit more predictability in the log output.
4620 4. Added a call to search_tidyup() just before forking a reception process. In
4621 theory, someone could use a lookup in the expansion of smtp_accept_max_
4622 per_host which, without the tidyup, could leave open a database connection.
4624 5. Added the variables $recipient_data and $sender_data which get set from a
4625 lookup success in an ACL "recipients" or "senders" condition, or a router
4626 "senders" option, similar to $domain_data and $local_part_data.
4628 6. Moved the writing of debug_print from before to after the "senders" test
4631 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4632 problems for message scanning, either using a data ACL, or using
4633 local_scan() because the Received: header was not generated till after they
4634 were called (in order to set the time as the time of reception completion).
4635 I have revised the way this works. The header is now generated after the
4636 body is received, but before the ACL or local_scan() are called. After they
4637 are run, the timestamp in the header is updated.
4643 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4644 before starting a queue runner without re-exec. This happened only when
4645 deliver_drop_privilege was set or when the Exim user was set to root. The
4646 effect of the bug was that timeouts during subsequent deliveries caused
4647 crashes instead of being properly handled. The handler is now left at its
4648 default (and expected) setting.
4650 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4651 message, again when deliver_drop_privilege is set or Exim is run as root.
4652 The bug described in (1) was not present in this case, but the tidying up
4653 of the other signals was missing. I have made the two cases consistent.
4655 3. The ignore_target_hosts setting on a manualroute router was being ignored
4656 for hosts that were looked up using the /MX notation.
4658 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4661 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4662 operated on the sender address. After changing the $sender_address to <>
4663 for the sender address verify, Exim was re-instated it as the original
4664 (before rewriting) address, but remembering that it had rewritten it, so it
4665 wasn't rewriting it again. This bug also had the effect of breaking the
4666 sender address verification caching when the sender address was rewritten.
4668 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4669 This has been changed so that if the ip literal address matches
4670 ignore_target_hosts, the router declines.
4672 7. Added expansion conditions match_domain, match_address, and match_local_
4673 part (NOT match_host).
4675 8. The placeholder for the Received: header didn't have a length field set.
4677 9. Added code to Exim itself and to exim_lock to test for a specific race
4678 condition that could lead to file corruption when using MBX delivery. The
4679 issue is with the lockfile that is created in /tmp. If this file is removed
4680 after a process has opened it but before that process has acquired a lock,
4681 there is the potential for a second process to recreate the file and also
4682 acquire a lock. This could lead to two Exim processes writing to the file
4683 at the same time. The added code performs the same test as UW imapd; it
4684 checks after acquiring the lock that its file descriptor still refers to
4685 the same named file.
4687 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4688 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4691 11. Added the smtp_active_hostname option. If used, this will typically be made
4692 to depend on the incoming interface address. Because $interface_address is
4693 not set up until the daemon has forked a reception process, error responses
4694 that can happen earlier (such as "too many connections") no longer contain
4697 12. If an expansion in a condition on a "warn" statement fails because a lookup
4698 defers, the "warn" statement is abandoned, and the next ACL statement is
4699 processed. Previously this caused the whole ACL to be aborted.
4701 13. Added the iplsearch lookup type.
4703 14. Added ident_timeout as a log selector.
4705 15. Added tls_certificate_verified as a log selector.
4707 16. Added a global option tls_require_ciphers (compare the smtp transport
4708 option of the same name). This controls incoming TLS connections.
4710 17. I finally figured out how to make tls_require_ciphers do a similar thing
4711 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4712 before starting the TLS session.
4714 18. Tabs are now shown as \t in -bP output.
4716 19. If the log selector return_path_on_delivery was set, Exim crashed when
4717 bouncing a message because it had too many Received: header lines.
4719 20. If two routers both had headers_remove settings, and the first one included
4720 a superfluous trailing colon, the final name in the first list and the
4721 first name in the second list were incorrectly joined into one item (with a
4722 colon in the middle).
4728 1. Added -C and -D options to the exinext utility, mainly to make it easier
4729 to include in the automated testing, but these could be helpful when
4730 multiple configurations are in use.
4732 2. The exinext utility was not formatting the output nicely when there was
4733 an alternate port involved in the retry record key, nor when there was a
4734 message id as well (for retries that were specific to a specific message
4735 and a specific host). It was also confused by IPv6 addresses, because of
4736 the additional colons they contain. I have fixed the IPv4 problem, and
4737 patched it up to do a reasonable job for IPv6.
4739 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4740 delivery, the log line now contains "pipelined" if PIPELINING was used.
4742 4. An SMTP transport process used to panic and die if the bind() call to set
4743 an explicit outgoing interface failed. This has been changed; it is now
4744 treated in the same way as a connect() failure.
4746 5. A reference to $sender_host_name in the part of a conditional expansion
4747 that was being skipped was still causing a DNS lookup. This no longer
4750 6. The def: expansion condition was not recognizing references to header lines
4751 that used bh_ and bheader_.
4753 7. Added the _cache feature to named lists.
4755 8. The code for checking quota_filecount in the appendfile transport was
4756 allowing one more file than it should have been.
4758 9. For compatibility with Sendmail, the command line option
4766 and sets the incoming protocol and host name (for trusted callers). The
4767 host name and its colon can be omitted when only the protocol is to be set.
4768 Note the Exim already has two private options, -pd and -ps, that refer to
4769 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4770 "s", but I don't think that's a major issue.
4772 10. A number of refactoring changes to the code, none of which should affect
4775 (a) The number of logging options was getting close to filling up the
4776 32-bit word that was used as a bit map. I have split them into two classes:
4777 those that are passed in the argument to log_write(), and those that are
4778 only ever tested independently outside of that function. These are now in
4779 separate 32-bit words, so there is plenty of room for expansion again.
4780 There is no change in the user interface or the logging behaviour.
4782 (b) When building, for example, log lines, the code previously used a
4783 macro that called string_cat() twice, in order to add two strings. This is
4784 not really sufficiently general. Furthermore, there was one instance where
4785 it was actually wrong because one of the argument was used twice, and in
4786 one call a function was used. (As it happened, calling the function twice
4787 did not affect the overall behaviour.) The macro has been replaced by a
4788 function that can join an arbitrary number of extra strings onto a growing
4791 (c) The code for expansion conditions now uses a table and a binary chop
4792 instead of a serial search (which was left over from when there were very
4793 few conditions). Also, it now recognizes conditions like "pam" even when
4794 the relevant support is not compiled in: a suitably worded error message is
4795 given if an attempt is made to use such a condition.
4797 11. Added ${time_interval:xxxxx}.
4799 12. A bug was causing one of the ddress fields not to be passed back correctly
4800 from remote delivery subprocesses. The field in question was not being
4801 subsequently used, so this caused to problems in practice.
4803 13. Added new log selectors queue_time and deliver_time.
4805 14. Might have fixed a bug in maildirsizefile handling that threw up
4806 "unexpected character" debug warnings, and recalculated the data
4807 unnecessarily. In any case, I expanded the warning message to give more
4810 15. Added the message "Restricted characters in address" to the statements in
4811 the default ACL that block characters like @ and % in local parts.
4813 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4814 Three changes have been made:
4816 (a) There was a serious bug; a negative response to MAIL caused the whole
4817 recipient domain to be cached as invalid, thereby blocking all messages
4818 to all local parts at the same domain, from all senders. This bug has
4819 been fixed. The domain is no longer cached after a negative response to
4820 MAIL if the sender used is not empty.
4822 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4825 (c) A new callout option, "use_sender" has been added for people who want
4826 the modified behaviour.
4832 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4835 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4836 indeed breaks things for older releases.
4838 3. Added additional logging to the case where there is a problem reading data
4839 from a filter that is running in a subprocess using a pipe, in order to
4840 try to track down a specific problem.
4842 4. Testing facility fudge: when running in the test harness and attempting
4843 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4844 getting "No route to host". Convert this to a timeout.
4846 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4849 6. Some OS don't have socklen_t but use size_t instead. This affects the
4850 fifth argument of getsockopt() amongst other things. This is now
4851 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4852 can be set for individual OS. I have set it for SunOS5, OSF1, and
4853 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4854 some earlier ones do not.
4856 7. Change 4.30/15 was not doing the test caselessly.
4858 8. The standard form for an IPv6 address literal was being rejected by address
4859 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4860 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4861 this, as well as the form without the "IPv6" on the front (but only when
4862 address literals are enabled, of course).
4864 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4866 10. Exim crashed if a message with an empty sender address specified by -f
4867 encountered a router with an errors_to setting. This could be provoked only
4868 by a command such as
4872 where an empty string was supplied; "<>" did not hit this bug.
4874 11. Installed PCRE release 4.5.
4876 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4877 remained set. It is now erased.
4879 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4880 times from message ids (which are base 36 rather than the normal 62).
4882 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4883 were being counted as actual protocol errors, and logged if the log
4884 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4885 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4886 and DATA following a set of rejected RCPTs do not count as protocol errors.
4887 In other words, Exim assumes they were pipelined, though this may not
4888 actually be the case. Of course, in all cases the client gets an
4889 appropriate error code.
4891 15. If a lookup fails in an ACL condition, a message about the failure may
4892 be available; it is used if testing the ACL cannot continue, because most
4893 such messages specify what the cause of the deferral is. However, some
4894 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4895 that caused an old message to be retained and used if a later statement
4896 caused a defer, replacing the real cause of the deferral.
4898 16. If an IP address had so many PTR records that the DNS lookup buffer
4899 was not large enough to hold them, Exim could crash while trying to process
4900 the truncated data. It now detects and logs this case.
4902 17. Further to 4.21/58, another change has been made: if (and only if) the
4903 first line of a message (the first header line) ends with CRLF, a bare LF
4904 in a subsequent header line has a space inserted after it, so as not to
4905 terminate the header.
4907 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4908 unnecessarily, used atoi() instead of strtol(), and didn't check the
4909 termination when getting file sizes from file names by regex.
4911 19. Completely re-implemented the support for maildirsize files, in the light
4912 of a number of problems with the previous contributed implementation
4913 (4.30/29). In particular:
4915 . If the quota is zero, the maildirsize file is maintained, but no quota is
4918 . If the maildir directory does not exist, it is created before any attempt
4919 to write a maildirsize file.
4921 . The quota value in the file is just a cache; if the quota is changed in
4922 the transport, the new value overrides.
4924 . A regular expression is available for excluding directories from the
4927 20. The autoreply transport checks the characters in options that define the
4928 message's headers; it allows continued headers, but it was checking with
4929 isspace() after an embedded newline instead of explicitly looking for a
4932 21. If all the "regular" hosts to which an address was routed had passed their
4933 expiry times, and had not reached their retry times, the address was
4934 bounced, even if fallback hosts were defined. Now Exim should go on to try
4937 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4938 equivalent code in the SMTP transport. Some hosts send humungous responses
4939 to HELO/EHLO, more than 1024 it seems.
4941 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4942 gives compiler warnings in some environments. I've now done it "properly",
4945 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4946 (because of problems with the built-in one) was declared to return uschar *
4947 instead of char *, causing compiler failure.
4949 25. Fixed a file descriptor leak when processing alias/forward files.
4951 26. Fixed a minor format string issue in dbfn.c.
4953 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4955 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4956 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4959 29. When a sender address is verified, it is cached, to save repeating the test
4960 when there is more than one recipient in a message. However, when the
4961 verification involves a callout, it is possible for different callout
4962 options to be set for different recipients. It is too complicated to keep
4963 track of this in the cache, so now Exim always runs a verification when a
4964 callout is required, relying on the callout cache for the optimization.
4965 The overhead is duplication of the address routing, but this should not be
4968 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4969 to be verified with callout=postmaster, and the main callout worked but the
4970 postmaster check failed, the verification correctly failed. However, if a
4971 subsequent RCPT command asked for sender verification *without* the
4972 postmaster check, incorrect caching caused this verification also to fail,
4975 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4976 it was not caching the DNS options (qualify_single, search_parents) that
4977 were used when the lookup failed. A subsequent lookup with different
4978 options therefore always gave the same answer, though there were cases
4979 where it should not have. (Example: a "domains = !$mx_any" option on a
4980 dnslookup router: the "domains" option is always processed without any
4981 widening, but the router might have qualify_single set.) Now Exim uses the
4982 cached value only when the same options are set.
4984 32. Added John Jetmore's "exipick" utility to the distribution.
4986 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4987 than a timeout (e.g. a certificate is required, and is not provided), an
4988 Exim server now closes the connection immediately. Previously it waited for
4989 the client to close - but if the client is SSL, it seems that they each
4990 wait for each other, leading to a delay before one of them times out.
4992 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4993 maintained 0.8.x compatibility because I don't think many are using it, and
4994 it is clearly obsolete.
4996 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4999 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5000 client certificate was expired. A simple patch fixes this, though I don't
5001 understand the full logic of why the verify callback is called multiple
5004 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5005 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5006 which causes problems with some clients (such as the Certicom SSL Plus
5007 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5008 disables the coutermeasure allowing Eudora to connect."
5010 38. Exim was not checking that a write() to a log file succeeded. This could
5011 lead to Bad Things if a log got too big, in particular if it hit a file
5012 size limit. Exim now panics and dies if it cannot write to a log file, just
5013 as it does if it cannot open a log file.
5015 39. Modified OS/Makefile-Linux so that it now contains
5017 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5019 The two -D definitions ensure that Exim is compiled with large file
5020 support, which makes it possible to handle log files that are bigger than
5023 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5024 instance) a domain was checked against a named list that involved a lookup,
5025 causing $domain_data to be set, then another domain was checked against the
5026 same list, then the first domain was re-checked, the value of $domain_data
5027 after the final check could be wrong. In particular, if the second check
5028 failed, it could be set empty. This bug probably also applied to
5031 41. The strip_trailing_dot option was not being applied to the address given
5032 with the -f command-line option.
5034 42. The code for reading a message's header from the spool was incrementing
5035 $received_count, but never initializing it. This meant that the value was
5036 incorrect (doubled) while delivering a message in the same process in which
5037 it was received. In the most common configuration of Exim, this never
5038 happens - a fresh exec is done - but it can happen when
5039 deliver_drop_privilege is set.
5041 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5042 it now includes up to 150 characters of the unexpected data in the log
5045 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5046 and building data strings. The size of both of these buffers was 10 000
5047 bytes - far larger than anybody would *ever* want, thought I. Needless to
5048 say, somebody hit the limit. I have increased the maximum line length to
5049 20 000 and the maximum data length of concatenated lines to 100 000. I have
5050 also fixed two bugs, because there was no checking on these buffers. Tsk,
5051 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5052 buffer is too small.
5054 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5055 lsearch lookups. Now it does.
5057 46. When parsing a route_list item in a manualroute router, a fixed-length
5058 buffer was used for the list of hosts. I made this 1024 bytes long,
5059 thinking that nobody would ever have a list of hosts that long. Wrong.
5060 Somebody had a whole pile of complicated expansion conditions, and the
5061 string was silently truncated, leading to an expansion error. It turns out
5062 that it is easier to change to an unlimited length (owing to other changes
5063 that have happened since this code was originally written) than to build
5064 structure for giving a limitation error. The length of the item that
5065 expands into the list of hosts is now unlimited.
5067 47. The lsearch lookup could not handle data where the length of text line was
5068 more than 4095 characters. Such lines were truncated, leading to shortened
5069 data being returned. It should now handle lines of any length.
5071 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5072 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5075 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5076 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5077 no longer happens in comment lines. A list of replacements is now placed
5078 at the head of all of the source files, except those whose only change is
5079 to replace PERL_COMMAND in the very first #! line.
5081 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5082 messages on the queue, with a bottom-up merge sort, using code contributed
5083 by Michael Haardt. This should make operations like -bp somewhat faster on
5084 large queues. It won't affect queue runners, except when queue_run_in_order
5087 51. Installed eximstats 1.31 in the distribution.
5089 52. Added support for SRV lookups to the dnslookup router.
5091 53. If an ACL referred to $message_body or $message_body_end, the value was not
5092 reset for any messages that followed in the same SMTP session.
5094 54. The store-handling optimization for building very long strings was not
5095 differentiating between the different store pools. I don't think this
5096 actually made any difference in practice, but I've tidied it.
5098 55. While running the routers to verify a sender address, $sender_address
5099 was still set to the sender address. This is wrong, because when routing to
5100 send a bounce to the sender, it would be empty. Therefore, I have changed
5101 it so that, while verifying a sender address, $sender_address is set to <>.
5102 (There is no change to what happens when verifying a recipient address.)
5104 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5105 target A or AAAA records (if not already returned) without resetting the
5106 qualify_single or search_parents options of the DNS resolver. These are
5107 inappropriate in this case because the targets of MX and SRV records must
5108 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5109 target that, when qualified, matched something in the local domain. These
5110 two options are now turned off when doing these lookups.
5112 57. It seems that at least some releases of Reiserfs (which does not have the
5113 concept of a fixed number of inodes) returns zero and not -1 for the
5114 number of available inodes. This interacted badly with check_spool_inodes,
5115 which assumed that -1 was the "no such thing" setting. What I have done is
5116 to check that the total number of inodes is greater than zero before doing
5117 the test of how many are available.
5119 58. When a "warn" ACL statement has a log_message modifier, the message is
5120 remembered, and not repeated. This is to avoid a lot of repetition when a
5121 message has many recipients that cause the same warning to be written.
5122 Howewer, Exim was preserving the list of already written lines for an
5123 entire SMTP session, which doesn't seem right. The memory is now reset if a
5124 new message is started.
5126 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5127 debugging output unless log_rewrite was also set.
5129 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5130 of (int)(handle) when we know that handle contains (void *)(-1).
5132 61. The Exim daemon panic-logs an error return when it closes the incoming
5133 connection. However "connection reset by peer" seems to be common, and
5134 isn't really an error worthy of noting specially, so that particular error
5137 62. When Exim is trying to find all the local interfaces, it used to panic and
5138 die if the ioctl to get the interface flags failed. However, it seems that
5139 on at least one OS (Solaris 9) it is possible to have an interface that is
5140 included in the list of interfaces, but for which you get a failure error
5141 for this call. This happens when the interface is not "plumbed" into a
5142 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5143 failure of the "get flags" call assumes that the interface is down.
5145 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5146 makes life easier for people who are doing arithmetic on fields extracted
5147 from dates, where you often get leading zeros that should not be
5148 interpreted as octal.
5150 64. Added qualify_domain to the redirect router, to override the global
5153 65. If a pathologically long header line contained very many addresses (the
5154 report of this problem mentioned 10 000) and each of them was rewritten,
5155 Exim could use up a very large amount of memory. (It kept on making new
5156 copies of the header line as it rewrote, and never released the old ones.)
5157 At the expense of a bit more processing, the header rewriting function has
5158 been changed so that it no longer eats memory in this way.
5160 66. The generation of the Received: header has been moved from the time that a
5161 message starts to be received, to the time that it finishes. The timestamp
5162 in the Received: header should now be very close to that of the <= log
5163 line. There are two side-effects of this change:
5165 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5166 logged header lines no longer include the local Received: line, because
5167 it has not yet been created. The same applies to a copy of the message
5168 that is returned to a non-SMTP sender when a message is rejected.
5170 (b) When a filter file is tested using -bf, no additional Received: header
5171 is added to the test message. After some thought, I decided that this
5174 This change does not affect the value of $received_for. It is still set
5175 after address rewriting, but before local_scan() is called.
5177 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5179 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5180 gave an unhelpful panic error message, and a defer error. I have managed to
5181 change this behaviour so that it now rejects any supplied certificate,
5182 which seems right, as the list of acceptable certificates is empty.
5184 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5185 gave an unhelpful defer error. I have not managed to make this reject any
5186 supplied certificates, but the error message it gives is "no certificate
5187 supplied", which is not helpful.
5189 70. exigrep's output now also includes lines that are not associated with any
5190 message, but which match the given pattern. Implemented by a patch from
5191 Martin Sluka, which also tidied up the Perl a bit.
5193 71. Recipient callout verification, like sender verification, was using <> in
5194 the MAIL FROM command. This isn't really the right thing, since the actual
5195 sender may affect whether the remote host accepts the recipient or not. I
5196 have changed it to use the actual sender in the callout; this means that
5197 the cache record is now keyed on a recipient/sender pair, not just the
5198 recipient address. There doesn't seem to be a real danger of callout loops,
5199 since a callout by the remote host to check the sender would use <>.
5200 [SEE ABOVE: changed after hitting problems.]
5202 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5203 temporary errors. However, in the case of such a code being given after
5204 the end of a data transmission (i.e. after ".") Exim was failing to write
5205 a retry record for the message. (Yes, there was some broken host that was
5206 actually sending 8xx at this point.)
5208 73. An unknown lookup type in a host list could cause Exim to panic-die when
5209 the list was checked. (An example that provoked this was putting <; in the
5210 middle of a list instead of at the start.) If this happened during a DATA
5211 ACL check, a -D file could be left lying around. This kind of configuration
5212 error no longer causes Exim to die; instead it causes a defer errror. The
5213 incident is still logged to the main and panic logs.
5215 74. Buglet left over from Exim 3 conversion. The message "too many messages
5216 in one connection" was written to the rejectlog but not the mainlog, except
5217 when address rewriting (yes!) was being logged.
5219 75. Added write_rejectlog option.
5221 76. When a system filter was run not as root (that is, when system_filter_user
5222 was set), the values of the $n variables were not being returned to the
5223 main process; thus, they were not subsequently available in the $sn
5226 77. Added +return_path_on_delivery log selector.
5228 78. A connection timeout was being treated differently from recipients deferred
5229 when testing hosts_max_try with a message that was older than the host's
5230 retry timeout. (The host should not be counted, thus allowing all hosts to
5231 be tried at least once before bouncing.) This may have been the cause of an
5232 occasionally reported bug whereby a message would remain on the queue
5233 longer than the retry timeout, but would be bounced if a delivery was
5234 forced. I say "may" because I never totally pinned down the problem;
5235 setting up timeout/retry tests is difficult. See also the next item.
5237 79. The ultimate address timeout was not being applied to errors that involved
5238 a combination of host plus message (for example, a timeout on a MAIL
5239 command). When an address resolved to a number of possible hosts, and they
5240 were not all tried for each delivery (e.g. because of hosts_max_try), a
5241 message could remain on the queue longer than the retry timeout.
5243 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5246 81. Fixed an obscure SMTP outgoing bug which required at least the following
5247 conditions: (a) there was another message waiting for the same server;
5248 (b) the server returned 5xx to all RCPT commands in the first message so
5249 that the message was not completed; (c) the server dropped the connection
5250 or gave a negative response to the RSET that Exim sends to abort the
5251 transaction. The observed case was a dropped connection after DATA that had
5252 been sent in pipelining mode. That is, the server had advertised PIPELINING
5253 but was not implementing it correctly. The effect of the bug was incorrect
5254 behaviour, such as trying another host, and this could lead to a crash.
5260 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5261 and daemon.c were passed as pointers to ints; they should have been
5262 pointers to socklen_t variables (which are typically unsigned ints).
5264 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5267 3. Fixed a really odd bug that affected only the testing scheme; patching a
5268 certain fixed string in the binary changed the value of another string that
5269 happened to be identical to the end of the original first string.
5271 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5272 name", it returns that address as the IP address. On some operating
5273 systems (e.g. Solaris), it also passes back the IP address string as the
5274 "host name". However, on others (e.g. Linux), it passes back an empty
5275 string. Exim wasn't checking for this, and was changing the host name to an
5276 empty string, assuming it had been canonicized.
5278 5. Although rare, it is permitted to have more than one PTR record for a given
5279 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5280 all the names associated with an address, because they do in Solaris.
5281 However, it seems that they do not in Linux for data that comes from the
5282 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5283 I found this out when I moved to a new Linux workstation and tried to run
5284 the Exim test suite.
5286 To get round this problem I have changed the code so that it now does its
5287 own call to the DNS to look up PTR records when searching for a host name.
5288 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5289 addresses that are only in /etc/hosts are still found.
5291 This behaviour is, however, controlled by an option called host_lookup_
5292 order, which defaults to "bydns:byaddr". If people want to use the other
5293 order, or indeed, just use one or the other means of lookup, they can
5294 specify it in this variable.
5296 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5297 some operating systems, this comes back from gethostbyaddr() as an empty
5298 string, and this is what Exim used to test for. However, it seems that in
5299 other systems, "." is yielded. Exim now tests for this case too.
5301 7. The values of check_spool_space and check_log_space are now held internally
5302 as a number of kilobytes instead of an absolute number of bytes. If a
5303 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5304 kilobyte. This means that much larger values can be stored.
5306 8. Exim monitor: an attempt to get the action menu when not actually pointing
5307 at a message produces an empty menu entitled "No message selected". This
5308 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5309 no entries in it ("Shell widget menu has zero width and/or height"). So I
5310 have added a single, blank menu entry in this case.
5312 9. Added ${quote_local_part.
5314 10. MIME decoding is now applied to the contents of Subject: header lines when
5317 11. Now that a reference to $sender_host_address automatically causes a reverse
5318 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5319 host lookup before query-style lookups in lists that might use this
5320 variable. This has therefore been abolished, and the "net-" prefix is no
5321 longer necessary for query-style lookups.
5323 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5324 have been a typo for LFLAGS, so it has been changed.
5326 13. The install script calls Exim with "-C /dev/null" in order to find the
5327 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5328 to be output. Howeve, since Exim outputs its version number before the
5329 error, it didn't break the script. It just looked ugly. I fixed this by
5330 always allowing "-C /dev/null" if the caller is root.
5332 14. Ignore overlarge ACL variable number when reading spool file - insurance
5333 against a later release with more variables having written the file.
5335 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5336 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5337 this, as well as the form without the "IPv6" on the front.
5339 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5340 OS/Makefile-Darwin file.
5342 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5343 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5345 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5346 variable. However, some OS use ints for inodes. I've added cast to long int
5347 to get rid of the compiler warning.
5349 19. I had forgotten to lock out "/../" in configuration file names when
5350 ALT_CONFIG_PREFIX was set.
5352 20. Routers used for verification do not need to specify transports. However,
5353 if such a router generated a host list, and callout was configured, Exim
5354 crashed, because it could not find a port number from the (non-existent)
5355 transport. It now assumes port 25 in this circumstance.
5357 21. Added the -t option to exigrep.
5359 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5360 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5361 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5362 them is set, LOOKUP_LSEARCH is forced.
5364 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5366 24. Added sender and host information to the "rejected by local_scan()" log
5367 line; previously there was no indication of these.
5369 25. Added .include_if_exists.
5371 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5372 Linux. It turns out that not all file systems support this. Apparently some
5373 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5374 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5375 not supported on the file descriptor, is now ignored when Exim is trying to
5376 sync a directory. This applies only to Linux.
5378 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5380 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5381 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5384 29. Added support for maildirsize files from supplied patch (modified a bit).
5386 30. The use of :fail: followed by an empty string could lead Exim to respond to
5387 sender verification failures with (e.g.):
5389 550 Verification failed for <xxx>
5390 550 Sender verify failed
5392 where the first response line was missing the '-' that indicates it is not
5393 the final line of the response.
5395 31. The loop for finding the name of the user that called Exim had a hardwired
5396 limit of 10; it now uses the value of finduser_retries, which is used for
5397 all other user lookups.
5399 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5402 33. Exim was neglecting to zero errno before one call of strtol() when
5403 expanding a string and expecting an integer value. On some systems this
5404 resulted in spurious "integer overflow" errors. Also, it was casting the
5405 result into an int without checking.
5407 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5408 did not work. The code looks as if it has *never* worked, though it appears
5409 to have been documented since at least releast 1.62. I have made it work.
5411 35. The "timeout_DNS" error in retry rules, also documented since at least
5412 1.62, also never worked. As it isn't clear exactly what this means, and
5413 clearly it isn't a major issue, I have abolished the feature by treating it
5414 as "timeout", and writing a warning to the main and panic logs.
5416 36. The display of retry rules for -brt wasn't always showing the error code
5419 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5420 timeout_connect_A, timeout_connect_MX.
5422 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5423 to the empty sender.
5425 39. The daemon was not analysing the content of -oX till after it had closed
5426 stderr and disconnected from the controlling terminal. This meant that any
5427 syntax errors were only noted on the panic log, and the return code from
5428 the command was 0. By re-arranging the code a little, I've made the
5429 decoding happen first, so such errors now appear on stderr, and the return
5430 code is 1. However, the actual setting up of the sockets still happens in
5431 the disconnected process, so errors there are still only recorded on the
5434 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5435 connections (as happens on some IP stacks) was logged at start up time as
5436 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5437 it from "IPv6 and IPv4", which means that two separate sockets are being
5440 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5441 says whether AF_INET or AF_INET6 was passed as an argument.
5443 42. Exiwhat output was messed up when time zones were included in log
5446 43. Exiwhat now gives more information about the daemon's listening ports,
5447 and whether -tls-on-connect was used.
5449 44. The "port" option of the smtp transport is now expanded.
5451 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5452 silently ignored. Now an error message is written to the main and panic
5455 46. There's a new ACL modifier called "logwrite" which writes to a log file
5456 as soon as it is encountered.
5458 47. Added $local_user_uid and $local_user_gid at routing time.
5460 48. Exim crashed when trying to verify a sender address that was being
5463 49. Exim was recognizing only a space character after ".include". It now also
5464 recognizes a tab character.
5466 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5467 extracting the relevant information from the specification. The man page no
5468 longer contains scrambled data for the -d option, and I've added a section
5469 at the front about calling Exim under different names.
5471 51. Added "extra_headers" argument to the "mail" command in filter files.
5473 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5476 53. Installed eximstats 1.29.
5478 54. Added transport_filter_timeout as a generic transport option.
5480 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5481 Cc: header lines. This was required by RFC 822, but it not required by RFC
5484 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5485 incoming messages that did not have them. Now it does so only if the
5486 message originates locally, that is, if there is no associated remote host
5487 address. When Resent- header lines are present, this applies to the Resent-
5488 lines rather than the non-Resent- lines.
5490 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5491 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5493 58. Messages for configuration errors now include the name of the main
5494 configuration file - useful now that there may be more than one file in a
5495 list (.included file names were always shown).
5497 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5498 for those rare installations that do not start the daemon as root or run it
5499 setuid root. I've cut out the call to initgroups() if the daemon is not
5502 60. The Exim user and group can now be bound into the binary as text strings
5503 that are looked up at the start of Exim's processing.
5505 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5507 62. Added $mailstore_basename variable.
5509 63. Installed patch to sieve.c from Michael Haardt.
5511 64. When Exim failed to open the panic log after failing to open the main log,
5512 the original message it was trying to log was written to stderr and debug
5513 output, but if they were not available (the usual case in production), it
5514 was lost. Now it is written to syslog before the two lines that record the
5515 failures to open the logs.
5517 65. Users' Exim filters run in subprocesses under the user's uid. It is
5518 possible for a "deliver" command or an alias in a "personal" command to
5519 provoke an address rewrite. If logging of address rewriting is configured,
5520 this fails because the process is not running as root or exim. There may be
5521 a better way of dealing with this, but for the moment (because 4.30 needs
5522 to be released), I have disabled address rewrite logging when running a
5523 filter in a non-root, non-exim process.
5529 1. The buildconfig auxiliary program wasn't quoting the value set for
5530 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5531 not defined. This bug was masked in 4.22 by the effect that was fixed in
5534 2. Some messages that were rejected after a message id was allocated were
5535 shown as "incomplete" by exigrep. It no longer does this for messages that
5536 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5538 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5539 have allow_domain_literals set, the ID did not get logged in the <= line.
5540 Domain literals are now always recognized in Message-ID: header lines.
5542 4. The first argument for a ${extract expansion item is the key name or field
5543 number. Leading and trailing spaces in this item were not being ignored,
5544 causing some misleading effects.
5546 5. When deliver_drop_privilege was set, single queue runner processes started
5547 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5548 same command in the process it spins off) were not dropping privilege.
5550 6. When the daemon running as "exim" started a queue runner, it always
5551 re-executed Exim in the spun-off process. This is a waste of effort when
5552 deliver_drop_privilege is set. The new process now just calls the
5553 queue-runner function directly.
5559 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5562 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5563 ignored. Though the use of -oP was forcing the writing of a pid file, it
5564 was always written to the default place.
5566 3. If the message "no IP address found for host xxxx" is generated during
5567 incoming verification, it is now followed by identification of the incoming
5568 connection (so you can more easily find what provoked it).
5570 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5572 5. Added some features to "harden" Exim a bit more against certain attacks:
5574 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5575 be put in Local/Makefile. This is like the never_users runtime option,
5576 but it cannot be overridden. The default setting is "root".
5578 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5579 prefix string with which any file named in a -C command line option
5582 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5583 is retained for -C and -D only if the caller of Exim is root. Without
5584 it, the exim user may also use -C and -D and retain privilege.
5586 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5587 command line option is disabled.
5589 6. Macro names set by the -D option must start with an upper case letter, just
5590 like macro names defined in the configuration file.
5592 7. Added "dereference=" facility to LDAP.
5594 8. Two instances of the typo "uknown" in the source files are fixed.
5596 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5597 the Configure-Makefile script screwed up while processing it.
5599 10. Incorporated PCRE 4.4.
5601 11. The SMTP synchronization check was not operating right at the start of an
5602 SMTP session. For example, it could not catch a HELO sent before the client
5603 waited for the greeting. There is now a check for outstanding input at the
5604 point when the greeting is written. Because of the duplex, asynchronous
5605 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5606 way, but not yet received, when the check is performed.
5608 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5609 on TCP/IP sockets, because this apparently causes some broken clients to
5612 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5613 unchanged) from the Cygwin maintainer.
5615 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5616 was supported instead of testing for mbx. Effectively a typo.
5618 15. The spa authenticator server code was not checking that the input it
5619 received was valid base64.
5621 16. The debug output line for the "set" modifier in ACLs was not showing the
5622 name of the variable that was being set.
5624 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5626 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5627 lookup on its own. Something else was needed to trigger the lookup. For
5628 example, a match in host_lookup or the need for a host name in a host list.
5629 Now, if $sender_host_name is referenced and the host name has not yet been
5630 looked up, a lookup is performed. If the lookup fails, the variable remains
5631 empty, and $host_lookup_failed is set to "1".
5633 19. Added "eqi" as a case-independent comparison operator.
5635 20. The saslauthd authentication condition could segfault if neither service
5636 nor realm was specified.
5638 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5639 error message that was logged was misleading, and incoming SMTP
5640 connections were dropped. The message is now more accurate, and temporary
5641 errors are given to SMTP connections.
5643 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5644 (except RSET) with a 421 error, until QUIT is received. However, it was
5645 failing to send a response to QUIT.
5647 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5648 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5649 ACL is now run after the helo_try_verify_hosts code.
5651 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5652 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5653 case-independent, so other case variants are also recognized). Apparently
5654 some systems use these upper case variants.
5656 25. If more than two messages were waiting for the same host, and a transport
5657 filter was specified for the transport, Exim sent two messages over the
5658 same TCP/IP connection, and then failed with "socket operation on non-
5659 socket" when it tried to send the third.
5661 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5663 27. The extern definition of crypt16() in expand.c was not being excluded when
5664 the OS had its own crypt16() function.
5666 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5667 as a preferred synonym for return_size_limit, both as an option and as an
5670 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5672 30. Changed the default configuration ACL to relax the local part checking rule
5673 for addresses that are not in any local domains. For these addresses,
5674 slashes and pipe symbols are allowed within local parts, but the sequence
5675 /../ is explicitly forbidden.
5677 31. SPA server authentication was not clearing the challenge buffer before
5680 32. log_message in a "warn" ACL statement was writing to the reject log as
5681 well as to the main log, which contradicts the documentation and doesn't
5682 seem right (because no rejection is happening). So I have stopped it.
5684 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5685 However, I am unable to do any testing of this.
5687 34. Fixed an infelicity in the appendfile transport. When checking directories
5688 for a mailbox, to see if any needed to be created, it was accidentally
5689 using path names with one or more superfluous leading slashes; tracing
5690 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5692 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5693 contents are added to the log line that is written for every discarded
5694 recipient. (Previously a log_message setting was ignored.)
5696 36. The ${quote: operator now quotes the string if it is empty.
5698 37. The install script runs exim in order to find its version number. If for
5699 some reason other than non-existence or emptiness, which it checks, it
5700 could not run './exim', it was installing it with an empty version number,
5701 i.e. as "exim-". This error state is now caught, and the installation is
5704 38. An argument was missing from the function that creates an error message
5705 when Exim fails to connect to the socket for saslauthd authentication.
5706 This could cause Exim to crash, or give a corrupted message.
5708 39. Added isip, isip4, and isip6 to ${if conditions.
5710 40. The ACL variables $acl_xx are now saved with the message, and can be
5711 accessed later in routers, transports, and filters.
5713 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5714 strings in the file are not string-expanded.
5716 42. If a MAIL command specified a SIZE value that was too large to fit into an
5717 int variable, the check against message_size_limit failed. Such values are
5718 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5719 day this will have to be increased, but I don't think I want to be around
5720 when emails are that large.
5727 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5728 iconv() is not standard in FreeBSD.
5730 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5731 IPv6 enabled. The observed symptom was a segmentation fault on return from
5732 the function os_common_find_running_interfaces() in src/os.c.
5734 3. In the check_special_case() function in daemon.c I had used "errno" as an
5735 argument name, which causes warnings on some systems. This was basically a
5736 typo, since it was named "eno" in the comments!
5738 4. The code that waits for the clock to tick (at a resolution of some fraction
5739 of a second) so as to ensure message-id uniqueness was always waiting for
5740 at least one whole tick, when it could have waited for less. [This is
5741 almost certainly not relevant at current processor speeds, where it is
5742 unlikely to ever wait at all. But we try to future-proof.]
5744 5. The function that sleeps for a time interval that includes fractions of a
5745 second contained a race. It did not block SIGALRM between setting the
5746 timer, and suspending (a couple of lines later). If the interval was short
5747 and the sigsuspend() was delayed until after it had expired, the suspension
5748 never ended. On busy systems this could lead to processes getting stuck for
5751 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5752 process, before it forks any delivery processes. The open lookup caching
5753 mechanism meant that the open file or database connection was passed into
5754 the delivery process. The problem was that delivery processes always tidy
5755 up cached lookup data. This could cause a problem for the next delivery
5756 process started by the queue runner, because the external queue runner
5757 process does not know about the closure. So the next delivery process
5758 still has data in the lookup cache. In the case of a file lookup, there was
5759 no problem because closing a file descriptor in a subprocess doesn't affect
5760 the parent. However, if the lookup was caching a connection to a database,
5761 the connection was closed, and the second delivery process was likely to
5762 see errors such as "PGSQL: query failed: server closed the connection
5763 unexpectedly". The problem has been fixed by closing all cached lookups
5764 in a queue runner before running a delivery process.
5766 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5767 seem to have the "const" qualifier which it has on other OS. I've
5770 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5771 *of the same type* (client or server) with the same public name that an
5772 error should be diagnosed.
5774 9. When Exim looked up a host name for an IP address, but failed to find the
5775 original IP address when looking up the host name (a safety check), it
5776 output the message "<ip address> does not match any IP for NULL", which was
5777 confusing, to say the least. The bug was that the host name should have
5778 appeared instead of "NULL".
5780 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5781 user that is built into the binary, and the -C or -D options is used, root
5782 privilege is dropped before the configuration file is read. In addition,
5783 logging is switched to stderr instead of the normal log files. If the
5784 configuration then re-defines the Exim user, the unprivileged environment
5785 is probably not what is expected, so Exim logs a panic warning message (but
5788 However, if deliver_drop_privilege is set, the unprivileged state may well
5789 be exactly what is intended, so the warning has been cut out in that case,
5790 and Exim is allowed to try to write to its normal log files.
5796 1. smtp_return_error_details was not giving details for temporary sender
5797 or receiver verification errors.
5799 2. Diagnose a configuration error if two authenticators have the same public
5802 3. Exim used not to create the message log file for a message until the first
5803 delivery attempt. This could be confusing when incoming messages were held
5804 for policy or load reasons. The message log file is now created at the time
5805 the message is received, and an initial "Received" line is written to it.
5807 4. The automatically generated man page for command line options had a minor
5808 bug that caused no ill effects; however, a more serious problem was that
5809 the procedure for building the man page automatically didn't always
5810 operate. Consequently, release 4.20 contains an out-of-date version. This
5811 shouldn't happen again.
5813 5. When building Exim with embedded Perl support, the script that builds the
5814 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5815 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5817 6. The freeze_tell option was not being used for messages that were frozen on
5818 arrival, either by an ACL or by local_scan().
5820 7. Added the smtp_incomplete_transaction log selector.
5822 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5823 was accepting AUTH without a new EHLO.
5825 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5826 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5829 10. Logging of TCP/IP connections (when configured) now happens in the main
5830 daemon process instead of the child process, so that the TCP/IP connection
5831 count is more accurate (but it can never be perfect).
5833 11. The use of "drop" in a nested ACL was not being handled correctly in the
5834 outer ACL. Now, if condition failure induced by the nested "drop" causes
5835 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5836 or "require"), the connection is dropped.
5838 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5839 that yield "discard" can now be used with an "accept" or a "discard" verb,
5840 but an error is generated for any others (because I can't see a useful way
5841 to define what should happen).
5843 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5844 are now processed in the same way as lines in the Exim configuration file.
5845 In particular, continuation lines are supported.
5847 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5849 15. Added -ti meaning -t -i.
5851 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5852 domains, and warn by logging if others are found.
5854 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5855 returned from GIFCONF when Exim is trying to find the list of interfaces on
5856 a host. The code in os.c has been modified to copy each ifreq to an aligned
5857 structure in all cases.
5859 Also, in some cases, the returned ifreq's were being copied to a 'struct
5860 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5861 means the last couple of bytes of an IPv6 address could be chopped if the
5862 ifreq contained only a normal sockaddr (14 bytes storage).
5864 18. Named domain lists were not supported in the hosts_treat_as_local option.
5865 An entry such as +xxxx was not recognized, and was treated as a literal
5868 19. Ensure that header lines added by a DATA ACL are included in the reject log
5869 if the ACL subsequently rejects the message.
5871 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5872 MD5 (which is deprecated).
5874 21. When testing a filter file using -bf, Exim was writing a message when it
5875 took the sender from a "From " line in the message, but it was not doing so
5876 when it took $return_path from a Return-Path: header line. It now does.
5878 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5879 with a valid header line field name (a series of printing characters
5880 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5882 23. Changed "disc" in the source to "disk" to conform to the documentation and
5883 the book and for uniformity.
5885 24. Ignore Sendmail's -Ooption=value command line item.
5887 25. When execve() failed while trying to run a command in a pipe transport,
5888 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5889 could be confused with a return value of 69 from the command itself. This
5890 has been changed to 127, the value the shell returns if it is asked to run
5891 a non-existent command. The wording for the related log line suggests a
5892 non-existent command as the problem.
5894 26. If received_header_text expands to an empty string, do not add a Received:
5895 header line to the message. (Well, it adds a token one on the spool, but
5896 marks it "old" so that it doesn't get used or transmitted.)
5898 27. Installed eximstats 1.28 (addition of -nt option).
5900 28. There was no check for failure on the call to getsockname() in the daemon
5901 code. This can fail if there is a shortage of resources on the system, with
5902 ENOMEM, for example. A temporary error is now given on failure.
5904 29. Contrary to the C standard, it seems that in some environments, the
5905 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5906 program. Exim now does this explicitly; it affects the formatting of
5907 timestamps using strftime().
5909 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5910 complaints. I've now initialized all the variables, to avoid this.
5912 32. Header lines added by a system filter were not being "seen" during
5913 transport-time rewrites.
5915 33. The info_callback() function passed to OpenSSL is set up with type void
5916 (*)(SSL *, int, int), as described somewhere. However, when calling the
5917 function (actually a macro) that sets it up, the type void(*)() is
5918 expected. I've put in a cast to prevent warnings from picky compilers.
5920 34. If a DNS black list lookup found a CNAME record, but there were no A
5921 records associated with the domain it pointed at, Exim crashed.
5923 35. If a DNS black list lookup returned more than one A record, Exim ignored
5924 all but the first. It now scans all returned addresses if a particular IP
5925 value is being sought. In this situation, the contents of the
5926 $dnslist_value variable are a list of all the addresses, separated by a
5929 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5930 to accept a host name and all its aliases if the forward lookup for any of
5931 them yielded the IP address of the incoming connection. Now it accepts only
5932 those names whose forward lookup yields the correct IP address. Any other
5933 names are discarded. This closes a loophole whereby a rogue DNS
5934 administrator could create reverse DNS records to break through a
5935 wildcarded host restriction in an ACL.
5937 37. If a user filter or a system filter that ran in a subprocess used any of
5938 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5939 the wrong values were passed to the pipe command ($thisaddress had the
5940 value of $0, $0 had the value of $1, etc). This bug was introduced by
5941 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5943 38. Improved the line breaking for long SMTP error messages from ACLs.
5944 Previously, if there was no break point between 40 and 75 characters, Exim
5945 left the rest of the message alone. Two changes have been made: (a) I've
5946 reduced the minimum length to 35 characters; (b) if it can't find a break
5947 point between 35 and 75 characters, it looks ahead and uses the first one
5948 that it finds. This may give the occasional overlong line, but at least the
5949 remaining text gets split now.
5951 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5952 file descriptors might be low, and that setting 1000 would always raise it.
5953 It turns out that in some environments, the limit is already over 1000 and
5954 that lowering it causes trouble. So now Exim takes care not to decrease it.
5956 40. When delivering a message, the value of $return_path is set to $sender_
5957 address at the start of routing (routers may change the value). By an
5958 oversight, this default was not being set up when an address was tested by
5959 -bt or -bv, which affected the outcome if any router or filter referred to
5962 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5963 log only when "message" or "log_message" are set. However, if one of the
5964 conditions was an address verification, or a call to a nested ACL, the
5965 messages generated by the underlying test were being passed through. This
5966 no longer happens. The underlying message is available in $acl_verify_
5967 message for both "message" and "log_message" expansions, so it can be
5968 passed through if needed.
5970 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5971 new expansion $bh_ to give the encoded byte string without charset
5972 translation. Translation happens only if iconv() is available; HAVE_ICONV
5973 indicates this at build time. HEADERS_CHARSET gives the charset to
5974 translate to; headers_charset can change it in the configuration, and
5975 "headers charset" can change it in an individual filter file.
5977 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5978 that creates RFC 2047 encoded "words" labels them as that charset instead
5979 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5980 expansion operator; (ii) when Exim creates a From: line for a local
5981 message; (iii) when a header line is rewritten to include a "phrase" part.
5983 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5984 buggy, causing it to skip the first lines of messages whose message ID
5985 ended in 'D'. This would not have bitten before Exim release 4.14, because
5986 message IDs were unlikely to end in 'D' before then. The effect was to have
5987 incorrect size information for certain domains.
5989 45. #include "config.h" was missing at the start of the crypt16.c module. This
5990 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5993 46. If there was a timeout during a "random" callout check, Exim treated it as
5994 a failure of the random address, and carried on sending RSET and the real
5995 address. If the delay was just some slowness somewhere, the response to the
5996 original RCPT would be taken as a response to RSET and so on, causing
5997 mayhem of various kinds.
5999 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6000 when I implemented it. It didn't allow for the fact that some option values
6001 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6002 the right test for positive values.
6004 48. Domain names in DNS records are case-independent. Exim always looks them up
6005 in lower case. Some resolvers return domain names in exactly the case they
6006 appear in the zone file, that is, they may contain uppercase letters. Not
6007 all resolvers do this - some return always lower case. Exim was treating a
6008 change of case by a resolver as a change of domain, similar to a widening
6009 of a domain abbreviation. This triggered its re-routing code and so it was
6010 trying to route what was effectively the same domain again. This normally
6011 caused routing to fail (because the router wouldn't handle the domain
6012 twice). Now Exim checks for this case specially, and just changes the
6013 casing of the domain that it ultimately uses when it transmits the message
6016 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6019 50. If a filter generated a file delivery with a non-absolute name (possible if
6020 no home directory exists for the router), the forbid_file option was not
6023 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6024 the existing equality matching.
6026 52. Exim was using ints instead of ino_t variables in some places where it was
6027 dealing with inode numbers.
6029 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6030 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6031 called TMPDIR, and if it finds it is different, it changes its value.
6033 54. The smtp_printf() function is now made available to local_scan() so
6034 additional output lines can be written before returning. There is also an
6035 smtp_fflush() function to enable the detection of a dropped connection.
6036 The variables smtp_input and smtp_batched_input are exported to
6039 55. Changed the default runtime configuration: the message "Unknown user"
6040 has been removed from the ACL, and instead placed on the localuser router,
6041 using the cannot_route_message feature. This means that any verification
6042 failures that generate their own messages won't get overridden. Similarly,
6043 the "Unrouteable address" message that was in the ACL for unverifiable
6044 relay addresses has also been removed.
6046 56. Added hosts_avoid_esmtp to the smtp transport.
6048 57. The exicyclog script was not checking for the esoteric option
6049 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6050 will work only if exicyclog is run under the appropriate euid.
6052 58. Following a discussion on the list, the rules by which Exim recognises line
6053 endings on incoming messages have been changed. The -dropcr and drop_cr
6054 options are now no-ops, retained only for backwards compatibility. The
6055 following line terminators are recognized: LF CRLF CR. However, special
6056 processing applies to CR:
6058 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6059 nor a local message in the state where . is a terminator.
6061 (ii) If a bare CR is encountered in a header line, an extra space is added
6062 after the line terminator so as not to end the header. The reasoning
6063 behind this is that bare CRs in header lines are most likely either
6064 to be mistakes, or people trying to play silly games.
6066 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6067 was being incorrectly given as 18 bytes larger than it should have been.
6068 This is a VOB (very old bug).
6070 60. This may never have affected anything current, but just in case it has:
6071 When the local host is found other than at the start of a list of hosts,
6072 the local host, those with the same MX, and any that follow, are discarded.
6073 When the list in question was part of a longer list of hosts, the following
6074 hosts (not currently being processed) were also being discarded. This no
6075 longer happens. I'm not sure if this situation could ever has previously
6078 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6081 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6082 header. This is something that is recommended in a new Internet Draft, and
6083 is something that is documented as being done by Sendmail. There are two
6084 possible values. For messages generated by the autoreply transport, Exim
6087 Auto-Submitted: auto-replied
6089 whereas for all other generated messages (e.g. bounces) it adds
6091 Auto-Submitted: auto-generated
6093 63. The "personal" condition in filters now includes a test for the
6094 Auto-Submitted: header. If it contains the string "auto-" the message it
6095 not considered personal.
6097 64. Added rcpt_include_affixes as a generic transport option.
6099 65. Added queue_only_override (default true).
6101 66. Added the syslog_duplication option.
6103 67. If what should have been the first header line of a message consisted of
6104 a space followed by a colon, Exim was mis-interpreting it as a header line.
6105 It isn't of course - it is syntactically invalid and should therefore be
6106 treated as the start of the message body. The misbehaviour could have
6107 caused a number of strange effects, including loss of data in subsequent
6108 header lines, and spool format errors.
6110 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6111 client host had authenticated. This control can now be exercised by an ACL
6112 for more flexibility.
6114 69. By default, callouts do not happen when testing with -bh. There is now a
6115 variant, -bhc, which does actually run the callout code, including
6116 consulting and updating the callout cache.
6118 70. Added support for saslauthd authentication, courtesy of Alexander
6121 71. If statvfs() failed on the spool or log directories while checking their
6122 size for availability, Exim confusingly gave the error "space shortage".
6123 Furthermore, in debugging mode it crashed with a floating point exception.
6124 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6125 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6126 serious problem, Exim now writes to the main and panic logs when this
6127 happens, with details of the failure. It then refuses to accept the
6128 incoming message, giving the message "spool directory problem" or "log
6129 directory problem" with a 421 code for SMTP messages.
6131 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6132 0, 1, and 2 exist, because some OS complain for execs without them (see
6133 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6134 descriptors. However, the code omitted to check that the open succeeded,
6135 causing mysterious errors if for some reason the permissions on /dev/null
6136 got screwed. Now Exim writes a message to the main and panic logs, and
6137 bombs out if it can't open /dev/null.
6139 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6140 interact so that it is all more flexible. It is supposed to remain
6141 backwards compatible. Also added extra_local_interfaces.
6143 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6144 to bomb out with an assertion failure - to the client this appears as a
6145 connection drop. This problem occurs in the part of the code that was taken
6146 from the Samba project. Fortunately, the assertion is in a very simple
6147 function, so I have fixed this by reproducing the function inline in the
6148 one place where it is called, and arranging for authentication to fail
6149 instead of killing the process with assert().
6151 75. The SPA client code was not working when the server requested OEM rather
6152 than Unicode encoding.
6154 76. Added code to make require_files with a specific uid setting more usable in
6155 the case where statting the file as root fails - usually a non-root-mounted
6156 NFS file system. When this happens and the failure is EACCES, Exim now
6157 forks a subprocess and does the per-uid checking as the relevant uid.
6159 77. Added process_log_path.
6161 78. If log_file_path was not explicitly set, a setting of check_log_space or
6162 check_log_inodes was ignored.
6164 79. If a space check for the spool or log partitions fails, the incident is now
6165 logged. Of course, in the latter case the data may get lost...
6167 80. Added the %p formatting code to string_format() so that it can be used to
6168 print addresses in debug_print(). Adjusted all the address printing in the
6169 debugging in store.c to use %p rather than %d.
6171 81. There was a concern that a line of code in smtp_in.c could overflow a
6172 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6173 initially expressed, the concern was not well-founded, because trailing
6174 spaces are removed early. However, if the trailing spaces were followed by
6175 a NULL, they did not get removed, so the overflow was possible. Two fixes
6178 (a) I re-wrote the offending code in a cleaner fashion.
6179 (b) If an incoming SMTP command contains a NULL character, it is rejected
6182 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6183 runs initgroups(), so that if the Exim user is in any additional groups,
6184 they will be used during message reception.
6190 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob