1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
15 TL/01 When running the test suite, the README says that variables such as
16 no_msglog_check are global and can be placed anywhere in a specific
17 test's script, however it was observed that placement needed to be near
18 the beginning for it to behave that way. Changed the runtest perl
19 script to read through the entire script once to detect and set these
20 variables, reset to the beginning of the script, and then run through
21 the script parsing/test process like normal.
23 TL/02 The BSD's have an arc4random API. One of the functions to induce
24 adding randomness was arc4random_stir(), but it has been removed in
25 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
26 function when detected.
28 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
29 cause callback expansion.
31 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
32 syntax errors in an expansion can be treated as a string instead of
33 logging or causing an error, due to the internal use of bool_lax
34 instead of bool when processing it.
36 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
37 server certificates when making smtp deliveries.
39 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
41 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
43 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
45 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
46 Merged patch from Sebastian Wiedenroth.
48 JH/05 Fix results-pipe from transport process. Several recipients, combined
49 with certificate use, exposed issues where response data items split
50 over buffer boundaries were not parsed properly. This eventually
51 resulted in duplicates being sent. This issue only became common enough
52 to notice due to the introduction of conection certificate information,
53 the item size being so much larger. Found and fixed by Wolfgang Breyha.
55 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
56 size buffer was used, resulting in syntax errors when an expansion
59 JH/07 Add support for directories of certificates when compiled with a GnuTLS
60 version 3.3.6 or later.
62 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
63 is EXPERIMENTAL_EVENT, the main-configuration and transport options
64 both become "event_action", the variables become $event_name, $event_data
65 and $event_defer_errno. There is a new variable $verify_mode, usable in
66 routers, transports and related events. The tls:cert event is now also
67 raised for inbound connections, if the main configuration event_action
70 TL/06 In test suite, disable OCSP for old versions of openssl which contained
71 early OCSP support, but no stapling (appears to be less than 1.0.0).
73 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
74 server certificate names available under the smtp transport option
75 "tls_verify_cert_hostname" now do not permit multi-component wildcard
78 JH/10 Time-related extraction expansions from certificates now use the main
79 option "timezone" setting for output formatting, and are consistent
80 between OpenSSL and GnuTLS compilations. Bug 1541.
82 JH/11 Fix a crash in mime ACL when meeting a zero-length parameter in the
85 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
86 include certificate info, eximon was claiming there were spoolfile
89 JH/13 Buf 1521: Fix ldap lookup for single-attr request, multiple-attr return.
91 JH/14 Log delivery-related information more consistently, using the sequence
92 "H=<name> [<ip>]" wherever possible.
97 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
98 checkers that were complaining about end of non-void function with no
101 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
102 This was a regression intruduced in 4.83 by another bugfix.
104 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
106 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
107 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
113 TF/01 Correctly close the server side of TLS when forking for delivery.
115 When a message was received over SMTP with TLS, Exim failed to clear up
116 the incoming connection properly after forking off the child process to
117 deliver the message. In some situations the subsequent outgoing
118 delivery connection happened to have the same fd number as the incoming
119 connection previously had. Exim would try to use TLS and fail, logging
120 a "Bad file descriptor" error.
122 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
123 utilities have not been installed.
125 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
126 temporary space as the ACL may create new global variables.
128 TL/01 LDAP support uses per connection or global context settings, depending
129 upon the detected version of the libraries at build time.
131 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
132 to extract and use the src ip:port in logging and expansions as if it
133 were a direct connection from the outside internet. PPv2 support was
134 updated based on HAProxy spec change in May 2014.
136 JH/02 Add ${listextract {number}{list}{success}{fail}}.
138 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
139 Properly escape header and check for NULL return.
141 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
144 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
146 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
147 characters in header names, implemented as a verify condition.
148 Contributed by Michael Fischer v. Mollard.
150 TL/05 Rename SPF condition results err_perm and err_temp to standardized
151 results permerror and temperror. Previous values are deprecated but
152 still accepted. In a future release, err_perm and err_temp will be
153 completely removed, which will be a backward incompatibility if the
154 ACL tests for either of these two old results. Patch contributed by
155 user bes-internal on the mailing list.
157 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
159 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
160 selectors, in both main and reject logs.
162 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
165 JH/07 Add malware type "sock" for talking to simple daemon.
167 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
169 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
170 routers/transports under cutthrough routing.
172 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
173 numbers. Touch up "bool" conditional to keep the same definition.
175 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
177 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
179 JH/12 Expand items in router/transport headers_add or headers_remove lists
180 individually rather than the list as a whole. Bug 1452.
182 Required for reasonable handling of multiple headers_ options when
183 they may be empty; requires that headers_remove items with embedded
184 colons must have them doubled (or the list-separator changed).
186 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
187 view the policy declared in the DMARC record. Currently, $dmarc_status
188 is a combined value of both the record presence and the result of the
191 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
193 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
194 dnslookup router and the smtp transport (applying to the forward
197 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
198 of ldap servers used for a specific lookup. Patch provided by Heiko
201 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
202 New variable $lookup_dnssec_authenticated for observability.
204 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
205 Patch submitted by Lars Timman.
207 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
209 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
210 Requires trusted mode and valid format message id, aborts otherwise.
211 Patch contributed by Heiko Schlichting.
213 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
214 certextract with support for various fields. Bug 1358.
216 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
217 is requested by default, modifiable by smtp transport option
220 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
221 operate on certificate variables to give certificate fingerprints
222 Also new ${sha256:cert_variable}.
224 JH/23 The PRDR feature is moved from being Experimental into the mainline.
226 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
227 Christian Aistleitner.
229 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
231 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
232 file. Patch from Wolfgang Breyha.
234 JH/25 Expand the coverage of the delivery $host and $host_address to
235 client authenticators run in verify callout. Bug 1476.
237 JH/26 Port service names are now accepted for tls_on_connect_ports, to
238 align with daemon_smtp_ports. Bug 72.
240 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
241 support and error reporting did not work properly.
243 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
244 and is readable. Patch from Andrew Colin Kissa.
246 TL/14 Enhance documentation of ${run expansion and how it parses the
247 commandline after expansion, particularly in the case when an
248 unquoted variable expansion results in an empty value.
250 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
252 PP/02 Fix internal collision of T_APL on systems which support RFC3123
253 by renaming away from it. Addresses GH issue 15, reported by
256 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
258 TL/15 SECURITY: prevent double expansion in math comparison functions
259 (can expand unsanitized data). Not remotely exploitable.
266 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
268 PP/02 Make -n do something, by making it not do something.
269 When combined with -bP, the name of an option is not output.
271 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
274 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
275 $sender_host_name and config options to manage this, and basic check
278 PP/05 DSCP support for outbound connections and control modifier for inbound.
280 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
281 (Only plugin which currently uses this is kerberos4, which nobody should
282 be using, but we should make it available and other future plugins might
283 conceivably use it, even though it would break NAT; stuff *should* be
284 using channel bindings instead).
286 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
287 name; added for Sendmail compatibility; requires admin caller.
288 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
289 just ignore it); requires trusted caller.
290 Also parse but ignore: -Ac -Am -X<logfile>
293 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
295 TL/02 Add +smtp_confirmation as a default logging option.
297 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
298 Patch by Magnus Holmgren from 2007-02-20.
300 TL/04 Bugzilla 1281 - Spec typo.
301 Bugzilla 1283 - Spec typo.
302 Bugzilla 1290 - Spec grammar fixes.
304 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
306 TL/06 Add Experimental DMARC support using libopendmarc libraries.
308 TL/07 Fix an out of order global option causing a segfault. Reported to dev
309 mailing list by by Dmitry Isaikin.
311 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
313 JH/02 Support "G" suffix to numbers in ${if comparisons.
315 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
317 NM/01 Bugzilla 1197 - Spec typo
318 Bugzilla 1196 - Spec examples corrections
320 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
322 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
323 gnutls_enable_pkcs11, but renamed to more accurately indicate its
326 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
327 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
329 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
330 "acl {{name}{arg}...}", and optional args on acl condition
333 JH/05 Permit multiple router/transport headers_add/remove lines.
335 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
337 JH/07 Avoid using a waiting database for a single-message-only transport.
338 Performance patch from Paul Fisher. Bugzilla 1262.
340 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
343 JH/09 Add $headers_added variable, with content from use of ACL modifier
344 add_header (but not yet added to the message). Bugzilla 199.
346 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
347 Pulled from Bugzilla 817 by Wolfgang Breyha.
349 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
351 (nb: this is the same fix as in Exim 4.80.1)
353 JH/11 Add A= logging on delivery lines, and a client_set_id option on
356 JH/12 Add optional authenticated_sender logging to A= and a log_selector
359 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
361 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
362 advertise SMTP AUTH mechanism to us, instead of a generic
363 protocol violation error. Also, make Exim more robust to bad
364 data from the Dovecot auth socket.
366 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
368 When a queue runner is handling a message, Exim first routes the
369 recipient addresses, during which it prunes them based on the retry
370 hints database. After that it attempts to deliver the message to
371 any remaining recipients. It then updates the hints database using
374 So if a recipient address works intermittently, it can get repeatedly
375 deferred at routing time. The retry hints record remains fresh so the
376 address never reaches the final cutoff time.
378 This is a fairly common occurrence when a user is bumping up against
379 their storage quota. Exim had some logic in its local delivery code
380 to deal with this. However it did not apply to per-recipient defers
381 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
383 This change adds a proper retry rule check during routing so that the
384 final cutoff time is checked against the message's age. We only do
385 this check if there is an address retry record and there is not a
386 domain retry record; this implies that previous attempts to handle
387 the address had the retry_use_local_parts option turned on. We use
388 this as an approximation for the destination being like a local
389 delivery, as in LMTP.
391 I suspect this new check makes the old local delivery cutoff check
392 redundant, but I have not verified this so I left the code in place.
394 TF/02 Correct gecos expansion when From: is a prefix of the username.
396 Test 0254 submits a message to Exim with the header
400 When I ran the test suite under the user fanf2, Exim expanded
401 the header to contain my full name, whereas it should have added
402 a Resent-Sender: header. It erroneously treats any prefix of the
403 username as equal to the username.
405 This change corrects that bug.
407 GF/01 DCC debug and logging tidyup
408 Error conditions log to paniclog rather than rejectlog.
409 Debug lines prefixed by "DCC: " to remove any ambiguity.
411 TF/03 Avoid unnecessary rebuilds of lookup-related code.
413 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
414 Bug spotted by Jeremy Harris; was flawed since initial commit.
415 Would have resulted in OCSP responses post-SNI triggering an Exim
416 NULL dereference and crash.
418 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
420 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
421 Bug detection, analysis and fix by Samuel Thibault.
422 Bugzilla 1331, Debian bug #698092.
424 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
426 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
427 Server implementation by Todd Lyons, client by JH.
428 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
429 config variable "prdr_enable" controls whether the server
430 advertises the facility. If the client requests PRDR a new
431 acl_data_smtp_prdr ACL is called once for each recipient, after
432 the body content is received and before the acl_smtp_data ACL.
433 The client is controlled by bolth of: a hosts_try_prdr option
434 on the smtp transport, and the server advertisement.
435 Default client logging of deliveries and rejections involving
436 PRDR are flagged with the string "PRDR".
438 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
439 fclose(). Diagnosis by Todd Lyons.
441 PP/17 Update configure.default to handle IPv6 localhost better.
442 Patch by Alain Williams (plus minor tweaks).
445 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
446 This is now consistent with GnuTLS, and is now documented: the
447 previous undocumented portable approach to treating the option as
448 unset was to force an expansion failure. That still works, and
449 an empty string is now equivalent.
451 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
452 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
453 not performing validation itself.
455 PP/20 Added force_command boolean option to pipe transport.
456 Patch from Nick Koston, of cPanel Inc.
458 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
461 TF/04 Added udpsend ACL modifer and hexquote expansion operator
463 PP/21 Fix eximon continuous updating with timestamped log-files.
464 Broken in a format-string cleanup in 4.80, missed when I repaired the
465 other false fix of the same issue.
466 Report and fix from Heiko Schlichting.
469 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
470 Report from Prashanth Katuri.
472 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
473 It's SecureTransport, so affects any MacOS clients which use the
474 system-integrated TLS libraries, including email clients.
476 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
477 using a MIME ACL for non-SMTP local injection.
478 Report and assistance in diagnosis by Warren Baker.
480 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
482 JH/16 Fix comparisons for 64b. Bugzilla 1385.
484 TL/09 Add expansion variable $authenticated_fail_id to keep track of
485 last id that failed so it may be referenced in subsequent ACL's.
487 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
490 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
491 ldap library initialization, allowing self-signed CA's to be
492 used. Also properly sets require_cert option later in code by
493 using NULL (global ldap config) instead of ldap handle (per
494 session). Bug diagnosis and testing by alxgomz.
496 TL/12 Enhanced documentation in the ratelimit.pl script provided in
497 the src/util/ subdirectory.
499 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
500 renamed to Transport Post Delivery Action by Jeremy Harris, as
503 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
504 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
505 redis_servers = needs to be configured which will be used by the redis
506 lookup. Patch from Warren Baker, of The Packet Hub.
508 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
510 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
511 hostname or reverse DNS when processing a host list. Used suggestions
512 from multiple comments on this bug.
514 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
516 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
517 Missed a few lines, added it to make the runtest require no keyboard
520 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
521 contains upper case chars. Make router use caseful_local_part.
523 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
524 support when GnuTLS has been built with p11-kit.
530 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
532 This, or similar/improved, will also be change PP/11 of 4.82.
538 PP/01 Handle short writes when writing local log-files.
539 In practice, only affects FreeBSD (8 onwards).
540 Bugzilla 1053, with thanks to Dmitry Isaikin.
542 NM/01 Bugzilla 949 - Documentation tweak
544 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
547 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
549 PP/02 Implemented gsasl authenticator.
551 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
553 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
554 `pkg-config foo` for cflags/libs.
556 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
557 with rest of GSASL and with heimdal_gssapi.
559 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
560 `pkg-config foo` for cflags/libs for the TLS implementation.
562 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
563 properties get this fed in as external SSF. A number of robustness
564 and debugging improvements to the cyrus_sasl authenticator.
566 PP/08 cyrus_sasl server now expands the server_realm option.
568 PP/09 Bugzilla 1214 - Log authentication information in reject log.
569 Patch by Jeremy Harris.
571 PP/10 Added dbmjz lookup type.
573 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
575 PP/12 MAIL args handles TAB as well as SP, for better interop with
576 non-compliant senders.
577 Analysis and variant patch by Todd Lyons.
579 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
580 Bug report from Lars Müller <lars@samba.org> (via SUSE),
581 Patch from Dirk Mueller <dmueller@suse.com>
583 PP/13 tls_peerdn now print-escaped for spool files.
584 Observed some $tls_peerdn in wild which contained \n, which resulted
585 in spool file corruption.
587 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
588 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
589 or write after TLS renegotiation, which otherwise led to messages
592 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
593 as a tracking header (ie: a signed header comes before the signature).
594 Patch from Wolfgang Breyha.
596 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
597 comma-sep list; embedded commas doubled.
599 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
601 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
603 Report and patch from Dmitry Banschikov.
605 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
606 Removed SSL_clear() after SSL_new() which led to protocol negotiation
607 failures. We appear to now support TLS1.1+ with Exim.
609 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
610 lets Exim select keys and certificates based upon TLS SNI from client.
611 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
612 before an outbound SMTP session. New log_selector, +tls_sni.
614 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
615 NULL dereference. Report and patch from Alun Jones.
617 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
618 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
619 Not seeing resolver debug output on NetBSD, but suspect this is a
620 resolver implementation change.
622 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
623 Left warnings. Added "eximon gdb" invocation mode.
625 PP/21 Defaulting "accept_8bitmime" to true, not false.
627 PP/22 Added -bw for inetd wait mode support.
629 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
630 locate the relevant includes and libraries. Made this the default.
632 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
633 Bugzilla 1246, report and most of solution from Tomasz Kusy.
635 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
636 This may cause build issues on older platforms.
638 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
639 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
640 gnutls_require_mac & gnutls_require_protocols (no longer supported).
641 Added SNI support via GnuTLS too.
642 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
644 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
646 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
647 Applied second patch from Janne, implementing suggestion to default
648 multiple-strings-in-record handling to match SPF spec.
650 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
652 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
653 read-only, out of scope).
654 Patch from Wolfgang Breyha, report from Stuart Northfield.
656 PP/29 Fix three issues highlighted by clang analyser static analysis.
657 Only crash-plausible issue would require the Cambridge-specific
658 iplookup router and a misconfiguration.
659 Report from Marcin Mirosław.
661 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
663 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
664 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
665 As part of this, removing so much warning spew let me fix some minor
666 real issues in debug logging.
668 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
669 assignment on my part. Fixed.
671 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
672 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
673 Janne Snabb (who went above and beyond: thank you).
675 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
676 string otherwise requires a connection and a bunch more work and it's
677 relatively easy to get wrong. Should also expose TLS library linkage
680 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
681 64-bit ${eval} (JH/03).
683 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
684 GNU libc to support some of the 64-bit stuff, should not lead to
685 conflicts. Defined before os.h is pulled in, so if a given platform
686 needs to override this, it can.
688 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
689 protection layer was required, which is not implemented.
690 Bugzilla 1254, patch from Wolfgang Breyha.
692 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
693 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
694 tls_dhparam take prime identifiers. Also unbreak combination of
695 OpenSSL+DH_params+TLSSNI.
697 PP/39 Disable SSLv2 by default in OpenSSL support.
703 PP/01 Solaris build fix for Oracle's LDAP libraries.
704 Bugzilla 1109, patch from Stephen Usher.
706 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
708 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
711 TF/02 Fix a couple more cases where we did not log the error message
712 when unlink() failed. See also change 4.74-TF/03.
714 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
715 lock up or crash if it happened to be inside a call to libc when it
716 got a SIGUSR1 from exiwhat.
718 The SIGUSR1 handler appends the current process status to the process
719 log which is later printed by exiwhat. It used to use the general
720 purpose logging code to do this, but several functions it calls are
721 not safe for signals.
723 The new output code in the SIGUSR1 handler is specific to the process
724 log, and simple enough that it's easy to inspect for signal safety.
725 Removing some special cases also simplifies the general logging code.
726 Removing the spurious timestamps from the process log simplifies
729 TF/04 Improved ratelimit ACL condition.
731 The /noupdate option has been deprecated in favour of /readonly which
732 has clearer semantics. The /leaky, /strict, and /readonly update modes
733 are mutually exclusive. The update mode is no longer included in the
734 database key; it just determines when the database is updated. (This
735 means that when you upgrde Exim will forget old rate measurements.)
737 Exim now checks that the per_* options are used with an update mode that
738 makes sense for the current ACL. For example, when Exim is processing a
739 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
740 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
741 must specify per_mail/readonly. If you omit the update mode it defaults to
742 /leaky where that makes sense (as before) or /readonly where required.
744 The /noupdate option is now undocumented but still supported for
745 backwards compatibility. It is equivalent to /readonly except that in
746 ACLs where /readonly is required you may specify /leaky/noupdate or
747 /strict/noupdate which are treated the same as /readonly.
749 A useful new feature is the /count= option. This is a generalization
750 of the per_byte option, so that you can measure the throughput of other
751 aggregate values. For example, the per_byte option is now equivalent
752 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
754 The per_rcpt option has been generalized using the /count= mechanism
755 (though it's more complicated than the per_byte equivalence). When it is
756 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
757 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
758 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
759 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
760 Note that using per_rcpt with a non-readonly update mode in more than
761 one ACL will cause the recipients to be double-counted. (The per_mail
762 and per_byte options don't have this problem.)
764 The handling of very low rates has changed slightly. If the computed rate
765 is less than the event's count (usually one) then this event is the first
766 after a long gap. In this case the rate is set to the same as this event's
767 count, so that the first message of a spam run is counted properly.
769 The major new feature is a mechanism for counting the rate of unique
770 events. The new per_addr option counts the number of different
771 recipients that someone has sent messages to in the last time period. It
772 behaves like per_rcpt if all the recipient addresses are different, but
773 duplicate recipient addresses do not increase the measured rate. Like
774 the /count= option this is a general mechanism, so the per_addr option
775 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
776 example, measure the rate that a client uses different sender addresses
777 with the options per_mail/unique=$sender_address. There are further
778 details in the main documentation.
780 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
782 TF/06 Removed a few PCRE remnants.
784 TF/07 Automatically extract Exim's version number from tags in the git
785 repository when doing development or release builds.
787 PP/02 Raise smtp_cmd_buffer_size to 16kB.
788 Bugzilla 879. Patch from Paul Fisher.
790 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
791 Heavily based on revision 40f9a89a from Simon Arlott's tree.
794 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
796 PP/05 Variable $av_failed, true if the AV scanner deferred.
797 Bugzilla 1078. Patch from John Horne.
799 PP/06 Stop make process more reliably on build failure.
800 Bugzilla 1087. Patch from Heiko Schlittermann.
802 PP/07 Make maildir_use_size_file an _expandable_ boolean.
803 Bugzilla 1089. Patch from Heiko Schlittermann.
805 PP/08 Handle ${run} returning more data than OS pipe buffer size.
806 Bugzilla 1131. Patch from Holger Weiß.
808 PP/09 Handle IPv6 addresses with SPF.
809 Bugzilla 860. Patch from Wolfgang Breyha.
811 PP/10 GnuTLS: support TLS 1.2 & 1.1.
813 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
816 PP/11 match_* no longer expand right-hand-side by default.
817 New compile-time build option, EXPAND_LISTMATCH_RHS.
818 New expansion conditions, "inlist", "inlisti".
820 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
822 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
824 PP/14 fix log_write() format string regression from TF/03.
825 Bugzilla 1152. Patch from Dmitry Isaikin.
831 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
833 PP/02 Harmonised TLS library version reporting; only show if debugging.
834 Layout now matches that introduced for other libraries in 4.74 PP/03.
836 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
838 PP/04 New "dns_use_edns0" global option.
840 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
843 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
844 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
846 TK/01 Updated PolarSSL code to 0.14.2.
847 Bugzilla 1097. Patch from Andreas Metzler.
849 PP/07 Catch divide-by-zero in ${eval:...}.
852 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
855 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
856 format-string attack -- SECURITY: remote arbitrary code execution.
858 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
859 time unintentionally subject to list matching rules, letting the header
860 cause arbitrary Exim lookups (of items which can occur in lists, *not*
861 arbitrary string expansion). This allowed for information disclosure.
863 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
864 INT_MIN/-1 -- value coerced to INT_MAX.
870 NM/01 Workround for PCRE version dependency in version reporting
873 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
874 This fixes portability to compilers other than gcc, notably
875 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
877 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
878 makefiles for portability to HP-UX and POSIX correctness.
880 PP/01 Permit LOOKUP_foo enabling on the make command-line.
881 Also via indented variable definition in the Makefile.
882 (Debugging by Oliver Heesakkers).
884 PP/02 Restore caching of spamd results with expanded spamd_address.
885 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
887 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
888 Improves build reliability. Fix from: Frank Elsner
890 NM/02 Fix wide character breakage in the rfc2047 coding
891 Fixes bug 1064. Patch from Andrey N. Oktyabrski
893 NM/03 Allow underscore in dnslist lookups
894 Fixes bug 1026. Patch from Graeme Fowler
896 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
897 Code patches from Adam Ciarcinski of NetBSD.
899 NM/04 Fixed exiqgrep to cope with mailq missing size issue
902 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
903 is logged, to avoid truncation. Patch from John Horne.
905 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
906 Patch from Jakob Hirsch.
908 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
909 SQL string expansion failure details.
910 Patch from Andrey Oktyabrski.
912 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
913 Patch from Simon Arlott.
915 PP/09 New lookups functionality failed to compile on old gcc which rejects
916 extern declarations in function scope.
917 Patch from Oliver Fleischmann
919 PP/10 Use sig_atomic_t for flags set from signal handlers.
920 Check getgroups() return and improve debugging.
921 Fixed developed for diagnosis in bug 927 (which turned out to be
924 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
925 Patch from Mark Zealey.
927 PP/12 Bugzilla 1056: Improved spamd server selection.
928 Patch from Mark Zealey.
930 PP/13 Bugzilla 1086: Deal with maildir quota file races.
931 Based on patch from Heiko Schlittermann.
933 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
934 Patch from Uwe Doering, sign-off by Michael Haardt.
936 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
937 variable declaration deep within a block. Bug and patch from
940 PP/15 lookups-Makefile IRIX compatibilty coercion.
942 PP/16 Make DISABLE_DKIM build knob functional.
944 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
945 Patch by Simon Arlott
947 TF/03 Fix valgrind.h portability to C89 compilers that do not support
948 variable argument macros. Our copy now differs from upstream.
954 TF/01 Failure to get a lock on a hints database can have serious
955 consequences so log it to the panic log.
957 TF/02 Log LMTP confirmation messages in the same way as SMTP,
958 controlled using the smtp_confirmation log selector.
960 TF/03 Include the error message when we fail to unlink a spool file.
962 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
963 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
964 for maintaining out-of-tree patches for some time.
966 PP/01 Bugzilla 139: Documentation and portability issues.
967 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
968 Handle per-OS dynamic-module compilation flags.
970 PP/02 Let /dev/null have normal permissions.
971 The 4.73 fixes were a little too stringent and complained about the
972 permissions on /dev/null. Exempt it from some checks.
973 Reported by Andreas M. Kirchwitz.
975 PP/03 Report version information for many libraries, including
976 Exim version information for dynamically loaded libraries. Created
977 version.h, now support a version extension string for distributors
978 who patch heavily. Dynamic module ABI change.
980 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
981 privilege escalation vulnerability whereby the Exim run-time user
982 can cause root to append content of the attacker's choosing to
985 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
988 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
989 If dropping privileges for untrusted macros, we disabled normal logging
990 on the basis that it would fail; for the Exim run-time user, this is not
991 the case, and it resulted in successful deliveries going unlogged.
992 Fixed. Reported by Andreas Metzler.
998 PP/01 Date: & Message-Id: revert to normally being appended to a message,
999 only prepend for the Resent-* case. Fixes regression introduced in
1000 Exim 4.70 by NM/22 for Bugzilla 607.
1002 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1003 increasing numbers of administrators be bitten by this.
1005 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1007 PP/03 Bugzilla 994: added openssl_options main configuration option.
1009 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1011 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1013 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1015 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1016 without a peer certificate, leading to a segfault because of an
1017 assumption that peers always have certificates. Be a little more
1018 paranoid. Problem reported by Martin Tscholak.
1020 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1021 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1022 NB: ClamAV planning to remove STREAM in "middle of 2010".
1023 CL also introduces -bmalware, various -d+acl logging additions and
1024 more caution in buffer sizes.
1026 PP/09 Implemented reverse_ip expansion operator.
1028 PP/10 Bugzilla 937: provide a "debug" ACL control.
1030 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1032 PP/12 Bugzilla 973: Implement --version.
1034 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1036 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1038 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1040 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1041 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1042 ignore trailing whitespace.
1044 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1046 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1049 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1050 Notification from Dr Andrew Aitchison.
1052 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1053 ExtendedDetectionInfo response format.
1054 Notification from John Horne.
1056 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1059 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1060 XSL and documented dependency on system catalogs, with examples of how
1063 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1066 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1067 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1068 configuration file which is writeable by the Exim user or group.
1070 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1071 of configuration files to cover files specified with the -C option if
1072 they are going to be used with root privileges, not just the default
1075 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1076 option (effectively making it always true).
1078 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1079 files to be used while preserving root privileges.
1081 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1082 that rogue child processes cannot use them.
1084 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1085 run-time user, instead of root.
1087 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1088 Exim run-time user without dropping privileges.
1090 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1091 result string, instead of calling string_vformat() twice with the same
1094 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1095 for other users. Others should always drop root privileges if they use
1096 -C on the command line, even for a whitelisted configure file.
1098 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1100 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1106 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1107 $data_path, and $header_path variables; fixed documentation bugs and
1110 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1111 exipick to access non-standard spools, including the "frozen" queue
1114 NM/01 Bugzilla 965: Support mysql stored procedures.
1115 Patch from Alain Williams
1117 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1119 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1120 Patch from Andreas Metzler
1122 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1123 Patch from Kirill Miazine
1125 NM/05 Bugzilla 671: Added umask to procmail example.
1127 JJ/03 installed exipick 20100323.0, fixing doc bug
1129 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1130 directory. Notification and patch from Dan Rosenberg.
1132 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1134 TK/02 Improve log output when DKIM signing operation fails.
1136 MH/01 Treat the transport option dkim_domain as a colon separated
1137 list, not as a single string, and sign the message with each element,
1138 omitting multiple occurences of the same signer.
1140 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1141 Bugzilla 985, 986. Patch by Simon Arlott
1143 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1144 Patch by Simon Arlott
1146 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1147 MBX locking. Notification from Dan Rosenberg.
1153 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1155 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1157 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1159 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1161 NM/04 Bugzilla 915: Flag broken perl installation during build.
1167 TK/01 Added patch by Johannes Berg that expands the main option
1168 "spamd_address" if it starts with a dollar sign.
1170 TK/02 Write list of recipients to X-Envelope-Sender header when building
1171 the mbox-format spool file for content scanning (suggested by Jakob
1174 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1175 (http://www.dcc-servers.net/) support via dccifd. Activated by
1176 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1178 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1179 by Mark Daniel Reidel <mr@df.eu>.
1181 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1182 When building exim an external PCRE library is now needed -
1183 PCRE is a system library on the majority of modern systems.
1184 See entry on PCRE_LIBS in EDITME file.
1186 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1187 conversation. Added nologin parameter to request.
1188 Patch contributed by Kirill Miazine.
1190 TF/01 Do not log submission mode rewrites if they do not change the address.
1192 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1194 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1195 log files in place. Contributed by Roberto Lima.
1197 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1199 TF/03 Bugzilla 615: When checking the local_parts router precondition
1200 after a local_part_suffix or local_part_prefix option, Exim now
1201 does not use the address's named list lookup cache, since this
1202 contains cached lookups for the whole local part.
1204 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1205 Robert Millan. Documentation is in experimental-spec.txt.
1207 TF/04 Bugzilla 668: Fix parallel build (make -j).
1209 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1211 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1212 Patch provided by Jan Srzednicki.
1214 TF/05 Leading white space used to be stripped from $spam_report which
1215 wrecked the formatting. Now it is preserved.
1217 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1218 that they are available at delivery time.
1220 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1222 TF/08 TLS error reporting now respects the incoming_interface and
1223 incoming_port log selectors.
1225 TF/09 Produce a more useful error message if an SMTP transport's hosts
1226 setting expands to an empty string.
1228 NM/06 Bugzilla 744: EXPN did not work under TLS.
1229 Patch provided by Phil Pennock.
1231 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1232 Patch provided by Richard Godbee.
1234 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1235 acl_smtp_notquit, added index entry.
1237 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1238 Patch provided by Eugene Bujak.
1240 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1241 accept(). Patch provided by Maxim Dounin.
1243 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1244 Patch provided by Phil Pennock.
1246 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1248 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1249 Patch provided by Brad "anomie" Jorsch.
1251 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1252 Patch provided by Dean Brooks.
1254 TK/05 Add native DKIM support (does not depend on external libraries).
1256 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1257 Patch provided by Graeme Fowler.
1259 NM/16 Bugzilla 851: Documentation example syntax fix.
1261 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1263 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1266 NM/19 Bugzilla 745: TLS version reporting.
1267 Patch provided by Phil Pennock.
1269 NM/20 Bugzilla 167: bool: condition support.
1270 Patch provided by Phil Pennock.
1272 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1273 clients. Patch provided by Phil Pennock.
1275 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1276 Patch provided by Brad "anomie" Jorsch.
1278 NM/23 Bugzilla 687: Fix misparses in eximstats.
1279 Patch provided by Heiko Schlittermann.
1281 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1282 Patch provided by Heiko Schlittermann.
1284 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1285 plus update to original patch.
1287 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1289 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1290 Patch provided by David Brownlee.
1292 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1294 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1296 NM/30 Bugzilla 888: TLS documentation bugfixes.
1298 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1300 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1301 Unlike the original bugzilla I have changed all shell scripts in src tree.
1303 NM/33 Bugzilla 898: Transport filter timeout fix.
1304 Patch by Todd Rinaldo.
1306 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1307 Patch by Serge Demonchaux.
1309 NM/35 Bugzilla 39: Base64 decode bug fixes.
1310 Patch by Jakob Hirsch.
1312 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1314 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1316 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1318 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1324 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1325 ALT-N's libdkim that I have put here:
1326 http://duncanthrax.net/exim-experimental/
1328 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1329 were called 'true' and it seems that C99 defines that as a reserved
1330 keyword to be used with 'bool' variable types. That means you could
1331 not include C99-style headers which use bools without triggering
1332 build errors in sieve.c.
1334 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1335 as mailq or other aliases. Changed the --help handling significantly
1336 to do whats expected. exim_usage() emits usage/help information.
1338 SC/01 Added the -bylocaldomain option to eximstats.
1340 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1342 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1344 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1350 PH/01 Another patch from the Sieve maintainer.
1352 PH/02 When an IPv6 address is converted to a string for single-key lookup
1353 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1354 dots are used instead of colons so that keys in lsearch files need not
1355 contain colons. This was done some time before quoting was made available
1356 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1357 (notated using the quote facility) so as to distinguish them from IPv4
1358 keys. This meant that lookups for IP addresses in host lists did not work
1359 for iplsearch lookups.
1361 This has been fixed by arranging for IPv6 addresses to be expressed with
1362 colons if the lookup type is iplsearch. This is not incompatible, because
1363 previously such lookups could never work.
1365 The situation is now rather anomolous, since one *can* have colons in
1366 ordinary lsearch keys. However, making the change in all cases is
1367 incompatible and would probably break a number of configurations.
1369 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1372 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1373 conversion specification without a maximum field width, thereby enabling
1374 a rogue spamd server to cause a buffer overflow. While nobody in their
1375 right mind would setup Exim to query an untrusted spamd server, an
1376 attacker that gains access to a server running spamd could potentially
1377 exploit this vulnerability to run arbitrary code as the Exim user.
1379 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1380 $primary_hostname instead of what libspf2 thinks the hosts name is.
1382 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1383 a directory entry by the name of the lookup key. Previously, if a
1384 symlink pointed to a non-existing file or a file in a directory that
1385 Exim lacked permissions to read, a lookup for a key matching that
1386 symlink would fail. Now it is enough that a matching directory entry
1387 exists, symlink or not. (Bugzilla 503.)
1389 PH/03 The body_linecount and body_zerocount variables are now exported in the
1392 PH/04 Added the $dnslist_matched variable.
1394 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1395 This means they are set thereafter only if the connection becomes
1398 PH/06 Added the client_condition to authenticators so that some can be skipped
1399 by clients under certain conditions.
1401 PH/07 The error message for a badly-placed control=no_multiline_responses left
1402 "_responses" off the end of the name.
1404 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1406 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1407 (without spaces) instead of just copying the configuration text.
1409 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1411 PH/11 Added $max_received_linelength.
1413 PH/12 Added +ignore_defer and +include_defer to host lists.
1415 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1416 way in which PCRE > 7.0 is built.
1418 PH/14 Implemented queue_only_load_latch.
1420 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1421 MAIL command. The effect was to mangle the value on 64-bit systems.
1423 PH/16 Another patch from the Sieve maintainer.
1425 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1427 PH/18 If a system quota error occurred while trying to create the file for
1428 a maildir delivery, the message "Mailbox is full" was not appended to the
1429 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1430 only to a quota excession during the actual writing of the file.
1432 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1433 characters?) which causes problems in log lines. The DN values are now
1434 passed through string_printing() before being added to log lines.
1436 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1437 and InterBase are left for another time.)
1439 PH/21 Added message_body_newlines option.
1441 PH/22 Guard against possible overflow in moan_check_errorcopy().
1443 PH/23 POSIX allows open() to be a macro; guard against that.
1445 PH/24 If the recipient of an error message contained an @ in the local part
1446 (suitably quoted, of course), incorrect values were put in $domain and
1447 $local_part during the evaluation of errors_copy.
1453 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1454 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1457 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1458 issue a MAIL command.
1460 PH/02 In an ACL statement such as
1462 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1464 if a client was not listed at all, or was listed with a value other than
1465 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1466 the condition was not true (as it should be), so access was not denied.
1467 The bug was that the ! inversion was incorrectly passed on to the second
1468 item. This has been fixed.
1470 PH/03 Added additional dnslists conditions == and =& which are different from
1471 = and & when the dns lookup returns more than one IP address.
1473 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1474 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1476 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1477 FSYNC, which compiles an option called disable_fsync that allows for
1478 bypassing fsync(). The documentation is heavily laced with warnings.
1480 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1482 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1483 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1484 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1485 including adding "make clean"; (3) Added -fPIC when compiling the test
1486 dynamically loaded module, to get rid of a warning.
1488 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1489 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1490 = 0s. The bug is otherwise harmless.
1492 PH/07 There was a bug in the dovecot authenticator such that the value of
1493 $auth1 could be overwritten, and so not correctly preserved, after a
1494 successful authentication. This usually meant that the value preserved by
1495 the server_setid option was incorrect.
1497 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1499 PH/09 Installed PCRE release 7.0.
1501 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1502 run for batched SMTP input. It is now run at the start of every message
1503 in the batch. While fixing this I discovered that the process information
1504 (output by running exiwhat) was not always getting set for -bs and -bS
1505 input. This is fixed, and it now also says "batched" for BSMTP.
1507 PH/11 Added control=no_pipelining.
1509 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1510 patch, slightly modified), and move the expansion of helo_data till after
1511 the connection is made in the smtp transport (so it can use these
1514 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1516 PH/14 Added log_selector = +pid.
1518 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1520 PH/16 Add ${if forany and ${if forall.
1522 PH/17 Added dsn_from option to vary the From: line in DSNs.
1524 PH/18 Flush SMTP output before performing a callout, unless control =
1525 no_callout_flush is set.
1527 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1528 was true (the default) a successful delivery failed to delete the retry
1529 item, thus causing premature timeout of the address. The bug is now
1532 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1534 PH/21 Long custom messages for fakedefer and fakereject are now split up
1535 into multiline reponses in the same way that messages for "deny" and
1536 other ACL rejections are.
1538 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1539 with slight modification.
1541 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1542 draft, changing the syntax and factoring some duplicate code".
1544 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1545 for deliveries of the second and subsequent messages over the same SMTP
1548 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1549 ${reduce, with only minor "tidies".
1551 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1553 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1554 expansion side effects.
1556 PH/27 When a message times out after an over-quota error from an Exim-imposed
1557 quota, the bounce message says "mailbox is full". This message was not
1558 being given when it was a system quota that was exceeded. It now should
1561 MH/03 Made $recipients available in local_scan(). local_scan() already has
1562 better access to the recipient list through recipients_list[], but
1563 $recipients can be useful in postmaster-provided expansion strings.
1565 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1566 in the case of a MAIL command with additional options following the
1567 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1568 were accidentally chopped off.
1570 PH/29 SMTP synchronization checks are implemented when a command is read -
1571 there is a check that no more input is waiting when there shouldn't be
1572 any. However, for some commands, a delay in an ACL can mean that it is
1573 some time before the response is written. In this time, more input might
1574 arrive, invalidly. So now there are extra checks after an ACL has run for
1575 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1576 pipelining has not been advertised.
1578 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1580 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1581 returned in the SMTP response when smtp_return_error_details was false.
1582 This has been fixed.
1584 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1585 instead of the C I/O that was originally supplied, because problems were
1586 reported on Solaris.
1588 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1589 Exim which did not show up earlier: it was assuming that a call to
1590 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1591 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1592 was a macro that became an assignment, so it seemed to work. This has
1593 changed to a proper function call with a void return, hence the compile
1594 error. Exim's code has been fixed.
1596 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1599 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1601 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1603 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1604 because it turns out to be too big or there is a timeout) there is no
1605 "Completed" line in the log. When some messages of this type were
1606 selected by exigrep, they were listed as "not completed". Others were
1607 picked up by some special patterns. I have improved the selection
1608 criteria to be more general.
1610 PH/38 The host_find_failed option in the manualroute router can now be set
1611 to "ignore", to completely ignore a host whose IP address cannot be
1612 found. If all hosts are ignored, the behaviour is controlled by the new
1613 host_all_ignored option.
1615 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1616 homing or because of multiple MX records with /mx) generated more than
1617 one IP address, and the following item turned out to be the local host,
1618 all the secondary addresses of the first item were incorrectly removed
1619 from the list, along with the local host and any following hosts (which
1620 is what is supposed to happen).
1622 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1623 whoever called Exim into the -H file. In the case of the daemon it was
1624 behaving confusingly. When first started, it used values for whoever
1625 started the daemon, but after a SIGHUP it used the Exim user (because it
1626 calls itself on a restart). I have changed the code so that it now always
1629 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1630 message are rejected with the same error (e.g. no authentication or bad
1631 sender address), and a DATA command is nevertheless sent (as can happen
1632 with PIPELINING or a stupid MUA), the error message that was given to the
1633 RCPT commands is included in the rejection of the DATA command. This is
1634 intended to be helpful for MUAs that show only the final error to their
1637 PH/42 Another patch from the Sieve maintainer.
1639 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1640 Eximstats - Fixed some broken HTML links and added missing column headers
1642 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1643 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1645 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1647 PH/43 Yet another patch from the Sieve maintainer.
1649 PH/44 I found a way to check for a TCP/IP connection going away before sending
1650 the response to the final '.' that terminates a message, but only in the
1651 case where the client has not sent further data following the '.'
1652 (unfortunately, this is allowed). However, in many cases there won't be
1653 any further data because there won't be any more messages to send. A call
1654 to select() can be used: if it shows that the input is "ready", there is
1655 either input waiting, or the socket has been closed. An attempt to read
1656 the next input character can distinguish the two cases. Previously, Exim
1657 would have sent an OK response which the client would never have see.
1658 This could lead to message repetition. This fix should cure that, at
1659 least in a lot of common cases.
1661 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1662 advertised in response to EHLO.
1668 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1669 fixed by 4.65/MH/01 (is this a record?) are fixed:
1671 (i) An empty string was always treated as zero by the numeric comparison
1672 operators. This behaviour has been restored.
1674 (ii) It is documented that the numeric comparison operators always treat
1675 their arguments as decimal numbers. This was broken in that numbers
1676 starting with 0 were being interpreted as octal.
1678 While fixing these problems I realized that there was another issue that
1679 hadn't been noticed. Values of message_size_limit (both the global option
1680 and the transport option) were treated as octal if they started with 0.
1681 The documentation was vague. These values are now always treated as
1682 decimal, and I will make that clear in the documentation.
1688 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1689 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1692 MH/01 Don't check that the operands of numeric comparison operators are
1693 integers when their expansion is in "skipping" mode (fixes bug
1694 introduced by 4.64-PH/07).
1696 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1697 child addresses, Exim now panics and dies. Previously, because the count
1698 is held in a short int, deliveries were likely to be lost. As such a
1699 large number of recipients for a single message is ridiculous
1700 (performance will be very, very poor), I have chosen to impose a limit
1701 rather than extend the field.
1707 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1708 leftover -K file (the existence of which was triggered by #402).
1709 While we were at it, introduced process PID as part of the -K
1710 filename. This should rule out race conditions when creating
1713 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1714 processing considerably. Previous code took too long for large mails,
1715 triggering a timeout which in turn triggers #401.
1717 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1718 in the DK code in transports.c. sendfile() is not really portable,
1719 hence the _LINUX specificness.
1721 TF/01 In the add_headers option to the mail command in an Exim filter,
1722 there was a bug that Exim would claim a syntax error in any
1723 header after the first one which had an odd number of characters
1726 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1727 callout verification, Exim cached a "reject" for the entire domain. This
1728 is correct for most verifications, but it is not correct for a recipient
1729 verification with use_sender or use_postmaster set, because in that case
1730 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1731 case of MAIL FROM:<> rejection from other early rejections (e.g.
1732 rejection of HELO). When verifying a recipient using a non-null MAIL
1733 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1734 Whatever the result of the callout, the value of the domain cache is
1735 left unchanged (for any other kind of callout, getting as far as trying
1736 RCPT means that the domain itself is ok).
1738 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1741 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1742 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1743 session, the daemon ignores ECONNECT errors and logs others; it now
1744 ignores EPIPE as well.
1746 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1747 (quoted-printable decoding).
1749 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1750 later the small subsequent patch to fix an introduced bug.
1752 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1754 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1756 PH/08 An error is now given if message_size_limit is specified negative.
1758 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1759 to be given (somewhat) arbitrary names.
1761 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1764 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1765 miscellaneous code fixes
1767 PH/10 Added the log_reject_target ACL modifier to specify where to log
1770 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1771 hostname. This is wrong, because it relates to the incoming message (and
1772 probably the interface on which it is arriving) and not to the outgoing
1773 callout (which could be using a different interface). This has been
1774 changed to use the value of the helo_data option from the smtp transport
1775 instead - this is what is used when a message is actually being sent. If
1776 there is no remote transport (possible with a router that sets up host
1777 addresses), $smtp_active_hostname is used.
1779 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1780 tweaks were necessary in order to get it to work (see also 21 below):
1781 (a) The code assumed that strncpy() returns a negative number on buffer
1782 overflow, which isn't the case. Replaced with Exim's string_format()
1784 (b) There were several signed/unsigned issues. I just did the minimum
1785 hacking in of casts. There is scope for a larger refactoring.
1786 (c) The code used strcasecmp() which is not a standard C function.
1787 Replaced with Exim's strcmpic() function.
1788 (d) The code set only $1; it now sets $auth1 as well.
1789 (e) A simple test gave the error "authentication client didn't specify
1790 service in request". It would seem that Dovecot has changed its
1791 interface. Fortunately there's a specification; I followed it and
1792 changed what the client sends and it appears to be working now.
1794 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1797 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1798 address A is aliased to B and C, where B exists and C does not. Without
1799 -v the output is "A verified" because verification stops after a
1800 successful redirection if more than one address is generated. However,
1801 with -v the child addresses are also verified. Exim was outputting "A
1802 failed to verify" and then showing the successful verification for C,
1803 with its parentage. It now outputs "B failed to verify", showing B's
1804 parentage before showing the successful verification of C.
1806 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1807 look up a TXT record in a specific list after matching in a combined
1810 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1811 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1812 they consult the DNS. I had assumed they would set it the way they
1813 wanted; and indeed my experiments on Linux seem to show that in some
1814 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1815 To be on the safe side, however, I have now made the interface to
1816 host_find_byname() similar to host_find_bydns(), with an argument
1817 containing the DNS resolver options. The host_find_byname() function now
1818 sets these options at its start, just as host_find_bydns() does. The smtp
1819 transport options dns_qualify_single and dns_search_parents are passed to
1820 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1821 of host_find_byname() use the default settings of RES_DEFNAMES
1822 (qualify_single) but not RES_DNSRCH (search_parents).
1824 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1825 spool_read_header() do less string testing, by means of a preliminary
1826 switch on the second character of optional "-foo" lines. (This is
1827 overdue, caused by the large number of possibilities that now exist.
1828 Originally there were few.) While I was there, I also converted the
1829 str(n)cmp tests so they don't re-test the leading "-" and the first
1830 character, in the hope this might squeeze out yet more improvement.
1832 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1833 flag allowing group syntax was set by the header_syntax check but not
1834 turned off, possible causing trouble later; (2) The flag was not being
1835 set at all for the header_verify test, causing "group"-style headers to
1836 be rejected. I have now set it in this case, and also caused header_
1837 verify to ignore an empty address taken from a group. While doing this, I
1838 came across some other cases where the code for allowing group syntax
1839 while scanning a header line wasn't quite right (mostly, not resetting
1840 the flag correctly in the right place). These bugs could have caused
1841 trouble for malformed header lines. I hope it is now all correct.
1843 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1844 with the "reply" argument non-NULL. The code, however (which originally
1845 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1846 but it didn't always do it. This confused somebody who was copying the
1847 code for some other use. I have removed all the tests.
1849 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1850 feature that was used to support insecure browsers during the U.S. crypto
1851 embargo. It requires special client support, and Exim is probably the
1852 only MTA that supported it -- and would never use it because real RSA is
1853 always available. This code has been removed, because it had the bad
1854 effect of slowing Exim down by computing (never used) parameters for the
1855 RSA_EXPORT functionality.
1857 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1858 authenticator to fail if there's a tab character in the incoming data
1859 (there should never be unless someone is messing about, as it's supposed
1860 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1861 if the connection is using TLS or if the remote IP is the same as the
1862 local IP, and the "valid-client-cert option" if a client certificate has
1865 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1866 authenticators. This can be used for authorization after authentication
1867 succeeds. (In the case of plaintext, it servers for both authentication
1870 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1871 if any retry times were supplied.
1873 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1874 connection, where there is no client IP address to check. In this
1875 situation, the verify now always succeeds.
1877 PH/25 Applied John Jetmore's -Mset patch.
1879 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1881 PH/27 In a string expansion for a processed (not raw) header when multiple
1882 headers of the same name were present, leading whitespace was being
1883 removed from all of them, but trailing whitespace was being removed only
1884 from the last one. Now trailing whitespace is removed from each header
1885 before concatenation. Completely empty headers in a concatenation (as
1886 before) are ignored.
1888 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1889 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1891 PH/29 [Removed. This was a change that I later backed out, and forgot to
1892 correct the ChangeLog entry (that I had efficiently created) before
1893 committing the later change.]
1895 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1896 address errors (4xx response to RCPT) over the same connection as other
1897 messages routed to the same hosts. Such deliveries are always "forced",
1898 so retry times are not inspected. This resulted in far too many retries
1899 for the affected addresses. The effect occurred only when there were more
1900 hosts than the hosts_max_try setting in the smtp transport when it had
1901 the 4xx errors. Those hosts that it had tried were not added to the list
1902 of hosts for which the message was waiting, so if all were tried, there
1903 was no problem. Two fixes have been applied:
1905 (i) If there are any address or message errors in an SMTP delivery, none
1906 of the hosts (tried or untried) are now added to the list of hosts
1907 for which the message is waiting, so the message should not be a
1908 candidate for sending over the same connection that was used for a
1909 successful delivery of some other message. This seems entirely
1910 reasonable: after all the message is NOT "waiting for some host".
1911 This is so "obvious" that I'm not sure why it wasn't done
1912 previously. Hope I haven't missed anything, but it can't do any
1913 harm, as the worst effect is to miss an optimization.
1915 (ii) If, despite (i), such a delivery is accidentally attempted, the
1916 routing retry time is respected, so at least it doesn't keep
1917 hammering the server.
1919 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1920 in ${readsocket because some servers need this prod.
1922 PH/32 Added some extra debug output when updating a wait-xxx database.
1924 PH/33 The hint "could be header name not terminated by colon", which has been
1925 given for certain expansion errors for a long time, was not being given
1926 for the ${if def:h_colon_omitted{... case.
1928 PH/34 The spec says: "With one important exception, whenever a domain list is
1929 being scanned, $domain contains the subject domain." There was at least
1930 one case where this was not true.
1932 PH/35 The error "getsockname() failed: connection reset by peer" was being
1933 written to the panic log as well as the main log, but it isn't really
1934 panic-worthy as it just means the connection died rather early on. I have
1935 removed the panic log writing for the ECONNRESET error when getsockname()
1938 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1939 runs only) independently of the message's sender address. This meant
1940 that, if the 4xx error was in fact related to the sender, a different
1941 message to the same recipient with a different sender could confuse
1942 things. In particualar, this can happen when sending to a greylisting
1943 server, but other circumstances could also provoke similar problems.
1944 I have changed the default so that the retry time for these errors is now
1945 based a combination of the sender and recipient addresses. This change
1946 can be overridden by setting address_retry_include_sender=false in the
1949 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1950 remote server are returned as part of bounce messages. This was not
1951 happening for LMTP over a pipe (the lmtp transport), but now it is the
1952 same for both kinds of LMTP.
1954 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1955 in header lines that were in fact CNAMEs. This is no longer the case.
1957 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1958 and queue runs started by the daemon processed all messages. This has
1959 been fixed so that -R and -S can now usefully be given with -q<time>.
1961 PH/40 Import PCRE release 6.7 (fixes some bugs).
1963 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1965 PH/42 Give an error if -q is specified more than once.
1967 PH/43 Renamed the variables $interface_address and $interface_port as
1968 $received_ip_address and $received_port, to make it clear that these
1969 values apply to message reception, and not to the outgoing interface when
1970 a message is delivered. (The old names remain recognized, of course.)
1972 PH/44 There was no timeout on the connect() call when using a Unix domain
1973 socket in the ${readsocket expansion. There now is.
1975 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1976 be meaningful with "accept".
1978 SC/01 Eximstats V1.43
1979 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1981 SC/02 Eximstats V1.44
1982 Use a glob alias rather than an array ref in the generated
1983 parser. This improves both readability and performance.
1985 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1986 Collect SpamAssassin and rejection statistics.
1987 Don't display local sender or destination tables unless
1988 there is data to show.
1989 Added average volumes into the top table text output.
1991 SC/04 Eximstats V1.46
1992 Collect data on the number of addresses (recipients)
1993 as well as the number of messages.
1995 SC/05 Eximstats V1.47
1996 Added 'Message too big' to the list of mail rejection
1997 reasons (thanks to Marco Gaiarin).
1999 SC/06 Eximstats V1.48
2000 Mainlog lines which have GMT offsets and are too short to
2001 have a flag are now skipped.
2003 SC/07 Eximstats V1.49 (Alain Williams)
2004 Added the -emptyok flag.
2006 SC/08 Eximstats V1.50
2007 Fixes for obtaining the IP address from reject messages.
2009 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2010 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2011 whitesspace changes from 4.64-PH/27
2013 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2016 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2017 are found, allow negative numbers in numeric criteria)
2019 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2021 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2022 to match changes made in 4.64-PH/43
2024 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2026 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2027 "domain" is actually an IP address, and force a failure. This locks out
2028 those revolvers/nameservers that support "A-for-A" lookups, in
2029 contravention of the specifications.
2031 PH/48 When a host name was looked up from an IP address, and the subsequent
2032 forward lookup of the name timed out, the host name was left in
2033 $sender_host_name, contrary to the specification.
2035 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2036 restricted to single-key lookups, Exim was not diagnosing an error if
2037 * or *@ was used with a query-style lookup.
2039 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2041 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2042 long ago, but noone interested enough thought of it. Let's just say that
2043 the "1.1" means that there are some new functions that weren't there at
2044 some point in the past.
2046 PH/51 Error processing for expansion failure of helo_data from an smtp
2047 transport during callout processing was broken.
2049 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2050 tested/used via the -bh/-bhc/-bs options.
2052 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2053 bug, fixed in subsequent PCRE releases).
2055 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2056 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2058 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2064 SC/01 Use a glob alias rather than an array ref in eximstats generated
2065 parser. This improves both readability and performance.
2067 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2068 Don't display local sender or destination tables in eximstats unless
2069 there is data to show.
2070 Added average volumes into the eximstats top table text output.
2072 SC/03 Collect data on the number of addresses (recipients) as well
2073 as the number of messages in eximstats.
2075 TF/01 Correct an error in the documentation for the redirect router. Exim
2076 does (usually) call initgroups() when daemonizing.
2078 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2079 with consistent privilege compared to when running as a daemon.
2081 TF/03 Note in the spec that $authenticated_id is not set for local
2082 submissions from trusted users.
2084 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2085 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2087 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2088 by adding some example configuration directives to the default
2089 configuration file. A little bit of work is required to uncomment the
2090 directives and define how usernames and passwords are checked, but
2091 there is now a framework to start from.
2093 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2094 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2095 without this. I don't know how relevant this is to other LDAP libraries.
2097 PH/02 Add the verb name to the "unknown ACL verb" error.
2099 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2101 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2103 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2104 directory not expanded when it should be if an expanded home directory
2105 was set for the address (which is overridden by the transport).
2107 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2110 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2111 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2112 because it is too late at that time, and has no effect.
2114 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2115 security issue with \' (bugzilla #107). I could not use the
2116 PQescapeStringConn() function, because it needs a PGconn value as one of
2119 PH/08 When testing addresses using -bt, indicate those final addresses that
2120 are duplicates that would not cause an additional delivery. At least one
2121 person was confused, thinking that -bt output corresponded to deliveries.
2122 (Suppressing duplicates isn't a good idea as you lose the information
2123 about possibly different redirections that led to the duplicates.)
2125 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2126 systems where poll() doesn't work, in particular OS X.
2128 PH/10 Added more information to debugging output for retry time not reached.
2130 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2131 operations in malware.c.
2133 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2136 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2137 syslog_duplication set false, log lines that would normally be written
2138 both the the main log and to the reject log were not written to syslog at
2141 PH/14 In the default configuration, change the use of "message" in ACL warn
2142 statements to "add_header".
2144 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2145 not followed by a command (e.g. "seen endif").
2147 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2148 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2151 PH/17 Added extra conditions to the default value of delay_warning_condition
2155 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2156 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2157 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2160 The Auto-Submitted: and various List- headers are standardised, whereas I
2161 don't think Precedence: ever was.
2163 PH/18 Refactored debugging code in route_finduser() to show more information,
2164 in particular, the error code if getpwnam() issues one.
2166 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2167 This is apparently needed in addition to the PH/07 change above to avoid
2168 any possible encoding problems.
2170 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2171 but not after initializing Perl.
2173 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2174 output them only if debugging. By default they are written stderr,
2175 apparently, which is not desirable.
2177 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2180 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2183 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2185 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2186 authenticated or an ident call has been made. Suppress the default
2187 values for $authenticated_id and $authenticated_sender (but permit -oMai
2188 and -oMas) when testing with -bh.
2190 PH/24 Re-jigged the order of the tests in the default configuration so that the
2191 tests for valid domains and recipients precede the DNS black list and CSA
2192 tests, on the grounds that those ones are more expensive.
2194 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2195 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2196 This bug exists in every version of Exim that I still have, right back to
2199 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2200 However, an attempt to turn on case-sensitivity in a regex key by
2201 including (?-i) didn't work because the subject string was already
2202 lowercased, and the effects were non-intuitive. It turns out that a
2203 one-line patch can be used to allow (?-i) to work as expected.
2209 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2210 other effects) broke the use of negated acl sub-conditions.
2212 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2215 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2216 "Deny" causes Exim to reject the incoming connection with a 554 error.
2217 Unfortunately, if there is a major crisis, such as a disk failure,
2218 tcp-wrappers gives "deny", whereas what one would like would be some
2219 kind of temporary error. A kludge has been added to help with this.
2220 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2221 554 error is used if errno is still zero or contains ENOENT (which occurs
2222 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2225 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2227 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2228 errors. Otherwise a message that provokes a temporary error (when other
2229 messages do not) can cause a whole host to time out.
2231 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2232 addresses were routed directly to files or pipes from a redirect router.
2233 File deliveries just didn't batch; pipe deliveries might have suffered
2236 PH/06 A failure to get a lock for a hints database would erroneously always say
2237 "Failed to get write lock", even when it was really a read lock.
2239 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2240 of 0600. This has been changed to use the value of the lockfile_mode
2241 option (which defaults to 0600).
2243 PH/08 Applied small patch from the Sieve maintainer.
2245 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2246 folder from quota calculations, a direct delivery into this folder messed
2247 up the contents of the maildirsize file. This was because the regex was
2248 used only to exclude .Trash (or whatever) when the size of the mailbox
2249 was calculated. There was no check that a delivery was happening into an
2250 excluded directory. This bug has been fixed by ignoring all quota
2251 processing for deliveries into excluded directories.
2253 PH/10 Added the maildirfolder_create_regex option to appendfile.
2259 PH/01 The code for finding all the local interface addresses on a FreeBSD
2260 system running IPv6 was broken. This may well have applied to all BSD
2261 systems, as well as to others that have similar system calls. The broken
2262 code found IPv4 interfaces correctly, but gave incorrect values for the
2263 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2264 that it would not match correctly against @[] and not recognize the IPv6
2267 PH/02 The ipliteral router was not recognizing addresses of the form user@
2268 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2270 PH/03 Added disable_ipv6.
2272 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2273 decoded form, because it is most often used to construct To: headers
2274 lines in autoreplies, and the decoded form may well be syntactically
2275 invalid. However, $reply_address has leading white space removed, and all
2276 newlines turned into spaces so that the autoreply transport does not
2279 PH/05 If group was specified without a user on a router, and no group or user
2280 was specified on a transport, the group from the router was ignored.
2282 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2283 for visible compile-time settings that can be used to change these
2284 numbers, for those that want even more. Backwards compatibility with old
2285 spool files has been maintained. However, going back to a previous Exim
2286 release will lost any variables that are in spool files.
2288 PH/07 Two small changes when running in the test harness: increase delay when
2289 passing a TCP/IP connection to a new process, in case the original
2290 process has to generate a bounce, and remove special handling of
2291 127.0.0.2 (sic), which is no longer necessary.
2293 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2294 be the same on different OS.
2296 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2299 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2300 whether --show-vars was specified or not
2302 JJ/02 exipick: Added support for new ACL variable spool format introduced
2305 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2306 syntactically invalid From: or Reply-to: line, and a filter used this to
2307 generate an autoreply, and therefore failed to obtain an address for the
2308 autoreply, Exim could try to deliver to a non-existent relative file
2309 name, causing unrelated and misleading errors. What now happens is that
2310 it logs this as a hard delivery error, but does not attempt to create a
2313 PH/11 The exinext utility has a -C option for testing purposes, but although
2314 the given file was scanned by exinext itself; it wasn't being passed on
2315 when Exim was called.
2317 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2318 an end-of-file indication when reading a command response.
2320 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2321 compiled. In many other places in Exim, IPv6 addresses are always
2322 recognized, so I have changed this. It also means that IPv4 domain
2323 literals of the form [IPV4:n.n.n.n] are now always recognized.
2325 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2326 used if the router is not running as root, for example, when verifying at
2327 ACL time, or when using -bh. The debugging output from this situation was
2328 non-existent - all you got was a failure to exec. I have made two
2331 (a) Failures to set uid/gid, the current directory, or a process leader
2332 in a subprocess such as that created by queryprogram now generate
2333 suitable debugging ouput when -d is set.
2335 (b) The queryprogram router detects when it is not running as root,
2336 outputs suitable debugging information if -d is set, and then runs
2337 the subprocess without attempting to change uid/gid.
2339 PH/15 Minor change to Makefile for building test_host (undocumented testing
2342 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2343 additional section of a DNS packet that returns MX or SRV records.
2344 Instead, it always explicitly searches for A/AAAA records. This avoids
2345 major problems that occur when a DNS server includes only records of one
2346 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2347 fixed another bug: if SRV records were looked up and the corresponding
2348 address records were *not* found in the additional section, the port
2349 values from the SRV records were lost.
2351 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2352 using the correct key (the original address) when searching the retry
2353 rules in order to find which one to use for generating the retry hint.
2355 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2356 adding the default one. Similarly, if it contains a Reply-To: header, the
2357 errors_reply_to option, if set, is not used.
2359 PH/19 When calculating a retry time, Exim used to measure the "time since
2360 failure" by looking at the "first failed" field in the retry record. Now
2361 it does not use this if it is later than than the arrival time of the
2362 message. Instead it uses the arrival time. This makes for better
2363 behaviour in cases where some deliveries succeed, thus re-setting the
2364 "first failed" field. An example is a quota failure for a huge message
2365 when small messages continue to be delivered. Without this change, the
2366 "time since failure" will always be short, possible causing more frequent
2367 delivery attempts for the huge message than are intended.
2368 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2370 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2371 $1, $2, $3) because the numerical variables can be reset during some
2372 expansion items (e.g. "match"), thereby losing the authentication data.
2374 PH/21 Make -bV show the size of off_t variables so that the test suite can
2375 decide whether to run tests for quotas > 2G.
2377 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2378 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2379 filecount value is greater than 2G or if a quota value is greater than 2G
2380 on a system where the size of off_t is not greater than 4, a panic error
2383 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2384 never match. The debug and -bh output now contains an explicit error
2385 message indicating a malformed IPv4 address or mask.
2387 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2388 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2389 PH/23 above applies.
2391 PH/25 Do not write to syslog when running in the test harness. The only
2392 occasion when this arises is a failure to open the main or panic logs
2393 (for which there is an explicit test).
2395 PH/26 Added the /no_tell option to "control=freeze".
2397 PH/27 If a host name lookup failed very early in a connection, for example, if
2398 the IP address matched host_lookup and the reverse lookup yielded a name
2399 that did not have a forward lookup, an error message of the form "no IP
2400 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2401 could be logged. Now it outputs the IP address instead of "NULL".
2403 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2404 allows the sender and the authenticated sender to be set when
2405 submitting a message from within Exim. Since child_open_exim() is
2406 documented for local_scan(), the new function should be too.
2408 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2409 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2410 results in an empty string is now treated as unset.
2412 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2414 PH/31 Added sender_verify_fail logging option.
2416 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2417 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2418 tidied the source and removed it altogether.
2420 PH/33 When a queue run was abandoned because the load average was too high, a
2421 log line was always written; now it is written only if the queue_run log
2422 selector is set. In addition, the log line for abandonment now contains
2423 information about the queue run such as the pid. This is always present
2424 in "start" and "stop" lines but was omitted from the "abandon" line.
2426 PH/34 Omit spaces between a header name and the colon in the error message that
2427 is given when verify = headers_syntax fails (if there are lots of them,
2428 the message gets confusing).
2430 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2431 names, as there are now some PTR records that contain slashes. This check
2432 is only to protect against broken name servers that fall over on strange
2433 characters, so the fact that it applies to all lookups doesn't matter.
2435 PH/36 Now that the new test suite is complete, we can remove some of the
2436 special code in Exim that was needed for the old test suite. For example,
2437 sorting DNS records because real resolvers return them in an arbitrary
2438 order. The new test suite's fake resolver always returns records in the
2441 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2442 bounces) except when queue_only is set, to avoid logging races between
2443 the different processes.
2445 PH/38 Panic-die if .include specifies a non-absolute path.
2447 PH/39 A tweak to the "H" retry rule from its user.
2449 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2450 a label. They prevented compilation on older perls.
2452 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2453 a warning to be raised on newish perls.
2455 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2456 on queue. Changes to match documented behaviour of showing count of
2457 messages matching specified criteria.
2459 PH/40 Changed the default ident timeout from 30s to 5s.
2461 PH/41 Added support for the use of login_cap features, on those BSD systems
2462 that have them, for controlling the resources used by pipe deliveries.
2464 PH/42 The content-scanning code uses fopen() to create files in which to put
2465 message data. Previously it was not paying any attention to the mode of
2466 the files. Exim runs with umask(0) because the rest of the code creates
2467 files with open(), and sets the required mode explicitly. Thus, these
2468 files were ending up world-writeable. This was not a big issue, because,
2469 being within the spool directory, they were not world-accessible. I have
2470 created a function called modefopen, which takes an additional mode
2471 argument. It sets umask(777), creates the file, chmods it to the required
2472 mode, then resets the umask. All the relevant calls to fopen() in the
2473 content scanning code have been changed to use this function.
2475 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2476 to 24 hours. This avoids potential overflow problems when processing G
2477 and H retry rules. I suspect nobody ever tinkers with this value.
2479 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2481 PH/45 When the plaintext authenticator is running as a client, the server's
2482 challenges are checked to ensure they are valid base64 strings. By
2483 default, the authentication attempt is cancelled if an invalid string is
2484 received. Setting client_ignore_invalid_base64 true ignores these errors.
2485 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2486 they are received. Thus, the responses can be made to depend on the
2487 challenges. If an invalid string is ignored, an empty string is placed in
2490 PH/46 Messages that are created by the autoreply transport now contains a
2491 References: header, in accordance with RFCs 2822 and 3834.
2493 PH/47 Added authenticated_sender_force to the smtp transport.
2495 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2497 PH/49 Installed latest patch from the Sieve maintainer.
2499 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2500 was also set, the appendfile transport was unnecessarily scanning a
2501 directory of message files (e.g. for maildir delivery) to find the count
2502 of files (along with the size), even though it did not need this
2503 information. It now does the scan only if it needs to find either the
2504 size of the count of files.
2506 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2508 PH/52 Two bugs concerned with error handling when the smtp transport is
2511 (i) Exim was not creating retry information for temporary errors given
2512 for individual recipients after the DATA command when the smtp transport
2513 was used in LMTP mode. This meant that they could be retried too
2514 frequently, and not timed out correctly.
2516 (ii) Exim was setting the flag that allows error details to be returned
2517 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2518 recipients that were returned after the DATA command.
2520 PH/53 This is related to PH/52, but is more general: for any failing address,
2521 when detailed error information was permitted to be returned to the
2522 sender, but the error was temporary, then after the final timeout, only
2523 "retry timeout exceeded" was returned. Now it returns the full error as
2524 well as "retry timeout exceeded".
2526 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2527 do this, and (what is worse) MTAs that accept it.
2529 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2530 will now be deprecated.
2532 PH/56 New os.c-cygwin from the Cygwin maintainer.
2534 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2535 formats (previously only available in exim formats via -bpr, -bpru,
2536 and -bpra. Now also available in native and exiqgrep formats)
2538 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2539 with very large, slow to parse queues
2541 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2543 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2545 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2546 responses to authentication challenges, though it was showing the
2547 challenges; (ii) I've removed the CR characters from the debug output for
2550 PH/58 Allow for the insertion of a newline as well as a space when a string
2551 is turned into more than one encoded-word during RFC 2047 encoding. The
2552 Sieve code now uses this.
2554 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2555 data_4xx, lost_connection, tls_required.
2557 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2558 message was being sent as an SMTP response.
2560 PH/61 Add -l and -k options to exicyclog.
2562 PH/62 When verifying, if an address was redirected to one new address, so that
2563 verification continued, and the new address failed or deferred after
2564 having set something in $address_data, the value of $address_data was not
2565 passed back to the ACL. This was different to the case when no
2566 redirection occurred. The value is now passed back in both cases.
2568 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2569 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2570 use login_cap.h, so on its own it isn't the distinguishing feature. The
2571 new name refers directly to the setclassresources() function.
2573 PH/65 Added configuration files for NetBSD3.
2575 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2577 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2578 is preferred over IPv4.
2580 PH/68 The bounce_return_message and bounce_return_body options were not being
2581 honoured for bounces generated during the reception of non-SMTP messages.
2582 In particular, this applied to messages rejected by the ACL. This bug has
2583 been fixed. However, if bounce_return_message is true and bounce_return_
2584 body is false, the headers that are returned for a non-SMTP message
2585 include only those that have been read before the error was detected.
2586 (In the case of an ACL rejection, they have all been read.)
2588 PH/69 The HTML version of the specification is now built in a directory called
2589 spec_html instead of spec.html, because the latter looks like a path with
2590 a MIME-type, and this confuses some software.
2592 PH/70 Catch two compiler warnings in sieve.c.
2594 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2595 function verify_get_ident() calls ip_connect() to connect a socket, but
2596 if the "connect()" function timed out, ip_connect() used to close the
2597 socket. However, verify_get_ident() also closes the socket later, and in
2598 between Exim writes to the log, which may get opened at this point. When
2599 the socket was closed in ip_connect(), the log could get the same file
2600 descriptor number as the socket. This naturally causes chaos. The fix is
2601 not to close the socket in ip_connect(); the socket should be closed by
2602 the function that creates it. There was only one place in the code where
2603 this was missing, in the iplookup router, which I don't think anybody now
2604 uses, but I've fixed it anyway.
2606 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2607 well as to direct DNS lookups. Otherwise the handling of names in host
2608 lists is inconsistent and therefore confusing.
2614 PH/01 Two changes to the default runtime configuration:
2616 (1) Move the checks for relay_from_hosts and authenticated clients from
2617 after to before the (commented out) DNS black list checks.
2619 (2) Add control=submission to the relay_from_hosts and authenticated
2620 clients checks, on the grounds that messages accepted by these
2621 statements are most likely to be submissions.
2623 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2625 (1) Generate an error if the third argument for the ${prvs expansion is
2628 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2631 (3) Reset the variables that are obtained from the first argument of
2632 ${prvscheck and used in the second argument before leaving the code,
2633 because their memory is reclaimed, so using them afterwards may do
2636 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2637 one (it's much easier than Tom thought :-).
2639 (5) Because of (4), we can now allow for the use of $prvscheck_result
2640 inside the third argument.
2642 PH/03 For some reason, the default setting of PATH when running a command from
2643 a pipe transport was just "/usr/bin". I have changed it to
2646 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2647 anything to be listed in the output from -bV.
2649 PH/05 When a filter generated an autoreply, the entire To: header line was
2650 quoted in the delivery log line, like this:
2652 => >A.N.Other <ano@some.domain> <original@ddress> ...
2654 This has been changed so that it extracts the operative address. There
2655 may be more than one such address. If so, they are comma-separated, like
2658 => >ano@some.domain,ona@other.domain <original@ddress> ...
2660 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2661 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2662 not being looked up in the rDNS to get a host name, Exim was showing the
2663 IP address twice in Received: lines, even though the IP addresses were
2664 identical. For example:
2666 Received: from [1.2.3.4] (helo=[1.2.3.4])
2668 However, if the real host name was known, it was omitting the HELO data
2669 if it matched the actual IP address. This has been tidied up so that it
2670 doesn't show the same IP address twice.
2672 PH/07 When both +timestamp and +memory debugging was on, the value given by
2673 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2674 called by the debug printing, thereby overwriting the timestamp buffer.
2675 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2677 PH/08 When the original message was included in an autoreply transport, it
2678 always said "this is a copy of the message, including all the headers",
2679 even if body_only or headers_only was set. It now gives an appropriate
2682 PH/09 Applied a patch from the Sieve maintainer which:
2684 o fixes some comments
2685 o adds the (disabled) notify extension core
2686 o adds some debug output for the result of if/elsif tests
2687 o points to the current vacation draft in the documentation
2688 and documents the missing references header update
2692 o fixes a bug in processing the envelope test (when testing
2693 multiple envelope elements, the last element determinted the
2696 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2697 Electronic Mail") by including:
2699 Auto-submitted: auto-generated
2701 in the messages that it generates (bounce messages and others, such as
2702 warnings). In the case of bounce messages for non-SMTP mesages, there was
2703 also a typo: it was using "Auto_submitted" (underscore instead of
2704 hyphen). Since every message generated by Exim is necessarily in response
2705 to another message, thes have all been changed to:
2707 Auto-Submitted: auto-replied
2709 in accordance with these statements in the RFC:
2711 The auto-replied keyword:
2713 - SHOULD be used on messages sent in direct response to another
2714 message by an automatic process,
2716 - MUST NOT be used on manually-generated messages,
2718 - MAY be used on Delivery Status Notifications (DSNs) and Message
2719 Disposition Notifications (MDNs),
2721 - MUST NOT be used on messages generated by automatic or periodic
2722 processes, except for messages which are automatic responses to
2725 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2726 to the default Received: header definition.
2728 PH/12 Added log selector acl_warn_skipped (default on).
2730 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2731 variables because (a) they are in the wrong storage pool and (b) even if
2732 they were copied, it wouldn't work properly because of the caching.
2734 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2735 checking when decoding. Apparently there are clients that generate
2736 overlong encoded strings. Why am I not surprised?
2738 PH/15 If the first argument of "${if match_address" was not empty, but did not
2739 contain an "@" character, Exim crashed. Now it writes a panic log message
2740 and treats the condition as false.
2742 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2744 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2745 extension "envelope-auth". The code is finished and in agreement with
2746 other implementations, but there is no documentation so far and in fact,
2747 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2748 not changing the active code.
2750 Print executed "if" and "elsif" statements when debugging is used. This
2751 helps a great deal to understand what a filter does.
2753 Document more things not specified clearly in RFC3028. I had all this
2754 sorted out, when out of a sudden new issues came to my mind. Oops."
2756 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2759 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2760 canonical form (as documented). However, after a host name lookup from
2761 the IP address, check_host() was doing a simple string comparison with
2762 addresses acquired from the DNS when checking that the found name did
2763 have the original IP as one of its addresses. Since any found IPv6
2764 addresses are likely to be in abbreviated form, the comparison could
2765 fail. Luckily, there already exists a function for doing the comparison
2766 by converting both addresses to binary, so now that is used instead of
2767 the text comparison.
2769 PH/20 There was another similar case to PH/19, when a complete host name was
2770 given in a host list; looking up its IP address could give an abbreviated
2771 form, whereas the current host's name might or might not be abbreviated.
2772 The same fix has been applied.
2778 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2779 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2782 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2783 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2785 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2787 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2788 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2789 Also, for "Auto-Submitted", treat anything other than "no" as
2790 non-personal, in accordance with RFC 3834. (Previously it treated
2791 anything starting "auto-" as non-personal.)
2793 TF/01 The control=submission/name=... option had a problem with syntax
2794 errors if the name included a slash character. The /name= option
2795 now slurps the rest of the string, so it can include any characters
2796 but it must come last in the list of options (after /sender_retain
2799 PH/05 Some modifications to the interface to the fake nameserver for the new
2807 TK/01 Added the "success_on_redirect" address verification option. See
2808 NewStuff for rationale and an example.
2810 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2812 PH/02 Patch to exigrep to allow it to work on syslog lines.
2814 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2815 fread() to skip over the body file's header line, because in Cygwin the
2816 header line is locked and is inaccessible.
2818 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2819 co-exist for some time) to make it clear that it is the Exim ID that is
2820 referenced, not the Message-ID: header line.
2822 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2823 string_format() function, because snprintf() does not exist on all
2826 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2827 ${stat: expansion item.
2829 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2830 protocol synchronization error", to keep the pedants happy.
2832 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2833 well as for IRIX systems, when gcc is being used. See the host.c source
2836 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2838 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2841 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2842 transport and to the smtp transport in LMTP mode.
2844 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2846 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2847 run a filter in a subprocess. This could lead to confusion in subsequent
2848 lookups in the parent process. There should also be a search_tidyup() at
2849 the end of the subprocess.
2851 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2852 only if the host matched helo_try_verify_hosts, which caused the
2853 verification to occur when the EHLO/HELO command was issued. The ACL just
2854 tested the remembered result. Now, if a previous verification attempt has
2855 not happened, "verify = helo" does it there and then.
2857 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2859 TK/03 Fix log output including CR from clamd.
2861 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2862 exist provoked a memory error which could cause a segfault.
2864 PH/15 Installed PCRE 6.2
2866 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2868 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2869 of the problem. Specifically, suggested +O2 rather than +O1 for the
2872 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2874 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2875 record, the port was not correctly logged when the outgoing_port log
2876 selector was set (it logged the transort's default port).
2878 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2879 fallback_hosts, and "hosts" in the smtp transport.
2881 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2882 host errors such as "Connection refused".
2884 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2885 authentication with radiusclient 0.4.9:
2887 - Error returned from rc_read_config was caught wrongly
2888 - Username/password not passed on to radius server due to wrong length.
2890 The presumption is that some radiusclient API changes for 4.51/PH/17
2891 were not taken care of correctly. The code is still untested by me (my
2892 Linux distribution still has 0.3.2 of radiusclient), but it was
2893 contributed by a Radius user.
2895 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2896 expanding the "port" option of the smtp transport.
2898 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2899 while reading a MIME header. Thanks to Tom Hughes for a patch.
2901 PH/24 Include config.h inside local_scan.h so that configuration settings are
2904 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2905 that in an ACL for RCPT (for example), you can examine exactly what was
2908 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2909 commands, but it was not correctly comparing the address with the actual
2910 client host address. Thus, it would show the EHLO address in Received:
2911 header lines when this was not necessary.
2913 PH/27 Added the % operator to ${eval:}.
2915 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2916 it should be ignoring failures (because with -C, for example, it has lost
2917 privilege). It wasn't ignoring creation failures other than "already
2920 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2923 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2924 because an input line was too long, either on its own, or by virtue of
2925 too many continuations, the temporary file was not being removed, and the
2926 return code was incorrect.
2928 PH/31 Missing "BOOL" in function definition in filtertest.c.
2930 PH/32 Applied Sieve patches from the maintainer.
2932 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2934 PH/33 Added "verify = not_blind".
2936 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2937 Local/Makefile (with some defaults set). These are used in built scripts
2938 such as exicyclog, but they have never been used in the exim_install
2939 script (though there are many overriding facilities there). I have
2940 arranged that the exim_install script now takes note of these two
2943 PH/35 Installed configuration files for Dragonfly.
2945 PH/36 When a locally submitted message by a trusted user did not contain a
2946 From: header, and the sender address was obtained from -f or from an SMTP
2947 MAIL command, and the trusted user did not use -F to supply a sender
2948 name, $originator_name was incorrectly used when constructing a From:
2949 header. Furthermore, $originator_name was used for submission mode
2950 messages from external hosts without From: headers in a similar way,
2951 which is clearly wrong.
2953 PH/37 Added control=suppress_local_fixups.
2955 PH/38 When log_selector = +received_sender was set, and the addition of the
2956 sender made the log line's construction buffer exactly full, or one byte
2957 less than full, an overflow happened when the terminating "\n" was
2960 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2961 when the result of a list match is failure because a DNS lookup failed.
2963 PH/40 RM_COMMAND is now used in the building process.
2965 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2966 the "build-* directories that it finds.
2968 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2969 address in a domain literal was a prefix of an interface address.
2971 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2972 when verifying a sender address, unless rewrite_headers is false.
2974 PH/44 Wrote a long comment about why errors_to addresses are verified as
2975 recipients, not senders.
2977 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2978 the ratelimit ACL was added.
2980 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2982 PH/46 Added extra information about PostgreSQL errors to the error string.
2984 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2985 avoiding the need to install special zones in a real server. This is
2986 backwards compatible; if it can't find the fake resolver, it drops back.
2987 Thus, both old and new test suites can be run.
2989 TF/02 Added util/ratelimit.pl
2991 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2992 clock is set back in time.
2994 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2995 Candler <B.Candler@pobox.com>.
2997 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2998 OK for addresses that are the result of redirections.
3000 PH/48 A number of further additions for the benefit of the new test suite,
3001 including a fake gethostbyname() that interfaces to the fake DNS resolver
3004 TF/06 The fix for widen_domains has also been applied to qualify_single and
3005 search_parents which are the other dnslookup options that can cause
3008 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3011 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3013 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3014 service" if it encountered an MX record with an empty target hostname.
3015 The message is now "an MX or SRV record indicated no SMTP service".
3017 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3018 if the DNS of the sending site is misconfigured. This is quite a
3019 common situation. This change restores the behaviour of treating a
3020 helo verification defer as a failure.
3022 PH/51 If self=fail was set on a router, the bounce message did not include the
3023 actual error message.
3029 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3031 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3032 command itself ended in error, the underlying message about the transport
3033 filter timeout was being overwritten with the pipe command error. Now the
3034 underlying error message should be appended to the second error message.
3036 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3038 PH/02 Reduce the amount of output that "make" produces by default. Full output
3039 can still be requested.
3041 PH/03 The warning log line about a condition test deferring for a "warn" verb
3042 was being output only once per connection, rather than after each
3043 occurrence (because it was using the same function as for successful
3044 "warn" verbs). This seems wrong, so I have changed it.
3046 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3047 it should not have, which might have caused a crash in the right
3048 circumstances, but probably never did.
3050 PH/04 Installed a modified version of Tony Finch's patch to make submission
3051 mode fix the return path as well as the Sender: header line, and to
3052 add a /name= option so that you can make the user's friendly name appear
3055 TF/03 Added the control = fakedefer ACL modifier.
3057 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3058 Mark Lowes for thorough testing.
3060 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3062 TK/03 Merged latest SRS patch from Miles Wilton.
3064 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3065 with the definition in sysexits.h (which is #included earlier).
3066 Fortunately, Exim does not actually use EX_OK. The code used to try to
3067 preserve the sysexits.h value, by assumimg that macro definitions were
3068 scanned for macro replacements. I have been disabused of this notion,
3069 so now the code just undefines EX_OK before #including unistd.h.
3071 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3072 in the smtp transport. When a block could not be written in a single
3073 write() function, the timeout was being re-applied to each part-write.
3074 This seems wrong - if the receiver was accepting one byte at a time it
3075 would take for ever. The timeout is now adjusted when this happens. It
3076 doesn't have to be particularly precise.
3078 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3079 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3081 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3082 without a domain if the check to <postmaster@domain> fails.
3084 SC/01 Eximstats: added -xls and the ability to specify output files
3085 (patch written by Frank Heydlauf).
3087 SC/02 Eximstats: use FileHandles for outputing results.
3089 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3091 SC/04 Eximstats: fixed display of large numbers with -nvr option
3093 SC/05 Eximstats: fixed merging of reports with empty tables.
3095 SC/06 Eximstats: added the -include_original_destination flag
3097 SC/07 Eximstats: removed tabs and trailing whitespace.
3099 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3101 TK/06 MBOX spool code: Add real "From " MBOX separator line
3102 so the .eml file is really in mbox format (even though
3103 most programs do not really care). Patch from Alex Miller.
3105 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3106 The latter is generated from $received_to and is only set if the
3107 message has one envelope recipient. SA can use these headers,
3108 obviously out-of-the-box. Patch from Alex Miller.
3110 PH/08 The ${def test on a variable was returning false if the variable's
3111 value was "0", contrary to what the specification has always said!
3112 The result should be true unless the variable is empty.
3114 PH/09 The syntax error of a character other than { following "${if
3115 def:variable_name" (after optional whitespace) was not being diagnosed.
3116 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3117 accidental colon was present, for example, could give incorrect results.
3119 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3120 result is used (not including appendfile, where other changes are about
3123 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3124 This involved changing a lot of size variables from int to off_t. It
3125 should work with maildirs and everything.
3127 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3128 spamd dying while we are connected to it.
3130 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3133 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3134 function for BDB 4.3.
3136 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3138 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3139 the sources for PCRE have been re-arranged and more files are now
3142 PH/15 The code I had for printing potentially long long variables in PH/11
3143 above was not the best (it lost precision). The length of off_t variables
3144 is now inspected at build time, and an appropriate printing format (%ld
3145 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3146 to be "long long int" or "long int". This is needed for the internal
3147 formatting function string_vformat().
3149 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3150 the configuration file to be ":syslog", then the script "guesses" where
3151 the logs files are, rather than using the compiled in default. In our
3152 case the guess is not the same as the compiled default, so the script
3153 suddenly stopped working when I started to use syslog. The patch checks
3154 to see if log_file_path is "". If so, it attempts to read it from exim
3155 with no configuration file to get the compiled in version, before it
3156 falls back to the previous guessing code."
3158 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3159 implementing BATV in an Exim configuration. See NewStuff for the gory
3162 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3163 Makefile that are specific to HP-UX.
3165 PH/18 If the "use_postmaster" option was set for a recipient callout together
3166 with the "random" option, the postmaster address was used as the MAIL
3167 FROM address for the random test, but not for the subsequent recipient
3168 test. It is now used for both.
3170 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3171 patch removes a few documentation additions to RFC 3028, because the
3172 latest draft now contains them. It adds the new en;ascii-case comparator
3173 and a new error check for 8bit text in MIME parts. Comparator and
3174 require names are now matched exactly. I enabled the subaddress
3175 extension, but it is not well tested yet (read: it works for me)."
3177 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3178 rework some of the code of TK/09 above to avoid the hardwired use of
3179 "%lld" and "long long". Replaced the call to snprintf() with a call to
3182 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3183 records point to non-existent hosts", "retry timeout exceeded", and
3184 "retry time not reached for any host after a long failure period".
3186 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3187 experimental DomainKeys support:
3189 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3190 (2) On an error such as an illegally used "control", the wrong name for
3191 the control was given.
3193 These problems did NOT occur unless DomainKeys support was compiled.
3195 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3197 PH/24 Added ${if match_ip condition.
3199 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3200 databases so that it will be absolutely obvious if a crash occurs in the
3201 DB library. This is a regular occurrence (often caused by mis-matched
3204 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3205 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3206 picked up on a user's system that detects such things. There doesn't seem
3207 to be a gcc warning option for this - only an attribute that has to be
3208 put on the function's prototype. It seems that in Fedora Core 4 they have
3209 set this on a number of new functions. No doubt there will be more in due
3212 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3213 specify a transport. However, if an address that was verified by such a
3214 router was the subject of a callout, Exim crashed because it tried to
3215 read the rcpt_include_affixes from the non-existent transport. Now it
3216 just assumes that the setting of that option is false. This bug was
3217 introduced by 4.51/PH/31.
3219 PH/28 Changed -d+all to exclude +memory, because that information is very
3220 rarely of interest, but it makes the output a lot bigger. People tend to
3221 do -d+all out of habit.
3223 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3224 code in os-type was giving problems when libc.so lives in lib64, like on
3227 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3228 aren't the modern standard, and it seems that some systems' include files
3229 don't always have them. Exim was already checking for some of the newer
3230 ones like T_AAAA, and defining it itself. I've added checks for all the
3231 record types that Exim uses.
3233 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3234 not automatically generating a new one, as it is supposed to. This
3235 prevented TLS from working. If the file did exist, but contained invalid
3236 data, a new version was generated, as expected. It was only the case of a
3237 non-existent file that was broken.
3239 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3240 with a change in libdomainkeys > 0.64.
3242 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3243 from DNS. If the selector record carries the flag, it now has
3244 precedence over the domain-wide flag.
3246 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3248 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3249 the use of a port name that isn't defined in /etc/services) occurred, the
3250 message was deferred as in a normal delivery, and thus remained on the
3251 spool, instead of being failed because of the mua_wrapper setting. This
3252 is now fixed, and I tidied up some of the mua_wrapper messages at the
3255 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3256 the messages in a hash of arrays rather than using individual hashes.
3257 This is a bit cleaner and results in dramatic memory savings, albeit
3258 at a slight CPU cost.
3260 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3261 as requested by Marc Sherman.
3263 SC/10 Eximstats: added histograms for user specified patterns as requested
3266 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3268 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3269 fopen() in the content-scanning modules that did not already have it.
3275 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3276 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3278 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3280 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3282 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3283 to test Sieve filters that use "vacation".
3285 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3286 that changes the way the GnuTLS parameters are stored in the cache file.
3287 The new format can be generated externally. For backward compatibility,
3288 if the data in the cache doesn't make sense, Exim assumes it has read an
3289 old-format file, and it generates new data and writes a new file. This
3290 means that you can't go back to an older release without removing the
3293 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3294 work if there are any delivery delays because "one_time" forces the
3295 parent to be marked "delivered", so its unseen clone is never tried
3296 again. For this reason, Exim now forbids the simultaneous setting of
3299 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3300 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3301 entry if you want to know the details. The fix, however, neglected to
3302 consider the case when local delivery batching is involved. The test for
3303 "previously delivered" was not happening when checking to see if an
3304 address could be batched with a previous (undelivered) one; under
3305 certain circumstances this could lead to multiple deliveries to the same
3308 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3309 in its include files, and this causes problems building Exim.
3311 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3312 header_syntax) but Exim was just ignoring anything given after a slash.
3313 In particular, this caused confusion with an attempt to use "verify =
3314 reverse_host_lookup/defer_ok". An error is now given when options are
3315 supplied for verify items that do not have them. (Maybe reverse_host_
3316 lookup should have a defer_ok option, but that's a different point.)
3318 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3319 defined by RFC 821) to 2048, because there were problems with some AUTH
3320 commands, and RFC 1869 says the size should be increased for extended
3321 SMTP commands that take arguments.
3323 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3326 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3327 "unknown" error; now it says that the functionality isn't in the binary.
3329 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3330 an address' error message when a string expansion fails (syntax or
3331 whatever). Otherwise the password may appear in the log. Following change
3332 PH/42 below, there is no longer a chance of it appearing in a bounce
3335 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3337 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3338 of its addresses was ever tried. (Bugzilla bug #2.)
3340 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3341 the result incorrectly in the debug output. (It correctly added a newline
3342 to what was transported.)
3344 TF/01 Added $received_time.
3346 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3347 commented out examples of how to interface to a virus scanner and to
3348 SpamAssassin. Also added commented examples of av_scanner and
3349 spamd_address settings.
3351 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3352 and controls are allowed in which ACLs. There were a couple of minor
3353 errors. Some of the entries in the conditions table (which is a table of
3354 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3355 negation of where the condition IS allowed.
3357 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3359 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3360 header file does not have a version number, so I've had to invent a new
3361 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3362 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3363 radiusclient), but it was contributed by a Radius user.
3365 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3366 files or directories, for OpenSSL.
3368 PH/19 When an Exim process that is running as root has to create an Exim log
3369 file, it does so in a subprocess that runs as exim:exim so as to get the
3370 ownership right at creation (otherwise, other Exim processes might see
3371 the file with the wrong ownership). There was no test for failure of this
3372 fork() call, which would lead to the process getting stuck as it waited
3373 for a non-existent subprocess. Forks do occasionally fail when resources
3374 run out. I reviewed all the other calls to fork(); they all seem to check
3377 PH/20 When checking for unexpected SMTP input at connect time (before writing
3378 the banner), Exim was not dealing correctly with a non-positive return
3379 from the read() function. If the client had disconnected by this time,
3380 the result was a log entry for a synchronization error with an empty
3381 string after "input=" when read() returned zero. If read() returned -1
3382 (an event I could not check), uninitialized data bytes were printed.
3383 There were reports of junk text (parts of files, etc) appearing after
3386 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3388 PH/22 Added support for macro redefinition, and (re)definition in between
3389 driver and ACL definitions.
3391 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3392 forgetting to use the resulting value; it was using the unexpanded value.
3394 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3395 hadn't been configured. The fix is from Juergen Kreileder, who
3396 understands it better than I do:
3398 "Here's what I see happening with three configured cyrus_sasl
3399 authenticators configured (plain, login, cram-md5):
3401 On startup auth_cyrus_sasl_init() gets called for each of these.
3402 This means three calls to sasl_listmech() without a specified mech_list.
3403 => SASL tests which mechs of all available mechs actually work
3404 => three warnings about OTP not working
3405 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3407 With the patch, sasl_listmech() also gets called three times. But now
3408 SASL's mech_list option is set to the server_mech specified in the the
3409 authenticator. Or in other words, the answer from sasl_listmech()
3410 gets limited to just the mech you're testing for (which is different
3412 => the return list contains just 'plain' or 'login', 'cram-md5' or
3413 nothing depending on the value of ob->server_mech.
3415 I've just tested the patch: Authentication still works fine,
3416 unavailable mechs specified in the exim configuration are still
3417 caught, and the auth.log warnings about OTP are gone."
3419 PH/25 When debugging is enabled, the contents of the command line are added
3420 to the debugging output, even when log_selector=+arguments is not
3423 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3424 answer is "GNU", and only if the return is "GNU/something" is the answer
3427 PH/27 $acl_verify_message is now set immediately after the failure of a
3428 verification in an ACL, and so is available in subsequent modifiers. In
3429 particular, the message can be preserved by coding like this:
3431 warn !verify = sender
3432 set acl_m0 = $acl_verify_message
3434 Previously, $acl_verify_message was set only while expanding "message"
3435 and "log_message" when a very denied access.
3437 PH/28 Modified OS/os.c-Linux with
3439 -#ifndef OS_LOAD_AVERAGE
3440 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3442 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3443 nomenclature these days.)
3445 PH/29 Installed patch from the Sieve maintainer that adds the options
3446 sieve_useraddress and sieve_subaddress to the redirect router.
3448 PH/30 In these circumstances:
3449 . Two addresses routed to the same list of hosts;
3450 . First host does not offer TLS;
3451 . First host accepts first address;
3452 . First host gives temporary error to second address;
3453 . Second host offers TLS and a TLS session is established;
3454 . Second host accepts second address.
3455 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3456 and peerdn, if requested) that were in fact used only for the second
3459 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3460 attention to any local part prefix or suffix that was matched by the
3461 router that accepted the address. It now behaves in the same way as it
3462 does for delivery: the affixes are removed from the local part unless
3463 rcpt_include_affixes is set on the transport.
3465 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3466 timeout during the DATA phase of an incoming message.
3468 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3469 applied a patch sanctioned by the Sieve maintainer.
3471 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3472 the uid or gid is negative. A case of a negative gid caused this to be
3473 noticed. The fix allows for either to be negative.
3475 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3476 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3479 PH/36 The change PH/12 above was broken. Fixed it.
3481 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3482 the grounds that routing the same address twice would always produce the
3483 same answer. This might have been true once, but it is certainly no
3484 longer true now. Routing a child address may depend on the previous
3485 routing that produced that child. Some complicated redirection strategies
3486 went wrong when messages had multiple recipients, and made Exim's
3487 behaviour dependent on the order in which the addresses were given.
3489 I have moved the duplicate checking until after the routing is complete.
3490 Exim scans the addresses that are assigned to local and remote
3491 transports, and removes any duplicates. This means that more work will be
3492 done, as duplicates will always all be routed, but duplicates are
3493 presumably rare, so I don't expect this is of any significance.
3495 For deliveries to pipes, files, and autoreplies, the duplicate checking
3496 still happens during the routing process, since they are not going to be
3499 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3500 It corrects a timeout issue with spamd. This is Ian's comment: "The
3501 background is that sometimes spamd either never reads data from a
3502 connection it has accepted, or it never writes response data. The exiscan
3503 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3504 blindly assumes that writes won't block so it may never time out."
3506 PH/39 Allow G after quota size as well as K and M.
3508 PH/40 The value set for $authenticated_id in an authenticator may not contain
3509 binary zeroes or newlines because the value is written to log lines and
3510 to spool files. There was no check on this. Now the value is run through
3511 the string_printing() function so that such characters are converted to
3512 printable escape sequences.
3514 PH/41 $message_linecount is a new variable that contains the total number of
3515 lines in the message. Compare $body_linecount, which is the count for the
3518 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3519 bounce and delay warning messages, except in certain special cases, which
3522 (a) An SMTP error message from a remote host;
3523 (b) A message specified in a :fail: redirection;
3524 (c) A message specified in a "fail" command in a system filter;
3525 (d) A message specified in a FAIL return from the queryprogram router;
3526 (e) A message specified by the cannot_route_message router option.
3528 In these cases only, Exim does include the error details in bounce and
3529 warning messages. There are also a few cases where bland messages such
3530 as "unrouteable address" or "local delivery error" are given.
3532 PH/43 $value is now also set for the "else" part of a ${run expansion.
3534 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3535 being worked on, but at least Exim now implements the latest version to
3538 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3539 process to complete was treated as a delivery failure, a timeout while
3540 writing the message to the pipe was logged, but erroneously treated as a
3541 successful delivery. Such timeouts include transport filter timeouts. For
3542 consistency with the overall process timeout, these timeouts are now
3543 treated as errors, giving rise to delivery failures by default. However,
3544 there is now a new Boolean option for the pipe transport called
3545 timeout_defer, which, if set TRUE, converts the failures into defers for
3546 both kinds of timeout. A transport filter timeout is now identified in
3549 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3550 systems where "make" and "gmake" are different, calling "gmake" at top
3551 level broke things. I've arranged for the value of $(MAKE) to be passed
3552 from the Makefile to this script so that it can call the same version of
3556 A note about Exim versions 4.44 and 4.50
3557 ----------------------------------------
3559 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3560 changes of various kinds. As a consequence, a big documentation update was
3561 needed. This delayed the release for rather longer than seemed good, especially
3562 in the light of a couple of (minor) security issues. Therefore, the changes
3563 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3564 release. So 4.44 and 4.50 are in effect two different branches that both start
3567 I have left the 4.50 change log unchanged; it contains all the changes since
3568 4.43. The change log for 4.44 is below; many of its items are identical to
3569 those for 4.50. This seems to be the most sensible way to preserve the
3570 historical information.
3576 1. Minor wording change to the doc/README.SIEVE file.
3578 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3579 computation of the current number of files was incorrect.
3581 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3582 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3585 4. Give more explanation in the error message when the command for a transport
3586 filter fails to execute.
3588 5. There are several places where Exim runs a non-Exim command in a
3589 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3590 was being done only for the command run by the queryprogram router. It is
3591 now done for all such subprocesses. The other cases are: ${run, transport
3592 filters, and the commands run by the lmtp and pipe transports.
3594 6. Added CONFIGURE_GROUP build-time option.
3596 7. Some older OS have a limit of 256 on the maximum number of file
3597 descriptors. Exim was using setrlimit() to set 1000 as a large value
3598 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3599 systems. I've change it so that if it can't get 1000, it tries for 256.
3601 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3602 was an oversight, and furthermore, ever since the addition of extra
3603 controls (e.g. 4.43/32), the checks on when to allow different forms of
3604 "control" were broken. There should now be diagnostics for all cases when a
3605 control that does not make sense is encountered.
3607 9. Added the /retain_sender option to "control=submission".
3609 10. $recipients is now available in the predata ACL (oversight).
3611 11. Tidy the search cache before the fork to do a delivery from a message
3612 received from the command line. Otherwise the child will trigger a lookup
3613 failure and thereby defer the delivery if it tries to use (for example) a
3614 cached ldap connection that the parent has called unbind on.
3616 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3617 of $address_data from the recipient verification was clobbered by the
3618 sender verification.
3620 13. The value of address_data from a sender verification is now available in
3621 $sender_address_data in subsequent conditions in the ACL statement.
3623 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3625 15. Added a new option "connect=<time>" to callout options, to set a different
3628 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3629 was its contents. (It was OK if the option was not defined at all.)
3631 17. A "Completed" log line is now written for messages that are removed from
3632 the spool by the -Mrm option.
3634 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3635 information about exactly what failed.
3637 19. Added -dd to debug only the daemon process.
3639 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3640 handles timeouts, both on the server side and network timeouts. Renamed the
3641 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3643 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3644 It is now set to "smtps".
3646 22. $host_address is now set to the target address during the checking of
3647 ignore_target_hosts.
3649 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3650 being passed; this would have caused $sender_host_name to have been used if
3651 matching the list had actually called for a host name (not very likely,
3652 since this list is usually IP addresses). A host name is now passed as
3655 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3656 code that specifies a non-restarting handler (typically sigaction() in
3657 modern systems) in an attempt to fix a rare and obscure crash bug.
3659 25. Narrowed the window for a race in the daemon that could cause it to ignore
3660 SIGCHLD signals. This is not a major problem, because they are used only to
3661 wake it up if nothing else does.
3663 26. A malformed maildirsize file could cause Exim to calculate negative values
3664 for the mailbox size or file count. Odd effects could occur as a result.
3665 The maildirsize information is now recalculated if the size or filecount
3668 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3669 support for a long time. Removed HAVE_SYS_VFS_H.
3671 28. Installed the latest version of exipick from John Jetmore.
3673 29. In an address list, if the pattern was not a regular expression, an empty
3674 subject address (from a bounce message) matched only if the pattern was an
3675 empty string. Non-empty patterns were not even tested. This was the wrong
3676 because it is perfectly reasonable to use an empty address as part of a
3677 database query. An empty address is now tested by patterns that are
3678 lookups. However, all the other forms of pattern expect the subject to
3679 contain a local part and a domain, and therefore, for them, an empty
3680 address still always fails if the pattern is not itself empty.
3682 30. Exim went into a mad DNS loop when attempting to do a callout where the
3683 host was specified on an smtp transport, and looking it up yielded more
3684 than one IP address.
3686 31. Re-factored the code for checking spool and log partition space into a
3687 function that finds that data and another that does the check. The former
3688 is then used to implement four new variables: $spool_space, $log_space,
3689 $spool_inodes, and $log_inodes.
3691 32. The RFC2047 encoding function was originally intended for short strings
3692 such as real names; it was not keeping to the 75-character limit for
3693 encoded words that the RFC imposes. It now respects the limit, and
3694 generates multiple encoded words if necessary. To be on the safe side, I
3695 have increased the buffer size for the ${rfc2047: expansion operator from
3698 33. It is now permitted to omit both strings after an "if" condition; if the
3699 condition is true, the result is "true". As before, when the second string
3700 is omitted, a false condition yields an empty string. This makes it less
3701 cumbersome to write custom ACL and router conditions.
3703 34. Failure to deliver a bounce message always caused it to be frozen, even if
3704 there was an errors_to setting on the router. The errors_to setting is now
3707 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3708 canonical form (fully expanded) before being placed in
3709 $sender_host_address.
3711 36. The table in the code that translates DNS record types into text (T_A to
3712 "A" for instance) was missing entries for NS and CNAME. It is just possible
3713 that this could have caused confusion if both these types were looked up
3714 for the same domain, because the text type is used as part of Exim's
3715 per-process caching. But the chance of anyone hitting this buglet seems
3718 37. The dnsdb lookup has been extended in a number of ways.
3720 (1) There is a new type, "zns", which walks up the domain tree until it
3721 finds some nameserver records. It should be used with care.
3723 (2) There is a new type, "mxh", which is like "mx" except that it returns
3724 just the host names, not the priorities.
3726 (3) It is now possible to give a list of domains (or IP addresses) to be
3727 looked up. The behaviour when one of the lookups defers can be
3728 controlled by a keyword.
3730 (4) It is now possible to specify the separator character for use when
3731 multiple records are returned.
3733 38. The dnslists ACL condition has been extended: it is now possible to supply
3734 a list of IP addresses and/or domains to be looked up in a particular DNS
3737 39. Added log_selector=+queue_time_overall.
3739 40. When running the queue in the test harness, wait just a tad after forking a
3740 delivery process, to get repeatability of debugging output.
3742 41. Include certificate and key file names in error message when GnuTLS fails
3743 to set them up, because the GnuTLS error message doesn't include the name
3744 of the failing file when there is a problem reading it.
3746 42. Allow both -bf and -bF in the same test run.
3748 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3750 44. The "Exiscan patch" is now merged into the mainline Exim source.
3752 45. Sometimes the final signoff response after QUIT could fail to get
3753 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3754 < 0 before doing a fflush(). This bug looks as though it goes back to the
3755 introduction of TLS in release 3.20, but "sometimes" must have been rare
3756 because the tests only now provoked it.
3758 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3759 (this can affect the format of dates).
3761 47. exim_tidydb, when checking for the continued existence of a message for
3762 which it has found a message-specific retry record, was not finding
3763 messages that were in split spool directories. Consequently, it was
3764 deleting retry records that should have stayed in existence.
3766 48. Steve fixed some bugs in eximstats.
3768 49. The SPA authentication driver was not abandoning authentication and moving
3769 on to the next authenticator when an expansion was forced to fail,
3770 contradicting the general specification for all authenticators. Instead it
3771 was generating a temporary error. It now behaves as specified.
3773 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3774 (the order specifies the preference for clients). The order is now AES256,
3775 AES128, 3DES, ARCFOUR128.
3777 51. Small patch to Sieve code - explicitly set From: when generating an
3780 52. Exim crashed if a remote delivery caused a very long error message to be
3781 recorded - for instance if somebody sent an entire SpamAssassin report back
3782 as a large number of 550 error lines. This bug was coincidentally fixed by
3783 increasing the size of one of Exim's internal buffers (big_buffer) that
3784 happened as part of the Exiscan merge. However, to be on the safe side, I
3785 have made the code more robust (and fixed the comments that describe what
3788 53. Now that there can be additional text after "Completed" in log lines (if
3789 the queue_time_overall log selector is set), a one-byte patch to exigrep
3790 was needed to allow it to recognize "Completed" as not the last thing in
3793 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3794 patch that reportedly fixes this has been added. I am not expert enough to
3795 create a test for it. This is what the patch creator wrote:
3797 "I found a little strange behaviour of ldap code when working with
3798 Windows 2003 AD Domain, where users was placed in more than one
3799 Organization Units. When I tried to give exim partial DN, the exit code
3800 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3801 But simultaneously result of request was absolutely normal ldap result,
3802 so I produce this patch..."
3804 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3805 REFERENCE, so I have modified the code to exclude the patch when that macro
3808 55. Some experimental protocols are using DNS PTR records for new purposes. The
3809 keys for these records are domain names, not reversed IP addresses. The
3810 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3811 leaves it alone. Component reversal etc. now happens only for IP addresses.
3814 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3816 57. Double the size of the debug message buffer (to 2048) so that more of very
3817 long debug lines gets shown.
3819 58. The exicyclog utility now does better if the number of log files to keep
3820 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3822 59. Two changes related to the smtp_active_hostname option:
3824 (1) $smtp_active_hostname is now available as a variable.
3825 (2) The default for smtp_banner uses $smtp_active_hostname instead
3826 of $primary_hostname.
3828 60. The host_aton() function is supposed to be passed a string that is known
3829 to be a valid IP address. However, in the case of IPv6 addresses, it was
3830 not checking this. This is a hostage to fortune. Exim now panics and dies
3831 if the condition is not met. A case was found where this could be provoked
3832 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3833 components; fortuitously, this particular loophole had already been fixed
3834 by change 4.50/55 above.
3836 If there are any other similar loopholes, the new check in host_aton()
3837 itself should stop them being exploited. The report I received stated that
3838 data on the command line could provoke the exploit when Exim was running as
3839 exim, but did not say which command line option was involved. All I could
3840 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3841 running as the user.
3844 61. There was a buffer overflow vulnerability in the SPA authentication code
3845 (which came originally from the Samba project). I have added a test to the
3846 spa_base64_to_bits() function which I hope fixes it.
3849 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3850 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3852 63. The daemon start-up calls getloadavg() while still root for those OS that
3853 need the first call to be done as root, but it missed one case: when
3854 deliver_queue_load_max is set with deliver_drop_privilege. This is
3855 necessary for the benefit of the queue runner, because there is no re-exec
3856 when deliver_drop_privilege is set.
3858 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3859 This has been fixed.
3861 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3862 was in use, was not putting the data itself into the right store pool;
3863 consequently, it could be overwritten for a subsequent message in the same
3864 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3867 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3869 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3870 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3871 the function were treating the return as a boolean value, which happened to
3872 work because 0=false and not-0=true, but is not correct code.
3874 68. The host_aton() function was not handling scoped IPv6 addresses (those
3875 with, for example, "%eth0" on the end) correctly.
3877 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3878 negated items (that is, ~something) in unsigned ints. Some compilers
3879 apparently mutter when there is no cast.
3881 70. If an address verification called from an ACL failed, and did not produce a
3882 user-specific message (i.e. there was only a "system" message), nothing was
3883 put in $acl_verify_message. In this situation, it now puts the system
3886 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3887 session; change 4.31/43 added the unwanted input to the log line - except
3888 that it did not do this in the start of session case. It now does.
3890 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3891 This is wrong and can cause the other end to generate a synchronization
3892 error if it is another Exim or anything else that does the synchronization
3893 check. A QUIT command is no longer sent after a timeout.
3895 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3896 during host lookups.
3898 74. The defer_ok option of callout verification was not working if it was used
3899 when verifying addresses in header lines, that is, for this case:
3901 verify = header_sender/callout=defer_ok
3903 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3904 those file descriptors could be used for SMTP connections. If anything
3905 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3906 could be sent to the SMTP client, causing chaos. The daemon now opens
3907 stdin, stdout, and stderr to /dev/null when it puts itself into the
3910 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3911 log by default. The user can override this with suitable Perl magic.
3913 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3914 the log message when discard triggers, was not working for the DATA ACL or
3915 for the non-SMTP ACL.
3917 78. Error message wording change in sieve.c.
3919 79. If smtp_accept_max_per_host was set, the number of connections could be
3920 restricted to fewer than expected, because the daemon was trying to set up
3921 a new connection before checking whether the processes handling previous
3922 connections had finished. The check for completed processes is now done
3923 earlier. On busy systems, this bug wouldn't be noticed because something
3924 else would have woken the daemon, and it would have reaped the completed
3927 80. If a message was submitted locally by a user whose login name contained one
3928 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3929 It caused a spool format error. I have fixed the spool reading code. A
3930 related problem was that the "from" clause in the Received: line became
3931 illegal because of the space(s). It is now covered by ${quote_local_part.
3933 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3936 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3938 83. Patch from Sieve maintainer to fix unterminated string problem in
3939 "vacation" handling.
3941 84. Some minor changes to the Linux configuration files to help with other
3942 OS variants using glibc.
3944 85. One more patch for Sieve to update vacation handling to latest spec.
3947 ----------------------------------------------------
3948 See the note above about the 4.44 and 4.50 releases.
3949 ----------------------------------------------------
3955 1. Change 4.43/35 introduced a bug that caused file counts to be
3956 incorrectly computed when quota_filecount was set in an appendfile
3959 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3960 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3963 3. Give more explanation in the error message when the command for a transport
3964 filter fails to execute.
3966 4. There are several places where Exim runs a non-Exim command in a
3967 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3968 was being done only for the command run by the queryprogram router. It is
3969 now done for all such subprocesses. The other cases are: ${run, transport
3970 filters, and the commands run by the lmtp and pipe transports.
3972 5. Some older OS have a limit of 256 on the maximum number of file
3973 descriptors. Exim was using setrlimit() to set 1000 as a large value
3974 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3975 systems. I've change it so that if it can't get 1000, it tries for 256.
3977 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3978 was an oversight, and furthermore, ever since the addition of extra
3979 controls (e.g. 4.43/32), the checks on when to allow different forms of
3980 "control" were broken. There should now be diagnostics for all cases when a
3981 control that does not make sense is encountered.
3983 7. $recipients is now available in the predata ACL (oversight).
3985 8. Tidy the search cache before the fork to do a delivery from a message
3986 received from the command line. Otherwise the child will trigger a lookup
3987 failure and thereby defer the delivery if it tries to use (for example) a
3988 cached ldap connection that the parent has called unbind on.
3990 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3991 of $address_data from the recipient verification was clobbered by the
3992 sender verification.
3994 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3995 was its contents. (It was OK if the option was not defined at all.)
3997 11. A "Completed" log line is now written for messages that are removed from
3998 the spool by the -Mrm option.
4000 12. $host_address is now set to the target address during the checking of
4001 ignore_target_hosts.
4003 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4004 being passed; this would have caused $sender_host_name to have been used if
4005 matching the list had actually called for a host name (not very likely,
4006 since this list is usually IP addresses). A host name is now passed as
4009 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4010 code that specifies a non-restarting handler (typically sigaction() in
4011 modern systems) in an attempt to fix a rare and obscure crash bug.
4013 15. Narrowed the window for a race in the daemon that could cause it to ignore
4014 SIGCHLD signals. This is not a major problem, because they are used only to
4015 wake it up if nothing else does.
4017 16. A malformed maildirsize file could cause Exim to calculate negative values
4018 for the mailbox size or file count. Odd effects could occur as a result.
4019 The maildirsize information is now recalculated if the size or filecount
4022 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4023 support for a long time. Removed HAVE_SYS_VFS_H.
4025 18. Updated exipick to current release from John Jetmore.
4027 19. Allow an empty sender to be matched against a lookup in an address list.
4028 Previously the only cases considered were a regular expression, or an
4031 20. Exim went into a mad DNS lookup loop when doing a callout where the
4032 host was specified on the transport, if the DNS lookup yielded more than
4035 21. The RFC2047 encoding function was originally intended for short strings
4036 such as real names; it was not keeping to the 75-character limit for
4037 encoded words that the RFC imposes. It now respects the limit, and
4038 generates multiple encoded words if necessary. To be on the safe side, I
4039 have increased the buffer size for the ${rfc2047: expansion operator from
4042 22. Failure to deliver a bounce message always caused it to be frozen, even if
4043 there was an errors_to setting on the router. The errors_to setting is now
4046 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4047 canonical form (fully expanded) before being placed in
4048 $sender_host_address.
4050 24. Updated eximstats to version 1.33
4052 25. Include certificate and key file names in error message when GnuTLS fails
4053 to set them up, because the GnuTLS error message doesn't include the name
4054 of the failing file when there is a problem reading it.
4056 26. Expand error message when OpenSSL has problems setting up cert/key files.
4059 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4060 (this can affect the format of dates).
4062 28. exim_tidydb, when checking for the continued existence of a message for
4063 which it has found a message-specific retry record, was not finding
4064 messages that were in split spool directories. Consequently, it was
4065 deleting retry records that should have stayed in existence.
4067 29. eximstats updated to version 1.35
4068 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4069 1.35 - bugfix such that pie charts by volume are generated correctly
4071 30. The SPA authentication driver was not abandoning authentication and moving
4072 on to the next authenticator when an expansion was forced to fail,
4073 contradicting the general specification for all authenticators. Instead it
4074 was generating a temporary error. It now behaves as specified.
4076 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4077 (the order specifies the preference for clients). The order is now AES256,
4078 AES128, 3DES, ARCFOUR128.
4080 31. Small patch to Sieve code - explicitly set From: when generating an
4083 32. Exim crashed if a remote delivery caused a very long error message to be
4084 recorded - for instance if somebody sent an entire SpamAssassin report back
4085 as a large number of 550 error lines. This bug was coincidentally fixed by
4086 increasing the size of one of Exim's internal buffers (big_buffer) that
4087 happened as part of the Exiscan merge. However, to be on the safe side, I
4088 have made the code more robust (and fixed the comments that describe what
4091 33. Some experimental protocols are using DNS PTR records for new purposes. The
4092 keys for these records are domain names, not reversed IP addresses. The
4093 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4094 leaves it alone. Component reversal etc. now happens only for IP addresses.
4097 34. The host_aton() function is supposed to be passed a string that is known
4098 to be a valid IP address. However, in the case of IPv6 addresses, it was
4099 not checking this. This is a hostage to fortune. Exim now panics and dies
4100 if the condition is not met. A case was found where this could be provoked
4101 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4102 components; fortuitously, this particular loophole had already been fixed
4103 by change 4.50/55 or 4.44/33 above.
4105 If there are any other similar loopholes, the new check in host_aton()
4106 itself should stop them being exploited. The report I received stated that
4107 data on the command line could provoke the exploit when Exim was running as
4108 exim, but did not say which command line option was involved. All I could
4109 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4110 running as the user.
4113 35. There was a buffer overflow vulnerability in the SPA authentication code
4114 (which came originally from the Samba project). I have added a test to the
4115 spa_base64_to_bits() function which I hope fixes it.
4118 36. The daemon start-up calls getloadavg() while still root for those OS that
4119 need the first call to be done as root, but it missed one case: when
4120 deliver_queue_load_max is set with deliver_drop_privilege. This is
4121 necessary for the benefit of the queue runner, because there is no re-exec
4122 when deliver_drop_privilege is set.
4124 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4125 was in use, was not putting the data itself into the right store pool;
4126 consequently, it could be overwritten for a subsequent message in the same
4127 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4130 38. Sometimes the final signoff response after QUIT could fail to get
4131 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4132 < 0 before doing a fflush(). This bug looks as though it goes back to the
4133 introduction of TLS in release 3.20, but "sometimes" must have been rare
4134 because the tests only now provoked it.
4140 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4141 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4142 Now it returns an integer. A number of calls were still expecting a T/F
4143 return. Fortuitously, in all cases, the tests worked in OK situations,
4144 which is the norm. However, things would have gone wrong on any write
4145 failures on the smtp file descriptor. This function is used when sending
4146 messages over SMTP and also when doing verify callouts.
4148 2. When Exim is called to do synchronous delivery of a locally submitted
4149 message (the -odf or -odi options), it no longer closes stderr before doing
4152 3. Implemented the mua_wrapper option.
4154 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4156 5. Implemented the functions header_remove(), header_testname(),
4157 header_add_at_position(), and receive_remove_recipient(), and exported them
4160 6. If an ACL "warn" statement specified the addition of headers, Exim already
4161 inserted X-ACL-Warn: at the start if there was no header name. However, it
4162 was not making this test for the second and subsequent header lines if
4163 there were newlines in the string. This meant that an invalid header could
4164 be inserted if Exim was badly configured.
4166 7. Allow an ACL "warn" statement to add header lines at the start or after all
4167 the Received: headers, as well as at the end.
4169 8. Added the rcpt_4xx retry error code.
4171 9. Added postmaster_mailfrom=xxx to callout verification option.
4173 10. Added mailfrom=xxxx to the callout verification option, for verify=
4176 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4177 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4179 12. Inserted some casts to stop certain compilers warning when using pointer
4180 differences as field lengths or precisions in printf-type calls (mostly
4181 affecting debugging statements).
4183 13. Added optional readline() support for -be (dynamically loaded).
4185 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4186 same clock tick as a message's arrival, so that its received time was the
4187 same as the "first fail" time on the retry record, and that message
4188 remained on the queue past the ultimate address timeout, every queue runner
4189 would try a delivery (because it was past the ultimate address timeout) but
4190 after another failure, the ultimate address timeout, which should have then
4191 bounced the address, did not kick in. This was a "< instead of <=" error;
4192 in most cases the first failure would have been in the next clock tick
4193 after the received time, and all would be well.
4195 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4196 being recognized when the domain list was tested by the match_domain
4197 condition in an expansion string.
4199 16. Added the ${str2b64: operator.
4201 17. Exim was always calling setrlimit() to set a large limit for the number of
4202 processes, without checking whether the existing limit was already
4203 adequate. (It did check for the limit on file descriptors.) Furthermore,
4204 errors from getrlimit() and setrlimit() were being ignored. Now they are
4205 logged to the main and panic logs, but Exim does carry on, to try to do its
4206 job under whatever limits there are.
4208 18. Imported PCRE 5.0.
4210 19. Trivial typo in log message " temporarily refused connection" (the leading
4213 20. If the log selector return_path_on_delivery was set and an address was
4214 redirected to /dev/null, the delivery process crashed because it assumed
4215 that a return path would always be set for a "successful" delivery. In this
4216 case, the whole delivery is bypassed as an optimization, and therefore no
4219 21. Internal re-arrangement: the function for sending a challenge and reading
4220 a response while authentication was assuming a zero-terminated challenge
4221 string. It's now changed to take a pointer and a length, to allow for
4222 binary data in such strings.
4224 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4226 23. Exim was not respecting finduser_retries when seeking the login of the
4227 uid under which it was called; it was always trying 10 times. (The default
4228 setting of finduser_retries is zero.) Also, it was sleeping after the final
4229 failure, which is pointless.
4231 24. Implemented tls_on_connect_ports.
4233 25. Implemented acl_smtp_predata.
4235 26. If the domain in control=submission is set empty, Exim assumes that the
4236 authenticated id is a complete email address when it generates From: or
4237 Sender: header lines.
4239 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4240 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4241 chown and chgrp in /bin and hostname in /usr/bin.
4243 28. Exim was keeping the "process log" file open after each use, just as it
4244 does for the main log. This opens the possibility of it remaining open for
4245 long periods when the USR1 signal hits a daemon. Occasional processlog
4246 errors were reported, that could have been caused by this. Anyway, it seems
4247 much more sensible not to leave this file open at all, so that is what now
4250 29. The long-running daemon process does not normally write to the log once it
4251 has entered its main loop, and it closes the log before doing so. This is
4252 so that log files can straightforwardly be renamed and moved. However,
4253 there are a couple of unusual error situations where the daemon does write
4254 log entries, and I had neglected to close the log afterwards.
4256 30. The text of an SMTP error response that was received during a remote
4257 delivery was being truncated at 512 bytes. This is too short for some of
4258 the long messages that one sometimes sees. I've increased the limit to
4261 31. It is now possible to make retry rules that apply only when a message has a
4262 specific sender, in particular, an empty sender.
4264 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4265 it possible to be selective about when SMTP synchronization is enforced.
4267 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4269 32. Implemented hosts_connection_nolog.
4271 33. Added an ACL for QUIT.
4273 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4276 35. Added mailbox_size and mailbox_filecount to appendfile.
4278 36. Added control = no_multiline_responses to ACLs.
4280 37. There was a bug in the logic of the code that waits for the clock to tick
4281 in the case where the clock went backwards by a substantial amount such
4282 that the microsecond fraction of "now" was more than the microsecond
4283 fraction of "then" (but the whole seconds number was less).
4285 38. Added support for the libradius Radius client library this is found on
4286 FreeBSD (previously only the radiusclient library was supported).
4292 1. When certain lookups returned multiple values in the form name=value, the
4293 quoting of the values was not always being done properly. Specifically:
4294 (a) If the value started with a double quote, but contained no whitespace,
4296 (b) If the value contained whitespace other than a space character (i.e.
4297 tabs or newlines or carriage returns) it was not quoted.
4298 This fix has been applied to the mysql and pgsql lookups by writing a
4299 separate quoting function and calling it from the lookup code. The fix
4300 should probably also be applied to nisplus, ibase and oracle lookups, but
4301 since I cannot test any of those, I have not disturbed their existing code.
4303 2. A hit in the callout cache for a specific address caused a log line with no
4304 reason for rejecting RCPT. Now it says "Previous (cached) callout
4305 verification failure".
4307 3. There was an off-by-one bug in the queryprogram router. An over-long
4308 return line was truncated at 256 instead of 255 characters, thereby
4309 overflowing its buffer with the terminating zero. As well as fixing this, I
4310 have increased the buffer size to 1024 (and made a note to document this).
4312 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4313 when Exim is waiting for an SMTP response from a remote server, Exim
4314 restarts its select() call on the socket, thereby resetting its timeout.
4315 This is not a problem when such interrupts are rare. Somebody set up a cron
4316 job to run exiwhat every 2 minutes, which is less than the normal select()
4317 timeout (5 or 10 minutes). This meant that the select() timeout never
4318 kicked in because it was always reset. I have fixed this by comparing the
4319 time when an interrupt arrives with the time at the start of the first call
4320 to select(). If more time than the timeout has elapsed, the interrupt is
4321 treated as a timeout.
4323 5. Some internal re-factoring in preparation for the addition of Sieve
4324 extensions (by MH). In particular, the "personal" test is moved to a
4325 separate function, and given an option for scanning Cc: and Bcc: (which is
4326 not set for Exim filters).
4328 6. When Exim created an email address using the login of the caller as the
4329 local part (e.g. when creating a From: or Sender: header line), it was not
4330 quoting the local part when it contained special characters such as @.
4332 7. Installed new OpenBSD configuration files.
4334 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4335 try to make them clearer.
4337 9. Callout options, other than the timeout value, were being ignored when
4338 verifying sender addresses in header lines. For example, when using
4340 verify = header_sender/callout=no_cache
4342 the cache was (incorrectly) being used.
4344 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4345 only the Cygwin environment.
4347 11. When return_path_on_delivery was set as a log selector, if different remote
4348 addresses in the same message used different return paths and parallel
4349 remote delivery occurred, the wrong values would sometimes be logged.
4350 (Whenever a remote delivery process finished, the return path value from
4351 the most recently started remote delivery process was logged.)
4353 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4354 lines when AUTH and/or TLS are in use. This is the "received protocol"
4355 field. Exim used to use "asmtp" for authenticated SMTP, without any
4356 indication (in the protocol name) for TLS use. Now it follows the RFC and
4357 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4358 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4359 names appear in log lines as well as in Received: header lines.
4361 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4362 extensions, and comparison tests, and to fix some bugs.
4364 14. Changes to the "personal" filter test:
4366 (1) The test was buggy in that it was just doing the equivalent of
4367 "contains" tests on header lines. For example, if a user's address was
4368 anne@some.where, the "personal" test would incorrectly be true for
4370 To: susanne@some.where
4372 This test is now done by extracting each address from the header in turn,
4373 and checking the entire address. Other tests that are part of "personal"
4374 are now done using regular expressions (for example, to check local parts
4375 of addresses in From: header lines).
4377 (2) The list of non-personal local parts in From: addresses has been
4378 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4379 taken from the Sieve specification recommendations.
4381 (3) If the message contains any header line starting with "List-" it is
4382 treated as non-personal.
4384 (4) The test for "circular" in the Subject: header line has been removed
4385 because it now seems ill-conceived.
4387 15. Minor typos in src/EDITME comments corrected.
4389 16. Installed latest exipick from John Jetmore.
4391 17. If headers_add on a router specified a text string that was too long for
4392 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4393 of string_sprintf() is now avoided.
4395 18. $message_body_size was not set (it was always zero) when running the DATA
4396 ACL and the local_scan() function.
4398 19. For the "mail" command in an Exim filter, no default was being set for
4399 the once_repeat time, causing a random time value to be used if "once" was
4400 specified. (If the value happened to be <= 0, no repeat happened.) The
4401 default is now 0s, meaning "never repeat". The "vacation" command was OK
4402 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4403 (I found it when inspecting the code).
4405 20. There is now an overall timeout for performing a callout verification. It
4406 defaults to 4 times the callout timeout, which applies to individual SMTP
4407 commands during the callout. The overall timeout applies when there is more
4408 than one host that can be tried. The timeout is checked before trying the
4409 next host. This prevents very long delays if there are a large number of
4410 hosts and all are timing out (e.g. when the network connections are timing
4411 out). The value of the overall timeout can be changed by specifying an
4412 additional sub-option for "callout", called "maxwait". For example:
4414 verify = sender/callout=5s,maxwait=20s
4416 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4417 to the end before writing, but this should make it even safer).
4419 22. Exim was forgetting that it had advertised PIPELINING for the second and
4420 subsequent messages on an SMTP connection. It was also not resetting its
4421 memory on STARTTLS and an internal HELO.
4423 23. When Exim logs an SMTP synchronization error within a session, it now
4424 records whether PIPELINING has been advertised or not.
4426 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4427 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4428 rather than long int.
4430 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4432 26. Added the never_mail option to autoreply.
4438 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4439 crash if the getsockname() call failed; this can happen if a connection is
4440 closed very soon after it is established. The problem was simply in the
4441 order in which certain operations were done, causing Exim to try to write
4442 to the SMTP stream before it had set up the file descriptor. The bug has
4443 been fixed by making things happen in the correct order.
4449 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4450 before the connection was closed, thus losing the rejection response.
4452 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4453 some early Solaris releases, but causes trouble in current releases where
4454 socklen_t is defined.
4456 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4459 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4462 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4463 that releasing the top of it at the end releases what was used for sub-
4464 expansions (unless the block got too big). However, discard this block if
4465 the first thing is a variable or header, so that we can use its block when
4466 it is dynamic (useful for very large $message_headers, for example).
4468 6. Lookups now cache *every* query, not just the most recent. A new, separate
4469 store pool is used for this. It can be recovered when all lookup caches are
4470 flushed. Lookups now release memory at the end of their result strings.
4471 This has involved some general refactoring of the lookup sources.
4473 7. Some code has been added to the store_xxx() functions to reduce the amount
4474 of flapping under certain conditions.
4476 8. log_incoming_interface used to affect only the <= reception log lines. Now
4477 it causes the local interface and port to be added to several more SMTP log
4478 lines, for example "SMTP connection from", and rejection lines.
4480 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4482 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4484 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4485 could be overwritten at the end of the current message (or the start of a
4486 new message if it was set in a HELO ACL). The value is now preserved for
4487 the duration of the SMTP connection.
4489 12. If a transport had a headers_rewrite setting, and a matching header line
4490 contained an unqualified address, that address was qualified, even if it
4491 did not match any rewriting rules. The underlying bug was that the values
4492 of the flags that permit the existence of unqualified sender and recipient
4493 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4494 non-local messages, and by -bnq for local messages) were not being
4495 preserved with the message after it was received.
4497 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4498 "next input=" as part of the text comprising the host identity instead of
4499 the correct text. The code was using the same buffer for two different
4500 strings. However, depending on which order the printing function evaluated
4501 its arguments, the bug did not always show up. Under Linux, for example, my
4502 test suite worked just fine.
4504 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4505 doesn't work with some older versions of Perl. It has been changed to "my",
4506 which in any case is probably the better facility to use.
4508 15. A really picky compiler found some instances of statements for creating
4509 error messages that either had too many or two few arguments for the format
4512 16. The size of the buffer for calls to the DNS resolver has been increased
4513 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4514 for addresses that have a lot of PTR records. This alleviates a problem; it
4515 does not fully solve it.
4517 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4518 buffer now truncates the list and logs the incident, which is the same
4519 action as happens when Exim is looking up a host name and its aliases.
4520 Previously in this situation something unpredictable would happen;
4521 sometimes it was "internal error: store_reset failed".
4523 18. If a server dropped the connection unexpectedly when an Exim client was
4524 using GnuTLS and trying to read a response, the client delivery process
4525 crashed while trying to generate an error log message.
4527 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4528 string, for example:
4530 warn message = H1: something\nH2: something
4532 the text was added as a single header line from Exim's point of view
4533 though it ended up OK in the delivered message. However, searching for the
4534 second and subsequent header lines using $h_h2: did not work. This has been
4535 fixed. Similarly, if a system filter added multiple headers in this way,
4536 the routers could not see them.
4538 20. Expanded the error message when iplsearch is called with an invalid key to
4539 suggest using net-iplsearch in a host list.
4541 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4542 ACLs are no longer actually imposed (and a message to that effect is
4545 22. If a "gecos" field in a passwd entry contained escaped characters, in
4546 particular, if it contained a \" sequence, Exim got it wrong when building
4547 a From: or a Sender: header from that name. A second bug also caused
4548 incorrect handling when an unquoted " was present following a character
4549 that needed quoting.
4551 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4552 was not being matched caselessly.
4554 24. Arranged for all hyphens in the exim.8 source to be escaped with
4557 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4558 properly. Recipient callout cache records were still being keyed to include
4559 the sender, even when use_sender was set false. This led to far more
4560 callouts that were necessary. The sender is no longer included in the key
4561 when use_sender is false.
4563 26. Added "control = submission" modifier to ACLs.
4565 27. Added the ${base62d: operator to decode base 62 numbers.
4567 28. dnsdb lookups can now access SRV records.
4569 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4570 the configuration file.
4572 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4573 (-v) mode. This makes the output for a verbose queue run more intelligible.
4575 31. Added a use_postmaster feature to recipient callouts.
4577 32. Added the $body_zerocount variable, containing the number of binary zero
4578 bytes in the message body.
4580 33. The time of last modification of the "new" subdirectory is now used as the
4581 "mailbox time last read" when there is a quota error for a maildir
4584 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4586 35. Added +ignore_unknown as a special item in host lists.
4588 36. Code for decoding IPv6 addresses in host lists is now included, even if
4589 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4590 address was recognized as an IP address, but was then not correctly decoded
4591 into binary, causing unexpected and incorrect effects when compared with
4598 1. Very minor rewording of debugging text in manualroute to say "list of
4599 hosts" instead of "hostlist".
4601 2. If verify=header_syntax was set, and a header line with an unqualified
4602 address (no domain) and a large number of spaces between the end of the
4603 name and the colon was received, the reception process suffered a buffer
4604 overflow, and (when I tested it) crashed. This was caused by some obsolete
4605 code that should have been removed. The fix is to remove it!
4607 3. When running in the test harness, delay a bit after writing a bounce
4608 message to get a bit more predictability in the log output.
4610 4. Added a call to search_tidyup() just before forking a reception process. In
4611 theory, someone could use a lookup in the expansion of smtp_accept_max_
4612 per_host which, without the tidyup, could leave open a database connection.
4614 5. Added the variables $recipient_data and $sender_data which get set from a
4615 lookup success in an ACL "recipients" or "senders" condition, or a router
4616 "senders" option, similar to $domain_data and $local_part_data.
4618 6. Moved the writing of debug_print from before to after the "senders" test
4621 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4622 problems for message scanning, either using a data ACL, or using
4623 local_scan() because the Received: header was not generated till after they
4624 were called (in order to set the time as the time of reception completion).
4625 I have revised the way this works. The header is now generated after the
4626 body is received, but before the ACL or local_scan() are called. After they
4627 are run, the timestamp in the header is updated.
4633 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4634 before starting a queue runner without re-exec. This happened only when
4635 deliver_drop_privilege was set or when the Exim user was set to root. The
4636 effect of the bug was that timeouts during subsequent deliveries caused
4637 crashes instead of being properly handled. The handler is now left at its
4638 default (and expected) setting.
4640 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4641 message, again when deliver_drop_privilege is set or Exim is run as root.
4642 The bug described in (1) was not present in this case, but the tidying up
4643 of the other signals was missing. I have made the two cases consistent.
4645 3. The ignore_target_hosts setting on a manualroute router was being ignored
4646 for hosts that were looked up using the /MX notation.
4648 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4651 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4652 operated on the sender address. After changing the $sender_address to <>
4653 for the sender address verify, Exim was re-instated it as the original
4654 (before rewriting) address, but remembering that it had rewritten it, so it
4655 wasn't rewriting it again. This bug also had the effect of breaking the
4656 sender address verification caching when the sender address was rewritten.
4658 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4659 This has been changed so that if the ip literal address matches
4660 ignore_target_hosts, the router declines.
4662 7. Added expansion conditions match_domain, match_address, and match_local_
4663 part (NOT match_host).
4665 8. The placeholder for the Received: header didn't have a length field set.
4667 9. Added code to Exim itself and to exim_lock to test for a specific race
4668 condition that could lead to file corruption when using MBX delivery. The
4669 issue is with the lockfile that is created in /tmp. If this file is removed
4670 after a process has opened it but before that process has acquired a lock,
4671 there is the potential for a second process to recreate the file and also
4672 acquire a lock. This could lead to two Exim processes writing to the file
4673 at the same time. The added code performs the same test as UW imapd; it
4674 checks after acquiring the lock that its file descriptor still refers to
4675 the same named file.
4677 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4678 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4681 11. Added the smtp_active_hostname option. If used, this will typically be made
4682 to depend on the incoming interface address. Because $interface_address is
4683 not set up until the daemon has forked a reception process, error responses
4684 that can happen earlier (such as "too many connections") no longer contain
4687 12. If an expansion in a condition on a "warn" statement fails because a lookup
4688 defers, the "warn" statement is abandoned, and the next ACL statement is
4689 processed. Previously this caused the whole ACL to be aborted.
4691 13. Added the iplsearch lookup type.
4693 14. Added ident_timeout as a log selector.
4695 15. Added tls_certificate_verified as a log selector.
4697 16. Added a global option tls_require_ciphers (compare the smtp transport
4698 option of the same name). This controls incoming TLS connections.
4700 17. I finally figured out how to make tls_require_ciphers do a similar thing
4701 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4702 before starting the TLS session.
4704 18. Tabs are now shown as \t in -bP output.
4706 19. If the log selector return_path_on_delivery was set, Exim crashed when
4707 bouncing a message because it had too many Received: header lines.
4709 20. If two routers both had headers_remove settings, and the first one included
4710 a superfluous trailing colon, the final name in the first list and the
4711 first name in the second list were incorrectly joined into one item (with a
4712 colon in the middle).
4718 1. Added -C and -D options to the exinext utility, mainly to make it easier
4719 to include in the automated testing, but these could be helpful when
4720 multiple configurations are in use.
4722 2. The exinext utility was not formatting the output nicely when there was
4723 an alternate port involved in the retry record key, nor when there was a
4724 message id as well (for retries that were specific to a specific message
4725 and a specific host). It was also confused by IPv6 addresses, because of
4726 the additional colons they contain. I have fixed the IPv4 problem, and
4727 patched it up to do a reasonable job for IPv6.
4729 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4730 delivery, the log line now contains "pipelined" if PIPELINING was used.
4732 4. An SMTP transport process used to panic and die if the bind() call to set
4733 an explicit outgoing interface failed. This has been changed; it is now
4734 treated in the same way as a connect() failure.
4736 5. A reference to $sender_host_name in the part of a conditional expansion
4737 that was being skipped was still causing a DNS lookup. This no longer
4740 6. The def: expansion condition was not recognizing references to header lines
4741 that used bh_ and bheader_.
4743 7. Added the _cache feature to named lists.
4745 8. The code for checking quota_filecount in the appendfile transport was
4746 allowing one more file than it should have been.
4748 9. For compatibility with Sendmail, the command line option
4756 and sets the incoming protocol and host name (for trusted callers). The
4757 host name and its colon can be omitted when only the protocol is to be set.
4758 Note the Exim already has two private options, -pd and -ps, that refer to
4759 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4760 "s", but I don't think that's a major issue.
4762 10. A number of refactoring changes to the code, none of which should affect
4765 (a) The number of logging options was getting close to filling up the
4766 32-bit word that was used as a bit map. I have split them into two classes:
4767 those that are passed in the argument to log_write(), and those that are
4768 only ever tested independently outside of that function. These are now in
4769 separate 32-bit words, so there is plenty of room for expansion again.
4770 There is no change in the user interface or the logging behaviour.
4772 (b) When building, for example, log lines, the code previously used a
4773 macro that called string_cat() twice, in order to add two strings. This is
4774 not really sufficiently general. Furthermore, there was one instance where
4775 it was actually wrong because one of the argument was used twice, and in
4776 one call a function was used. (As it happened, calling the function twice
4777 did not affect the overall behaviour.) The macro has been replaced by a
4778 function that can join an arbitrary number of extra strings onto a growing
4781 (c) The code for expansion conditions now uses a table and a binary chop
4782 instead of a serial search (which was left over from when there were very
4783 few conditions). Also, it now recognizes conditions like "pam" even when
4784 the relevant support is not compiled in: a suitably worded error message is
4785 given if an attempt is made to use such a condition.
4787 11. Added ${time_interval:xxxxx}.
4789 12. A bug was causing one of the ddress fields not to be passed back correctly
4790 from remote delivery subprocesses. The field in question was not being
4791 subsequently used, so this caused to problems in practice.
4793 13. Added new log selectors queue_time and deliver_time.
4795 14. Might have fixed a bug in maildirsizefile handling that threw up
4796 "unexpected character" debug warnings, and recalculated the data
4797 unnecessarily. In any case, I expanded the warning message to give more
4800 15. Added the message "Restricted characters in address" to the statements in
4801 the default ACL that block characters like @ and % in local parts.
4803 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4804 Three changes have been made:
4806 (a) There was a serious bug; a negative response to MAIL caused the whole
4807 recipient domain to be cached as invalid, thereby blocking all messages
4808 to all local parts at the same domain, from all senders. This bug has
4809 been fixed. The domain is no longer cached after a negative response to
4810 MAIL if the sender used is not empty.
4812 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4815 (c) A new callout option, "use_sender" has been added for people who want
4816 the modified behaviour.
4822 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4825 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4826 indeed breaks things for older releases.
4828 3. Added additional logging to the case where there is a problem reading data
4829 from a filter that is running in a subprocess using a pipe, in order to
4830 try to track down a specific problem.
4832 4. Testing facility fudge: when running in the test harness and attempting
4833 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4834 getting "No route to host". Convert this to a timeout.
4836 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4839 6. Some OS don't have socklen_t but use size_t instead. This affects the
4840 fifth argument of getsockopt() amongst other things. This is now
4841 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4842 can be set for individual OS. I have set it for SunOS5, OSF1, and
4843 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4844 some earlier ones do not.
4846 7. Change 4.30/15 was not doing the test caselessly.
4848 8. The standard form for an IPv6 address literal was being rejected by address
4849 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4850 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4851 this, as well as the form without the "IPv6" on the front (but only when
4852 address literals are enabled, of course).
4854 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4856 10. Exim crashed if a message with an empty sender address specified by -f
4857 encountered a router with an errors_to setting. This could be provoked only
4858 by a command such as
4862 where an empty string was supplied; "<>" did not hit this bug.
4864 11. Installed PCRE release 4.5.
4866 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4867 remained set. It is now erased.
4869 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4870 times from message ids (which are base 36 rather than the normal 62).
4872 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4873 were being counted as actual protocol errors, and logged if the log
4874 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4875 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4876 and DATA following a set of rejected RCPTs do not count as protocol errors.
4877 In other words, Exim assumes they were pipelined, though this may not
4878 actually be the case. Of course, in all cases the client gets an
4879 appropriate error code.
4881 15. If a lookup fails in an ACL condition, a message about the failure may
4882 be available; it is used if testing the ACL cannot continue, because most
4883 such messages specify what the cause of the deferral is. However, some
4884 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4885 that caused an old message to be retained and used if a later statement
4886 caused a defer, replacing the real cause of the deferral.
4888 16. If an IP address had so many PTR records that the DNS lookup buffer
4889 was not large enough to hold them, Exim could crash while trying to process
4890 the truncated data. It now detects and logs this case.
4892 17. Further to 4.21/58, another change has been made: if (and only if) the
4893 first line of a message (the first header line) ends with CRLF, a bare LF
4894 in a subsequent header line has a space inserted after it, so as not to
4895 terminate the header.
4897 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4898 unnecessarily, used atoi() instead of strtol(), and didn't check the
4899 termination when getting file sizes from file names by regex.
4901 19. Completely re-implemented the support for maildirsize files, in the light
4902 of a number of problems with the previous contributed implementation
4903 (4.30/29). In particular:
4905 . If the quota is zero, the maildirsize file is maintained, but no quota is
4908 . If the maildir directory does not exist, it is created before any attempt
4909 to write a maildirsize file.
4911 . The quota value in the file is just a cache; if the quota is changed in
4912 the transport, the new value overrides.
4914 . A regular expression is available for excluding directories from the
4917 20. The autoreply transport checks the characters in options that define the
4918 message's headers; it allows continued headers, but it was checking with
4919 isspace() after an embedded newline instead of explicitly looking for a
4922 21. If all the "regular" hosts to which an address was routed had passed their
4923 expiry times, and had not reached their retry times, the address was
4924 bounced, even if fallback hosts were defined. Now Exim should go on to try
4927 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4928 equivalent code in the SMTP transport. Some hosts send humungous responses
4929 to HELO/EHLO, more than 1024 it seems.
4931 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4932 gives compiler warnings in some environments. I've now done it "properly",
4935 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4936 (because of problems with the built-in one) was declared to return uschar *
4937 instead of char *, causing compiler failure.
4939 25. Fixed a file descriptor leak when processing alias/forward files.
4941 26. Fixed a minor format string issue in dbfn.c.
4943 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4945 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4946 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4949 29. When a sender address is verified, it is cached, to save repeating the test
4950 when there is more than one recipient in a message. However, when the
4951 verification involves a callout, it is possible for different callout
4952 options to be set for different recipients. It is too complicated to keep
4953 track of this in the cache, so now Exim always runs a verification when a
4954 callout is required, relying on the callout cache for the optimization.
4955 The overhead is duplication of the address routing, but this should not be
4958 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4959 to be verified with callout=postmaster, and the main callout worked but the
4960 postmaster check failed, the verification correctly failed. However, if a
4961 subsequent RCPT command asked for sender verification *without* the
4962 postmaster check, incorrect caching caused this verification also to fail,
4965 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4966 it was not caching the DNS options (qualify_single, search_parents) that
4967 were used when the lookup failed. A subsequent lookup with different
4968 options therefore always gave the same answer, though there were cases
4969 where it should not have. (Example: a "domains = !$mx_any" option on a
4970 dnslookup router: the "domains" option is always processed without any
4971 widening, but the router might have qualify_single set.) Now Exim uses the
4972 cached value only when the same options are set.
4974 32. Added John Jetmore's "exipick" utility to the distribution.
4976 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4977 than a timeout (e.g. a certificate is required, and is not provided), an
4978 Exim server now closes the connection immediately. Previously it waited for
4979 the client to close - but if the client is SSL, it seems that they each
4980 wait for each other, leading to a delay before one of them times out.
4982 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4983 maintained 0.8.x compatibility because I don't think many are using it, and
4984 it is clearly obsolete.
4986 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4989 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4990 client certificate was expired. A simple patch fixes this, though I don't
4991 understand the full logic of why the verify callback is called multiple
4994 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4995 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4996 which causes problems with some clients (such as the Certicom SSL Plus
4997 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4998 disables the coutermeasure allowing Eudora to connect."
5000 38. Exim was not checking that a write() to a log file succeeded. This could
5001 lead to Bad Things if a log got too big, in particular if it hit a file
5002 size limit. Exim now panics and dies if it cannot write to a log file, just
5003 as it does if it cannot open a log file.
5005 39. Modified OS/Makefile-Linux so that it now contains
5007 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5009 The two -D definitions ensure that Exim is compiled with large file
5010 support, which makes it possible to handle log files that are bigger than
5013 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5014 instance) a domain was checked against a named list that involved a lookup,
5015 causing $domain_data to be set, then another domain was checked against the
5016 same list, then the first domain was re-checked, the value of $domain_data
5017 after the final check could be wrong. In particular, if the second check
5018 failed, it could be set empty. This bug probably also applied to
5021 41. The strip_trailing_dot option was not being applied to the address given
5022 with the -f command-line option.
5024 42. The code for reading a message's header from the spool was incrementing
5025 $received_count, but never initializing it. This meant that the value was
5026 incorrect (doubled) while delivering a message in the same process in which
5027 it was received. In the most common configuration of Exim, this never
5028 happens - a fresh exec is done - but it can happen when
5029 deliver_drop_privilege is set.
5031 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5032 it now includes up to 150 characters of the unexpected data in the log
5035 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5036 and building data strings. The size of both of these buffers was 10 000
5037 bytes - far larger than anybody would *ever* want, thought I. Needless to
5038 say, somebody hit the limit. I have increased the maximum line length to
5039 20 000 and the maximum data length of concatenated lines to 100 000. I have
5040 also fixed two bugs, because there was no checking on these buffers. Tsk,
5041 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5042 buffer is too small.
5044 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5045 lsearch lookups. Now it does.
5047 46. When parsing a route_list item in a manualroute router, a fixed-length
5048 buffer was used for the list of hosts. I made this 1024 bytes long,
5049 thinking that nobody would ever have a list of hosts that long. Wrong.
5050 Somebody had a whole pile of complicated expansion conditions, and the
5051 string was silently truncated, leading to an expansion error. It turns out
5052 that it is easier to change to an unlimited length (owing to other changes
5053 that have happened since this code was originally written) than to build
5054 structure for giving a limitation error. The length of the item that
5055 expands into the list of hosts is now unlimited.
5057 47. The lsearch lookup could not handle data where the length of text line was
5058 more than 4095 characters. Such lines were truncated, leading to shortened
5059 data being returned. It should now handle lines of any length.
5061 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5062 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5065 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5066 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5067 no longer happens in comment lines. A list of replacements is now placed
5068 at the head of all of the source files, except those whose only change is
5069 to replace PERL_COMMAND in the very first #! line.
5071 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5072 messages on the queue, with a bottom-up merge sort, using code contributed
5073 by Michael Haardt. This should make operations like -bp somewhat faster on
5074 large queues. It won't affect queue runners, except when queue_run_in_order
5077 51. Installed eximstats 1.31 in the distribution.
5079 52. Added support for SRV lookups to the dnslookup router.
5081 53. If an ACL referred to $message_body or $message_body_end, the value was not
5082 reset for any messages that followed in the same SMTP session.
5084 54. The store-handling optimization for building very long strings was not
5085 differentiating between the different store pools. I don't think this
5086 actually made any difference in practice, but I've tidied it.
5088 55. While running the routers to verify a sender address, $sender_address
5089 was still set to the sender address. This is wrong, because when routing to
5090 send a bounce to the sender, it would be empty. Therefore, I have changed
5091 it so that, while verifying a sender address, $sender_address is set to <>.
5092 (There is no change to what happens when verifying a recipient address.)
5094 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5095 target A or AAAA records (if not already returned) without resetting the
5096 qualify_single or search_parents options of the DNS resolver. These are
5097 inappropriate in this case because the targets of MX and SRV records must
5098 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5099 target that, when qualified, matched something in the local domain. These
5100 two options are now turned off when doing these lookups.
5102 57. It seems that at least some releases of Reiserfs (which does not have the
5103 concept of a fixed number of inodes) returns zero and not -1 for the
5104 number of available inodes. This interacted badly with check_spool_inodes,
5105 which assumed that -1 was the "no such thing" setting. What I have done is
5106 to check that the total number of inodes is greater than zero before doing
5107 the test of how many are available.
5109 58. When a "warn" ACL statement has a log_message modifier, the message is
5110 remembered, and not repeated. This is to avoid a lot of repetition when a
5111 message has many recipients that cause the same warning to be written.
5112 Howewer, Exim was preserving the list of already written lines for an
5113 entire SMTP session, which doesn't seem right. The memory is now reset if a
5114 new message is started.
5116 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5117 debugging output unless log_rewrite was also set.
5119 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5120 of (int)(handle) when we know that handle contains (void *)(-1).
5122 61. The Exim daemon panic-logs an error return when it closes the incoming
5123 connection. However "connection reset by peer" seems to be common, and
5124 isn't really an error worthy of noting specially, so that particular error
5127 62. When Exim is trying to find all the local interfaces, it used to panic and
5128 die if the ioctl to get the interface flags failed. However, it seems that
5129 on at least one OS (Solaris 9) it is possible to have an interface that is
5130 included in the list of interfaces, but for which you get a failure error
5131 for this call. This happens when the interface is not "plumbed" into a
5132 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5133 failure of the "get flags" call assumes that the interface is down.
5135 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5136 makes life easier for people who are doing arithmetic on fields extracted
5137 from dates, where you often get leading zeros that should not be
5138 interpreted as octal.
5140 64. Added qualify_domain to the redirect router, to override the global
5143 65. If a pathologically long header line contained very many addresses (the
5144 report of this problem mentioned 10 000) and each of them was rewritten,
5145 Exim could use up a very large amount of memory. (It kept on making new
5146 copies of the header line as it rewrote, and never released the old ones.)
5147 At the expense of a bit more processing, the header rewriting function has
5148 been changed so that it no longer eats memory in this way.
5150 66. The generation of the Received: header has been moved from the time that a
5151 message starts to be received, to the time that it finishes. The timestamp
5152 in the Received: header should now be very close to that of the <= log
5153 line. There are two side-effects of this change:
5155 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5156 logged header lines no longer include the local Received: line, because
5157 it has not yet been created. The same applies to a copy of the message
5158 that is returned to a non-SMTP sender when a message is rejected.
5160 (b) When a filter file is tested using -bf, no additional Received: header
5161 is added to the test message. After some thought, I decided that this
5164 This change does not affect the value of $received_for. It is still set
5165 after address rewriting, but before local_scan() is called.
5167 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5169 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5170 gave an unhelpful panic error message, and a defer error. I have managed to
5171 change this behaviour so that it now rejects any supplied certificate,
5172 which seems right, as the list of acceptable certificates is empty.
5174 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5175 gave an unhelpful defer error. I have not managed to make this reject any
5176 supplied certificates, but the error message it gives is "no certificate
5177 supplied", which is not helpful.
5179 70. exigrep's output now also includes lines that are not associated with any
5180 message, but which match the given pattern. Implemented by a patch from
5181 Martin Sluka, which also tidied up the Perl a bit.
5183 71. Recipient callout verification, like sender verification, was using <> in
5184 the MAIL FROM command. This isn't really the right thing, since the actual
5185 sender may affect whether the remote host accepts the recipient or not. I
5186 have changed it to use the actual sender in the callout; this means that
5187 the cache record is now keyed on a recipient/sender pair, not just the
5188 recipient address. There doesn't seem to be a real danger of callout loops,
5189 since a callout by the remote host to check the sender would use <>.
5190 [SEE ABOVE: changed after hitting problems.]
5192 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5193 temporary errors. However, in the case of such a code being given after
5194 the end of a data transmission (i.e. after ".") Exim was failing to write
5195 a retry record for the message. (Yes, there was some broken host that was
5196 actually sending 8xx at this point.)
5198 73. An unknown lookup type in a host list could cause Exim to panic-die when
5199 the list was checked. (An example that provoked this was putting <; in the
5200 middle of a list instead of at the start.) If this happened during a DATA
5201 ACL check, a -D file could be left lying around. This kind of configuration
5202 error no longer causes Exim to die; instead it causes a defer errror. The
5203 incident is still logged to the main and panic logs.
5205 74. Buglet left over from Exim 3 conversion. The message "too many messages
5206 in one connection" was written to the rejectlog but not the mainlog, except
5207 when address rewriting (yes!) was being logged.
5209 75. Added write_rejectlog option.
5211 76. When a system filter was run not as root (that is, when system_filter_user
5212 was set), the values of the $n variables were not being returned to the
5213 main process; thus, they were not subsequently available in the $sn
5216 77. Added +return_path_on_delivery log selector.
5218 78. A connection timeout was being treated differently from recipients deferred
5219 when testing hosts_max_try with a message that was older than the host's
5220 retry timeout. (The host should not be counted, thus allowing all hosts to
5221 be tried at least once before bouncing.) This may have been the cause of an
5222 occasionally reported bug whereby a message would remain on the queue
5223 longer than the retry timeout, but would be bounced if a delivery was
5224 forced. I say "may" because I never totally pinned down the problem;
5225 setting up timeout/retry tests is difficult. See also the next item.
5227 79. The ultimate address timeout was not being applied to errors that involved
5228 a combination of host plus message (for example, a timeout on a MAIL
5229 command). When an address resolved to a number of possible hosts, and they
5230 were not all tried for each delivery (e.g. because of hosts_max_try), a
5231 message could remain on the queue longer than the retry timeout.
5233 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5236 81. Fixed an obscure SMTP outgoing bug which required at least the following
5237 conditions: (a) there was another message waiting for the same server;
5238 (b) the server returned 5xx to all RCPT commands in the first message so
5239 that the message was not completed; (c) the server dropped the connection
5240 or gave a negative response to the RSET that Exim sends to abort the
5241 transaction. The observed case was a dropped connection after DATA that had
5242 been sent in pipelining mode. That is, the server had advertised PIPELINING
5243 but was not implementing it correctly. The effect of the bug was incorrect
5244 behaviour, such as trying another host, and this could lead to a crash.
5250 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5251 and daemon.c were passed as pointers to ints; they should have been
5252 pointers to socklen_t variables (which are typically unsigned ints).
5254 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5257 3. Fixed a really odd bug that affected only the testing scheme; patching a
5258 certain fixed string in the binary changed the value of another string that
5259 happened to be identical to the end of the original first string.
5261 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5262 name", it returns that address as the IP address. On some operating
5263 systems (e.g. Solaris), it also passes back the IP address string as the
5264 "host name". However, on others (e.g. Linux), it passes back an empty
5265 string. Exim wasn't checking for this, and was changing the host name to an
5266 empty string, assuming it had been canonicized.
5268 5. Although rare, it is permitted to have more than one PTR record for a given
5269 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5270 all the names associated with an address, because they do in Solaris.
5271 However, it seems that they do not in Linux for data that comes from the
5272 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5273 I found this out when I moved to a new Linux workstation and tried to run
5274 the Exim test suite.
5276 To get round this problem I have changed the code so that it now does its
5277 own call to the DNS to look up PTR records when searching for a host name.
5278 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5279 addresses that are only in /etc/hosts are still found.
5281 This behaviour is, however, controlled by an option called host_lookup_
5282 order, which defaults to "bydns:byaddr". If people want to use the other
5283 order, or indeed, just use one or the other means of lookup, they can
5284 specify it in this variable.
5286 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5287 some operating systems, this comes back from gethostbyaddr() as an empty
5288 string, and this is what Exim used to test for. However, it seems that in
5289 other systems, "." is yielded. Exim now tests for this case too.
5291 7. The values of check_spool_space and check_log_space are now held internally
5292 as a number of kilobytes instead of an absolute number of bytes. If a
5293 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5294 kilobyte. This means that much larger values can be stored.
5296 8. Exim monitor: an attempt to get the action menu when not actually pointing
5297 at a message produces an empty menu entitled "No message selected". This
5298 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5299 no entries in it ("Shell widget menu has zero width and/or height"). So I
5300 have added a single, blank menu entry in this case.
5302 9. Added ${quote_local_part.
5304 10. MIME decoding is now applied to the contents of Subject: header lines when
5307 11. Now that a reference to $sender_host_address automatically causes a reverse
5308 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5309 host lookup before query-style lookups in lists that might use this
5310 variable. This has therefore been abolished, and the "net-" prefix is no
5311 longer necessary for query-style lookups.
5313 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5314 have been a typo for LFLAGS, so it has been changed.
5316 13. The install script calls Exim with "-C /dev/null" in order to find the
5317 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5318 to be output. Howeve, since Exim outputs its version number before the
5319 error, it didn't break the script. It just looked ugly. I fixed this by
5320 always allowing "-C /dev/null" if the caller is root.
5322 14. Ignore overlarge ACL variable number when reading spool file - insurance
5323 against a later release with more variables having written the file.
5325 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5326 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5327 this, as well as the form without the "IPv6" on the front.
5329 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5330 OS/Makefile-Darwin file.
5332 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5333 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5335 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5336 variable. However, some OS use ints for inodes. I've added cast to long int
5337 to get rid of the compiler warning.
5339 19. I had forgotten to lock out "/../" in configuration file names when
5340 ALT_CONFIG_PREFIX was set.
5342 20. Routers used for verification do not need to specify transports. However,
5343 if such a router generated a host list, and callout was configured, Exim
5344 crashed, because it could not find a port number from the (non-existent)
5345 transport. It now assumes port 25 in this circumstance.
5347 21. Added the -t option to exigrep.
5349 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5350 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5351 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5352 them is set, LOOKUP_LSEARCH is forced.
5354 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5356 24. Added sender and host information to the "rejected by local_scan()" log
5357 line; previously there was no indication of these.
5359 25. Added .include_if_exists.
5361 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5362 Linux. It turns out that not all file systems support this. Apparently some
5363 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5364 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5365 not supported on the file descriptor, is now ignored when Exim is trying to
5366 sync a directory. This applies only to Linux.
5368 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5370 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5371 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5374 29. Added support for maildirsize files from supplied patch (modified a bit).
5376 30. The use of :fail: followed by an empty string could lead Exim to respond to
5377 sender verification failures with (e.g.):
5379 550 Verification failed for <xxx>
5380 550 Sender verify failed
5382 where the first response line was missing the '-' that indicates it is not
5383 the final line of the response.
5385 31. The loop for finding the name of the user that called Exim had a hardwired
5386 limit of 10; it now uses the value of finduser_retries, which is used for
5387 all other user lookups.
5389 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5392 33. Exim was neglecting to zero errno before one call of strtol() when
5393 expanding a string and expecting an integer value. On some systems this
5394 resulted in spurious "integer overflow" errors. Also, it was casting the
5395 result into an int without checking.
5397 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5398 did not work. The code looks as if it has *never* worked, though it appears
5399 to have been documented since at least releast 1.62. I have made it work.
5401 35. The "timeout_DNS" error in retry rules, also documented since at least
5402 1.62, also never worked. As it isn't clear exactly what this means, and
5403 clearly it isn't a major issue, I have abolished the feature by treating it
5404 as "timeout", and writing a warning to the main and panic logs.
5406 36. The display of retry rules for -brt wasn't always showing the error code
5409 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5410 timeout_connect_A, timeout_connect_MX.
5412 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5413 to the empty sender.
5415 39. The daemon was not analysing the content of -oX till after it had closed
5416 stderr and disconnected from the controlling terminal. This meant that any
5417 syntax errors were only noted on the panic log, and the return code from
5418 the command was 0. By re-arranging the code a little, I've made the
5419 decoding happen first, so such errors now appear on stderr, and the return
5420 code is 1. However, the actual setting up of the sockets still happens in
5421 the disconnected process, so errors there are still only recorded on the
5424 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5425 connections (as happens on some IP stacks) was logged at start up time as
5426 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5427 it from "IPv6 and IPv4", which means that two separate sockets are being
5430 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5431 says whether AF_INET or AF_INET6 was passed as an argument.
5433 42. Exiwhat output was messed up when time zones were included in log
5436 43. Exiwhat now gives more information about the daemon's listening ports,
5437 and whether -tls-on-connect was used.
5439 44. The "port" option of the smtp transport is now expanded.
5441 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5442 silently ignored. Now an error message is written to the main and panic
5445 46. There's a new ACL modifier called "logwrite" which writes to a log file
5446 as soon as it is encountered.
5448 47. Added $local_user_uid and $local_user_gid at routing time.
5450 48. Exim crashed when trying to verify a sender address that was being
5453 49. Exim was recognizing only a space character after ".include". It now also
5454 recognizes a tab character.
5456 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5457 extracting the relevant information from the specification. The man page no
5458 longer contains scrambled data for the -d option, and I've added a section
5459 at the front about calling Exim under different names.
5461 51. Added "extra_headers" argument to the "mail" command in filter files.
5463 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5466 53. Installed eximstats 1.29.
5468 54. Added transport_filter_timeout as a generic transport option.
5470 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5471 Cc: header lines. This was required by RFC 822, but it not required by RFC
5474 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5475 incoming messages that did not have them. Now it does so only if the
5476 message originates locally, that is, if there is no associated remote host
5477 address. When Resent- header lines are present, this applies to the Resent-
5478 lines rather than the non-Resent- lines.
5480 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5481 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5483 58. Messages for configuration errors now include the name of the main
5484 configuration file - useful now that there may be more than one file in a
5485 list (.included file names were always shown).
5487 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5488 for those rare installations that do not start the daemon as root or run it
5489 setuid root. I've cut out the call to initgroups() if the daemon is not
5492 60. The Exim user and group can now be bound into the binary as text strings
5493 that are looked up at the start of Exim's processing.
5495 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5497 62. Added $mailstore_basename variable.
5499 63. Installed patch to sieve.c from Michael Haardt.
5501 64. When Exim failed to open the panic log after failing to open the main log,
5502 the original message it was trying to log was written to stderr and debug
5503 output, but if they were not available (the usual case in production), it
5504 was lost. Now it is written to syslog before the two lines that record the
5505 failures to open the logs.
5507 65. Users' Exim filters run in subprocesses under the user's uid. It is
5508 possible for a "deliver" command or an alias in a "personal" command to
5509 provoke an address rewrite. If logging of address rewriting is configured,
5510 this fails because the process is not running as root or exim. There may be
5511 a better way of dealing with this, but for the moment (because 4.30 needs
5512 to be released), I have disabled address rewrite logging when running a
5513 filter in a non-root, non-exim process.
5519 1. The buildconfig auxiliary program wasn't quoting the value set for
5520 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5521 not defined. This bug was masked in 4.22 by the effect that was fixed in
5524 2. Some messages that were rejected after a message id was allocated were
5525 shown as "incomplete" by exigrep. It no longer does this for messages that
5526 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5528 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5529 have allow_domain_literals set, the ID did not get logged in the <= line.
5530 Domain literals are now always recognized in Message-ID: header lines.
5532 4. The first argument for a ${extract expansion item is the key name or field
5533 number. Leading and trailing spaces in this item were not being ignored,
5534 causing some misleading effects.
5536 5. When deliver_drop_privilege was set, single queue runner processes started
5537 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5538 same command in the process it spins off) were not dropping privilege.
5540 6. When the daemon running as "exim" started a queue runner, it always
5541 re-executed Exim in the spun-off process. This is a waste of effort when
5542 deliver_drop_privilege is set. The new process now just calls the
5543 queue-runner function directly.
5549 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5552 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5553 ignored. Though the use of -oP was forcing the writing of a pid file, it
5554 was always written to the default place.
5556 3. If the message "no IP address found for host xxxx" is generated during
5557 incoming verification, it is now followed by identification of the incoming
5558 connection (so you can more easily find what provoked it).
5560 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5562 5. Added some features to "harden" Exim a bit more against certain attacks:
5564 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5565 be put in Local/Makefile. This is like the never_users runtime option,
5566 but it cannot be overridden. The default setting is "root".
5568 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5569 prefix string with which any file named in a -C command line option
5572 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5573 is retained for -C and -D only if the caller of Exim is root. Without
5574 it, the exim user may also use -C and -D and retain privilege.
5576 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5577 command line option is disabled.
5579 6. Macro names set by the -D option must start with an upper case letter, just
5580 like macro names defined in the configuration file.
5582 7. Added "dereference=" facility to LDAP.
5584 8. Two instances of the typo "uknown" in the source files are fixed.
5586 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5587 the Configure-Makefile script screwed up while processing it.
5589 10. Incorporated PCRE 4.4.
5591 11. The SMTP synchronization check was not operating right at the start of an
5592 SMTP session. For example, it could not catch a HELO sent before the client
5593 waited for the greeting. There is now a check for outstanding input at the
5594 point when the greeting is written. Because of the duplex, asynchronous
5595 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5596 way, but not yet received, when the check is performed.
5598 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5599 on TCP/IP sockets, because this apparently causes some broken clients to
5602 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5603 unchanged) from the Cygwin maintainer.
5605 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5606 was supported instead of testing for mbx. Effectively a typo.
5608 15. The spa authenticator server code was not checking that the input it
5609 received was valid base64.
5611 16. The debug output line for the "set" modifier in ACLs was not showing the
5612 name of the variable that was being set.
5614 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5616 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5617 lookup on its own. Something else was needed to trigger the lookup. For
5618 example, a match in host_lookup or the need for a host name in a host list.
5619 Now, if $sender_host_name is referenced and the host name has not yet been
5620 looked up, a lookup is performed. If the lookup fails, the variable remains
5621 empty, and $host_lookup_failed is set to "1".
5623 19. Added "eqi" as a case-independent comparison operator.
5625 20. The saslauthd authentication condition could segfault if neither service
5626 nor realm was specified.
5628 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5629 error message that was logged was misleading, and incoming SMTP
5630 connections were dropped. The message is now more accurate, and temporary
5631 errors are given to SMTP connections.
5633 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5634 (except RSET) with a 421 error, until QUIT is received. However, it was
5635 failing to send a response to QUIT.
5637 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5638 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5639 ACL is now run after the helo_try_verify_hosts code.
5641 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5642 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5643 case-independent, so other case variants are also recognized). Apparently
5644 some systems use these upper case variants.
5646 25. If more than two messages were waiting for the same host, and a transport
5647 filter was specified for the transport, Exim sent two messages over the
5648 same TCP/IP connection, and then failed with "socket operation on non-
5649 socket" when it tried to send the third.
5651 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5653 27. The extern definition of crypt16() in expand.c was not being excluded when
5654 the OS had its own crypt16() function.
5656 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5657 as a preferred synonym for return_size_limit, both as an option and as an
5660 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5662 30. Changed the default configuration ACL to relax the local part checking rule
5663 for addresses that are not in any local domains. For these addresses,
5664 slashes and pipe symbols are allowed within local parts, but the sequence
5665 /../ is explicitly forbidden.
5667 31. SPA server authentication was not clearing the challenge buffer before
5670 32. log_message in a "warn" ACL statement was writing to the reject log as
5671 well as to the main log, which contradicts the documentation and doesn't
5672 seem right (because no rejection is happening). So I have stopped it.
5674 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5675 However, I am unable to do any testing of this.
5677 34. Fixed an infelicity in the appendfile transport. When checking directories
5678 for a mailbox, to see if any needed to be created, it was accidentally
5679 using path names with one or more superfluous leading slashes; tracing
5680 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5682 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5683 contents are added to the log line that is written for every discarded
5684 recipient. (Previously a log_message setting was ignored.)
5686 36. The ${quote: operator now quotes the string if it is empty.
5688 37. The install script runs exim in order to find its version number. If for
5689 some reason other than non-existence or emptiness, which it checks, it
5690 could not run './exim', it was installing it with an empty version number,
5691 i.e. as "exim-". This error state is now caught, and the installation is
5694 38. An argument was missing from the function that creates an error message
5695 when Exim fails to connect to the socket for saslauthd authentication.
5696 This could cause Exim to crash, or give a corrupted message.
5698 39. Added isip, isip4, and isip6 to ${if conditions.
5700 40. The ACL variables $acl_xx are now saved with the message, and can be
5701 accessed later in routers, transports, and filters.
5703 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5704 strings in the file are not string-expanded.
5706 42. If a MAIL command specified a SIZE value that was too large to fit into an
5707 int variable, the check against message_size_limit failed. Such values are
5708 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5709 day this will have to be increased, but I don't think I want to be around
5710 when emails are that large.
5717 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5718 iconv() is not standard in FreeBSD.
5720 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5721 IPv6 enabled. The observed symptom was a segmentation fault on return from
5722 the function os_common_find_running_interfaces() in src/os.c.
5724 3. In the check_special_case() function in daemon.c I had used "errno" as an
5725 argument name, which causes warnings on some systems. This was basically a
5726 typo, since it was named "eno" in the comments!
5728 4. The code that waits for the clock to tick (at a resolution of some fraction
5729 of a second) so as to ensure message-id uniqueness was always waiting for
5730 at least one whole tick, when it could have waited for less. [This is
5731 almost certainly not relevant at current processor speeds, where it is
5732 unlikely to ever wait at all. But we try to future-proof.]
5734 5. The function that sleeps for a time interval that includes fractions of a
5735 second contained a race. It did not block SIGALRM between setting the
5736 timer, and suspending (a couple of lines later). If the interval was short
5737 and the sigsuspend() was delayed until after it had expired, the suspension
5738 never ended. On busy systems this could lead to processes getting stuck for
5741 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5742 process, before it forks any delivery processes. The open lookup caching
5743 mechanism meant that the open file or database connection was passed into
5744 the delivery process. The problem was that delivery processes always tidy
5745 up cached lookup data. This could cause a problem for the next delivery
5746 process started by the queue runner, because the external queue runner
5747 process does not know about the closure. So the next delivery process
5748 still has data in the lookup cache. In the case of a file lookup, there was
5749 no problem because closing a file descriptor in a subprocess doesn't affect
5750 the parent. However, if the lookup was caching a connection to a database,
5751 the connection was closed, and the second delivery process was likely to
5752 see errors such as "PGSQL: query failed: server closed the connection
5753 unexpectedly". The problem has been fixed by closing all cached lookups
5754 in a queue runner before running a delivery process.
5756 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5757 seem to have the "const" qualifier which it has on other OS. I've
5760 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5761 *of the same type* (client or server) with the same public name that an
5762 error should be diagnosed.
5764 9. When Exim looked up a host name for an IP address, but failed to find the
5765 original IP address when looking up the host name (a safety check), it
5766 output the message "<ip address> does not match any IP for NULL", which was
5767 confusing, to say the least. The bug was that the host name should have
5768 appeared instead of "NULL".
5770 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5771 user that is built into the binary, and the -C or -D options is used, root
5772 privilege is dropped before the configuration file is read. In addition,
5773 logging is switched to stderr instead of the normal log files. If the
5774 configuration then re-defines the Exim user, the unprivileged environment
5775 is probably not what is expected, so Exim logs a panic warning message (but
5778 However, if deliver_drop_privilege is set, the unprivileged state may well
5779 be exactly what is intended, so the warning has been cut out in that case,
5780 and Exim is allowed to try to write to its normal log files.
5786 1. smtp_return_error_details was not giving details for temporary sender
5787 or receiver verification errors.
5789 2. Diagnose a configuration error if two authenticators have the same public
5792 3. Exim used not to create the message log file for a message until the first
5793 delivery attempt. This could be confusing when incoming messages were held
5794 for policy or load reasons. The message log file is now created at the time
5795 the message is received, and an initial "Received" line is written to it.
5797 4. The automatically generated man page for command line options had a minor
5798 bug that caused no ill effects; however, a more serious problem was that
5799 the procedure for building the man page automatically didn't always
5800 operate. Consequently, release 4.20 contains an out-of-date version. This
5801 shouldn't happen again.
5803 5. When building Exim with embedded Perl support, the script that builds the
5804 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5805 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5807 6. The freeze_tell option was not being used for messages that were frozen on
5808 arrival, either by an ACL or by local_scan().
5810 7. Added the smtp_incomplete_transaction log selector.
5812 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5813 was accepting AUTH without a new EHLO.
5815 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5816 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5819 10. Logging of TCP/IP connections (when configured) now happens in the main
5820 daemon process instead of the child process, so that the TCP/IP connection
5821 count is more accurate (but it can never be perfect).
5823 11. The use of "drop" in a nested ACL was not being handled correctly in the
5824 outer ACL. Now, if condition failure induced by the nested "drop" causes
5825 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5826 or "require"), the connection is dropped.
5828 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5829 that yield "discard" can now be used with an "accept" or a "discard" verb,
5830 but an error is generated for any others (because I can't see a useful way
5831 to define what should happen).
5833 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5834 are now processed in the same way as lines in the Exim configuration file.
5835 In particular, continuation lines are supported.
5837 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5839 15. Added -ti meaning -t -i.
5841 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5842 domains, and warn by logging if others are found.
5844 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5845 returned from GIFCONF when Exim is trying to find the list of interfaces on
5846 a host. The code in os.c has been modified to copy each ifreq to an aligned
5847 structure in all cases.
5849 Also, in some cases, the returned ifreq's were being copied to a 'struct
5850 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5851 means the last couple of bytes of an IPv6 address could be chopped if the
5852 ifreq contained only a normal sockaddr (14 bytes storage).
5854 18. Named domain lists were not supported in the hosts_treat_as_local option.
5855 An entry such as +xxxx was not recognized, and was treated as a literal
5858 19. Ensure that header lines added by a DATA ACL are included in the reject log
5859 if the ACL subsequently rejects the message.
5861 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5862 MD5 (which is deprecated).
5864 21. When testing a filter file using -bf, Exim was writing a message when it
5865 took the sender from a "From " line in the message, but it was not doing so
5866 when it took $return_path from a Return-Path: header line. It now does.
5868 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5869 with a valid header line field name (a series of printing characters
5870 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5872 23. Changed "disc" in the source to "disk" to conform to the documentation and
5873 the book and for uniformity.
5875 24. Ignore Sendmail's -Ooption=value command line item.
5877 25. When execve() failed while trying to run a command in a pipe transport,
5878 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5879 could be confused with a return value of 69 from the command itself. This
5880 has been changed to 127, the value the shell returns if it is asked to run
5881 a non-existent command. The wording for the related log line suggests a
5882 non-existent command as the problem.
5884 26. If received_header_text expands to an empty string, do not add a Received:
5885 header line to the message. (Well, it adds a token one on the spool, but
5886 marks it "old" so that it doesn't get used or transmitted.)
5888 27. Installed eximstats 1.28 (addition of -nt option).
5890 28. There was no check for failure on the call to getsockname() in the daemon
5891 code. This can fail if there is a shortage of resources on the system, with
5892 ENOMEM, for example. A temporary error is now given on failure.
5894 29. Contrary to the C standard, it seems that in some environments, the
5895 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5896 program. Exim now does this explicitly; it affects the formatting of
5897 timestamps using strftime().
5899 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5900 complaints. I've now initialized all the variables, to avoid this.
5902 32. Header lines added by a system filter were not being "seen" during
5903 transport-time rewrites.
5905 33. The info_callback() function passed to OpenSSL is set up with type void
5906 (*)(SSL *, int, int), as described somewhere. However, when calling the
5907 function (actually a macro) that sets it up, the type void(*)() is
5908 expected. I've put in a cast to prevent warnings from picky compilers.
5910 34. If a DNS black list lookup found a CNAME record, but there were no A
5911 records associated with the domain it pointed at, Exim crashed.
5913 35. If a DNS black list lookup returned more than one A record, Exim ignored
5914 all but the first. It now scans all returned addresses if a particular IP
5915 value is being sought. In this situation, the contents of the
5916 $dnslist_value variable are a list of all the addresses, separated by a
5919 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5920 to accept a host name and all its aliases if the forward lookup for any of
5921 them yielded the IP address of the incoming connection. Now it accepts only
5922 those names whose forward lookup yields the correct IP address. Any other
5923 names are discarded. This closes a loophole whereby a rogue DNS
5924 administrator could create reverse DNS records to break through a
5925 wildcarded host restriction in an ACL.
5927 37. If a user filter or a system filter that ran in a subprocess used any of
5928 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5929 the wrong values were passed to the pipe command ($thisaddress had the
5930 value of $0, $0 had the value of $1, etc). This bug was introduced by
5931 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5933 38. Improved the line breaking for long SMTP error messages from ACLs.
5934 Previously, if there was no break point between 40 and 75 characters, Exim
5935 left the rest of the message alone. Two changes have been made: (a) I've
5936 reduced the minimum length to 35 characters; (b) if it can't find a break
5937 point between 35 and 75 characters, it looks ahead and uses the first one
5938 that it finds. This may give the occasional overlong line, but at least the
5939 remaining text gets split now.
5941 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5942 file descriptors might be low, and that setting 1000 would always raise it.
5943 It turns out that in some environments, the limit is already over 1000 and
5944 that lowering it causes trouble. So now Exim takes care not to decrease it.
5946 40. When delivering a message, the value of $return_path is set to $sender_
5947 address at the start of routing (routers may change the value). By an
5948 oversight, this default was not being set up when an address was tested by
5949 -bt or -bv, which affected the outcome if any router or filter referred to
5952 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5953 log only when "message" or "log_message" are set. However, if one of the
5954 conditions was an address verification, or a call to a nested ACL, the
5955 messages generated by the underlying test were being passed through. This
5956 no longer happens. The underlying message is available in $acl_verify_
5957 message for both "message" and "log_message" expansions, so it can be
5958 passed through if needed.
5960 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5961 new expansion $bh_ to give the encoded byte string without charset
5962 translation. Translation happens only if iconv() is available; HAVE_ICONV
5963 indicates this at build time. HEADERS_CHARSET gives the charset to
5964 translate to; headers_charset can change it in the configuration, and
5965 "headers charset" can change it in an individual filter file.
5967 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5968 that creates RFC 2047 encoded "words" labels them as that charset instead
5969 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5970 expansion operator; (ii) when Exim creates a From: line for a local
5971 message; (iii) when a header line is rewritten to include a "phrase" part.
5973 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5974 buggy, causing it to skip the first lines of messages whose message ID
5975 ended in 'D'. This would not have bitten before Exim release 4.14, because
5976 message IDs were unlikely to end in 'D' before then. The effect was to have
5977 incorrect size information for certain domains.
5979 45. #include "config.h" was missing at the start of the crypt16.c module. This
5980 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5983 46. If there was a timeout during a "random" callout check, Exim treated it as
5984 a failure of the random address, and carried on sending RSET and the real
5985 address. If the delay was just some slowness somewhere, the response to the
5986 original RCPT would be taken as a response to RSET and so on, causing
5987 mayhem of various kinds.
5989 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5990 when I implemented it. It didn't allow for the fact that some option values
5991 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5992 the right test for positive values.
5994 48. Domain names in DNS records are case-independent. Exim always looks them up
5995 in lower case. Some resolvers return domain names in exactly the case they
5996 appear in the zone file, that is, they may contain uppercase letters. Not
5997 all resolvers do this - some return always lower case. Exim was treating a
5998 change of case by a resolver as a change of domain, similar to a widening
5999 of a domain abbreviation. This triggered its re-routing code and so it was
6000 trying to route what was effectively the same domain again. This normally
6001 caused routing to fail (because the router wouldn't handle the domain
6002 twice). Now Exim checks for this case specially, and just changes the
6003 casing of the domain that it ultimately uses when it transmits the message
6006 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6009 50. If a filter generated a file delivery with a non-absolute name (possible if
6010 no home directory exists for the router), the forbid_file option was not
6013 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6014 the existing equality matching.
6016 52. Exim was using ints instead of ino_t variables in some places where it was
6017 dealing with inode numbers.
6019 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6020 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6021 called TMPDIR, and if it finds it is different, it changes its value.
6023 54. The smtp_printf() function is now made available to local_scan() so
6024 additional output lines can be written before returning. There is also an
6025 smtp_fflush() function to enable the detection of a dropped connection.
6026 The variables smtp_input and smtp_batched_input are exported to
6029 55. Changed the default runtime configuration: the message "Unknown user"
6030 has been removed from the ACL, and instead placed on the localuser router,
6031 using the cannot_route_message feature. This means that any verification
6032 failures that generate their own messages won't get overridden. Similarly,
6033 the "Unrouteable address" message that was in the ACL for unverifiable
6034 relay addresses has also been removed.
6036 56. Added hosts_avoid_esmtp to the smtp transport.
6038 57. The exicyclog script was not checking for the esoteric option
6039 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6040 will work only if exicyclog is run under the appropriate euid.
6042 58. Following a discussion on the list, the rules by which Exim recognises line
6043 endings on incoming messages have been changed. The -dropcr and drop_cr
6044 options are now no-ops, retained only for backwards compatibility. The
6045 following line terminators are recognized: LF CRLF CR. However, special
6046 processing applies to CR:
6048 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6049 nor a local message in the state where . is a terminator.
6051 (ii) If a bare CR is encountered in a header line, an extra space is added
6052 after the line terminator so as not to end the header. The reasoning
6053 behind this is that bare CRs in header lines are most likely either
6054 to be mistakes, or people trying to play silly games.
6056 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6057 was being incorrectly given as 18 bytes larger than it should have been.
6058 This is a VOB (very old bug).
6060 60. This may never have affected anything current, but just in case it has:
6061 When the local host is found other than at the start of a list of hosts,
6062 the local host, those with the same MX, and any that follow, are discarded.
6063 When the list in question was part of a longer list of hosts, the following
6064 hosts (not currently being processed) were also being discarded. This no
6065 longer happens. I'm not sure if this situation could ever has previously
6068 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6071 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6072 header. This is something that is recommended in a new Internet Draft, and
6073 is something that is documented as being done by Sendmail. There are two
6074 possible values. For messages generated by the autoreply transport, Exim
6077 Auto-Submitted: auto-replied
6079 whereas for all other generated messages (e.g. bounces) it adds
6081 Auto-Submitted: auto-generated
6083 63. The "personal" condition in filters now includes a test for the
6084 Auto-Submitted: header. If it contains the string "auto-" the message it
6085 not considered personal.
6087 64. Added rcpt_include_affixes as a generic transport option.
6089 65. Added queue_only_override (default true).
6091 66. Added the syslog_duplication option.
6093 67. If what should have been the first header line of a message consisted of
6094 a space followed by a colon, Exim was mis-interpreting it as a header line.
6095 It isn't of course - it is syntactically invalid and should therefore be
6096 treated as the start of the message body. The misbehaviour could have
6097 caused a number of strange effects, including loss of data in subsequent
6098 header lines, and spool format errors.
6100 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6101 client host had authenticated. This control can now be exercised by an ACL
6102 for more flexibility.
6104 69. By default, callouts do not happen when testing with -bh. There is now a
6105 variant, -bhc, which does actually run the callout code, including
6106 consulting and updating the callout cache.
6108 70. Added support for saslauthd authentication, courtesy of Alexander
6111 71. If statvfs() failed on the spool or log directories while checking their
6112 size for availability, Exim confusingly gave the error "space shortage".
6113 Furthermore, in debugging mode it crashed with a floating point exception.
6114 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6115 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6116 serious problem, Exim now writes to the main and panic logs when this
6117 happens, with details of the failure. It then refuses to accept the
6118 incoming message, giving the message "spool directory problem" or "log
6119 directory problem" with a 421 code for SMTP messages.
6121 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6122 0, 1, and 2 exist, because some OS complain for execs without them (see
6123 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6124 descriptors. However, the code omitted to check that the open succeeded,
6125 causing mysterious errors if for some reason the permissions on /dev/null
6126 got screwed. Now Exim writes a message to the main and panic logs, and
6127 bombs out if it can't open /dev/null.
6129 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6130 interact so that it is all more flexible. It is supposed to remain
6131 backwards compatible. Also added extra_local_interfaces.
6133 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6134 to bomb out with an assertion failure - to the client this appears as a
6135 connection drop. This problem occurs in the part of the code that was taken
6136 from the Samba project. Fortunately, the assertion is in a very simple
6137 function, so I have fixed this by reproducing the function inline in the
6138 one place where it is called, and arranging for authentication to fail
6139 instead of killing the process with assert().
6141 75. The SPA client code was not working when the server requested OEM rather
6142 than Unicode encoding.
6144 76. Added code to make require_files with a specific uid setting more usable in
6145 the case where statting the file as root fails - usually a non-root-mounted
6146 NFS file system. When this happens and the failure is EACCES, Exim now
6147 forks a subprocess and does the per-uid checking as the relevant uid.
6149 77. Added process_log_path.
6151 78. If log_file_path was not explicitly set, a setting of check_log_space or
6152 check_log_inodes was ignored.
6154 79. If a space check for the spool or log partitions fails, the incident is now
6155 logged. Of course, in the latter case the data may get lost...
6157 80. Added the %p formatting code to string_format() so that it can be used to
6158 print addresses in debug_print(). Adjusted all the address printing in the
6159 debugging in store.c to use %p rather than %d.
6161 81. There was a concern that a line of code in smtp_in.c could overflow a
6162 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6163 initially expressed, the concern was not well-founded, because trailing
6164 spaces are removed early. However, if the trailing spaces were followed by
6165 a NULL, they did not get removed, so the overflow was possible. Two fixes
6168 (a) I re-wrote the offending code in a cleaner fashion.
6169 (b) If an incoming SMTP command contains a NULL character, it is rejected
6172 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6173 runs initgroups(), so that if the Exim user is in any additional groups,
6174 they will be used during message reception.
6180 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob